Conference Agenda

Overview and details of the sessions of this conference. Please select a date or room to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

Session Overview
Date: Wednesday, 18/Mar/2015
9:00Begin Check-In
11:30 - 13:00Workshop 1
Room 158 

Introduction to Network Analytics

Cathleen M. Stuetzer

TU Dresden, Germany

1.) Title of the workshop

Introduction to Network Analytics

2.) The workshop includes 1.5 h of instruction (no break).

3.) Target groups

For those who are interested in network analytics and social network theory. The workshop is targeted especially for people who have no experience with network analytics. The material is suitable for both researchers and practitioners.

4.) Is the workshop geared at an exclusively German or an international audience?

International audience

5.) Workshop language

English/ German (depending on the audience)

6.) Description of the content of the workshop

This course provides an introduction to theories and methods for network analytics as a perspective in social sciences and aims at newcomers to the field. Network analytics refers to the process of extracting relational data for the purpose of exploring hidden structures within (virtual) organizations, and knowledge networks. The course aims to be a practical introduction and focuses on answering questions as what network analytics mean, where network analytics come from, and how network analytic concepts can be applied. The course demonstrates the importance of social network perspective in various research fields and their relevance in practice. The workshop begins with the discussion of fundamental principles of social network theory and goes on to the description of network data, common measurements, and analysis of network structures. The course presents techniques of data collection and provides an overview on suitable instruments for network analytics. Resources for further study will be listed.

7.) Goals of the workshop

The workshop attendees are introduced to understand the composition, the structure, and the dynamics of (social) networks and how to apply the knowledge to study human behavior in the digital age. The workshop aims at providing a well-organized introduction to theories and methods for network analytics and their relevance in practice.

8.) Necessary prior knowledge of participants

The introduction to key concepts requires no prior knowledge.

9.) Literature that participants need to read prior to participation

The course has no special requirements for participation.

10.) Recommended additional literature

Hennig, Marina, Ulrik Brandes, Juergen Pfeffer, and Ines Mergel. (2012). Studying Social Networks: A Guide to Empirical Research. Frankfurt: Campus Verlag.

Wasserman, S., and Faust, K. (1994). Social Network Analysis: Methods and Applications. Cambridge, ENG and New York: Cambridge University Press.

Freeman, Linton (2006). The Development of Social Network Analysis. Vancouver: Empirical Press.

Freeman, Linton (1979). Centrality in social networks: Conceptual clarification. Social Networks 1:215-239.

11.) Information about the instructor

Cathleen M. Stuetzer is a research assistant at the Institute for Sociology at the Johannes Gutenberg University Mainz (JGU), and a post-doctoral researcher at the Department of Educational Technology at the University of Technology in Dresden. In her empirical work, she focuses on the issues as „Evolution of Networks and Mechanisms of Elite Building in Science” and “Social Networks and Scientific Careers”. She is working as data scientist in diverse projects with the view to the impact of cooperation and collaboration in the social media age and concentrating on the handling of big data and development of quantitative and qualitative (online) research methods.

She earned her PhD (thesis) at TU Dresden in 2013 in the field of educational research with the topic “Distribution of Information and Knowledge in Collaborative Learning Systems”. She was honored with the GOR Thesis Award 2014 by the DGOF and nominated for the Young Researcher Award at WWW 2014 Conference in Seoul. Since 2011 she is engaging as reviewer at GOR and since 2014 an active member at DGOF (e.g. vice program chair GOR 15, program chair GOR16).


Dr. Cathleen M. Stuetzer

Dresden University of Technology

Faculty of Education / Department of Educational TechnologyOffice: Strehlener Straße 22/24, Room: 451, D-01069 Dresden, Germany

Phone: +49 351 463 36261



12.) Maximum number of participants


13.) Will participants need to bring their own devices in order to be able to access the Internet? Will they need to bring anything else to the workshop?

Participants should bring their own mobile devices.

13:30 - 16:30Workshop 2
Room 149 

R 101: Introduction to R, an open-source software for statistical computing

Frederik Funke

(1) LINK Institut (2), Germany

1.) Title of the workshop

R 101: Introduction to R, an open-source software for statistical computing

2.) The workshop includes 2.5 hours of instruction.

3.) Target groups

Persons who are interested in a first hands-on introduction to the free software R (see, which can be used as alternative to Excel and statistical packages like IBM SPSS Statistics or Stata.

4.) Is the workshop geared at an exclusively German or an international audience?

International audience

5.) Workshop language

Depends on audience: In German if all participants speak German, otherwise in English.

6.) Description of the content of the workshop

A basic introduction to descriptive statistical analyzes and to data visualization. Participants will become acquainted with basic procedures, guidelines, and general rules.

7.) Goals of the workshop

After the workshop participants will be able to perform basic operations (e.g., import of data sets, computation of new variables), use important functions (e.g., summary statistics, crosstabs), and build simple charts (e.g., bar charts, pie charts, box plots).

8.) Necessary prior knowledge of participants


9.) Literature that participants need to read prior to participation


10.) Recommended additional literature

In English:

Adler, J. (2012). R in a nutshell: A desktop quick reference (2nd ed.). Sebastopol, CA: O’Reilly.

Crawley, M. J. (2012). The R book (2nd ed.). Hoboken, NJ: Wiley.

Field, A., Miles, J., & Field, Z. (2012). Discovering Statistics using R. Thousand Oaks, CA: Sage.

In German:

Groß, J. (2010). Grundlegende Statistik mit R: Eine anwendungsorientierte Einführung in die Verwendung der Statistik Software R. Wiesbaden: Vieweg+Teubner.

Luhmann, M. (2013). R für Einsteiger: Einführung in die Statistiksoftware für Sozialwissenschaftler (3. Auflage). Weinheim: Beltz.

Manderscheid, K. (2012). Sozialwissenschaftliche Datenanalyse mit R. Wiesbaden: VS.

11.) Information about the instructor

Dr. Funke (see is freelance consultant and trainer for statistics and surveys, focussing on questionnaire design, design of experiments, and methods of data analysis. He offers trainings on R and IBM SPSS Statistics and helps to optimize workflows. Furthermore, he works as senior project manager online research at LINK Institut and as lecturer in higher education.

12.) Maximum number of participants

15 participants

13.) Will participants need to bring their own devices in order to be able to access the Internet? Will they need to bring anything else to the workshop?

Own laptop computers (R runs on Linux and on MacOS as well as on Windows) are required. Internet access could help but is not mandatory. All participants have to install software before the workshop. A detailed description on how to install the necessary software will be provided two weeks before the workshop.
13:30 - 16:30Workshop 3
Room 158 

The art of asking questions

Jon Puleston

Lightspeed GMI, United Kingdom

1.) Title of the workshop

The art of asking questions

2.) The workshop includes 2.5 hours of instruction.

3.) Target groups

Anyone who has ever the occasion to design an online survey and is looking to update their approach

4.) Is the workshop geared at an exclusively German or an international audience?

Anyone who can understand my mumbling English – of international appeal

5.) Workshop language


6.) Description of the content of the workshop

This workshop will take a detailed look at the nuts and bolts of question design exploring the key biasing effects that influence how respondents answer questions in surveys; it will look at the techniques you can use to get respondents to think in different ways and use different parts of their brains to answer questions and how you can employ reward mechanics in surveys to encourage more active participation; it will explore some of the new and emerging techniques and teach you all the most important considerations to ensure that your questions are mobile friendly.

7.) Goals of the workshop

Educate attendees on best practice for questionnaire design and inspire

8.) Necessary prior knowledge of participants


9.) Literature that participants need to read prior to participation


10.) Recommended additional literature


11.) Information about the instructor:

Jon Puleston is VP of Innovation at Lightspeed GMI, where he heads an international team specialising in the design of surveys. Over the last few years he and his team have won multiple awards for their ground breaking work exploring online research methodology. He a leading international speaker and blog writer and runs workshops and master classes on effective survey design. You can find out more on his blog

12.) Maximum number of participants


13.) Will participants need to bring their own devices in order to be able to access the Internet? Will they need to bring anything else to the workshop?

No need to bring anything
17:00 - 19:00DGOF members meeting
Session Chair: Otto Hellwig, respondi AG

The members meeting of the Deutsche Gesellschaft für Online-Forschung (Germany Society for Online Research is open to all members. Guests are welcome. The meeting will be in German.

Room 248 
19:30GOR Get-Together


Im Schnörres

Ferkulum 30, 50678 Cologne

The GOR 15 Get-Together is open to everyone with a valid conference or workshop ticket!

Sponsored by respondi

Date: Thursday, 19/Mar/2015
8:00Begin Check-In
9:00 - 10:15Opening & Keynote 1
Aula (Mainzer Str. 5) 

What can advertising teach us about effective survey design?

Jon Puleston

Lightspeed GMI, United Kingdom

The presentation explores what we can learn from other commercial creative art forms to design better pieces of research. It will examine how the thinking used to create great advertising can be used to create great surveys; how the techniques used by the gaming industry to make computer games more addictive can be used to help make surveys more compelling and how if you think more like a publisher rather than a research you can create more meaningful surveys. This will be a show and tell session with lots of “how to” examples that I hope will inspire you to go away and produce some great surveys in the future.

Jon Puleston is VP of Innovation at Lightspeed GMI, where he heads an international team specialising in the design of surveys. Over the last few years his team have won multiple awards for their ground breaking work exploring online research methodology particularly in the field of the gamification of surveys. He is a leading international speaker and blog writer and runs workshops and master classes on effective survey design. You can find out more on his blog
10:15 - 10:45Coffee Break
10:45Track A: Online Research Methodology and Internet Surveys

Sponsored by GESIS

10:45Track B: Internet and Society

Programme Partner: HIIG

10:45Track C: Social Media Research
10:45Track D: Angewandte Online-Forschung (Applied Online Research)

Programme Partner:

10:45 - 11:45A 2: Measurement
Session Chair: Oliver Bastian Tristan Franken, TU Dresden
Room 248 

Click, Touch, Slide: Impact of the Implementation of Graphical Rating Scales on Data Quality in Mobile and Desktop Settings

Frederik Funke1,2, Vera Toepoel3; 2LINK Institut, Germany; 3Utrecht University, The Netherlands

Relevance & Research Question: Rating scales (e.g., agree-disagree scales) can be implemented in different ways. Besides standard HTML radio buttons there are different graphical rating scales available. This study focuses on slider scales (e.g., Funke, Reips & Thomas, 2011) and visual analogue scales (VAS, e.g., Couper et al., 2006), two scales that differ in the way they are operated. VAS and radio buttons are operated by clicking only, which makes a marker appear on the previously empty scale. Slider scales consist of a handle that is visible in the beginning and has to be moved. There are different possible implementations: Sliders can either be operated by sliding only or by a combination of sliding and clicking. This study is aimed at identifying the best of implementing way graphical rating scales.

Methods & Data: The sample consisted of N = 4180 respondents that were randomly assigned to a questionnaire consisting of either radio buttons, slide-only sliders, click and slide sliders, or VAS. As the actual use could also depend on the respondent’s device, a comparison between desktop computers (N = 1406), smart phones (N = 1372), and tablets (N = 1402) was made. As second experimental factor the response scale consisted either of 5, 7, or 11 options.

Results: Item nonresponse was highest with sliders (7.3% and 7.7%) followed by VAS (5.5%) and it was lowest with radio buttons (3.8%). Especially respondents with a low formal education produced missing data with sliders that could only be slid but not clicked on. No difference was found in mean ratings, response time, and evaluation of the questionnaire. The way sliders were operated did affect the results. The number of response options did not affect the results systematically.

Added Value: Overall, higher rates of item nonresponse argue against the use of slider scales, especially against those that can only be operated by sliding. Overall, it is recommended to use radio buttons or VAS as graphical rating scales.

Dynamic Drag-and-Drop Rating Scales in Web Surveys

Tanja Kunz

Darmstadt University of Technology, Germany

Relevance & Research Question: In Web surveys, rating scales are typically presented in grid (or matrix) questions. Besides benefits such as the neat arrangement and efficient processing of rating scale items, grid questions carry an increased risk of respondents relying on cognitive shortcuts in order to reduce their cognitive and navigational effort. Even though in Web surveys a wide range of visual and dynamic features is available for the design of survey questions, new types of rating scale designs beyond grid questions using conventional radio buttons have rarely been used yet. In this study, two different rating scale procedures using drag-and-drop as a more interactive data input method are applied: Respondents need to drag the response options towards the rating scale items (“drag-response”), or in reverse, the rating scale items towards the response options (“drag-item”). Both drag-and-drop scales aim at encouraging respondents to process rating scale items more attentively and carefully instead of simply relying on cognitive shortcuts.

Methods & Data: In two randomized field experimental Web surveys conducted among university applicants (n=5,977 and n=7,395), various between-subjects designs were implemented to assess the effectiveness of the drag-response and drag-item scale in preventing the respondents’ susceptibility to cognitive shortcuts compared to a standard grid question in terms of different systematic response tendencies commonly encountered with rating scales such as nondifferentiation and primacy effects. Furthermore, item missing data and response times were examined.

Results: Findings revealed that the quality of answers to rating scales may profit from the respondents’ higher attentiveness and carefulness in both drag-and-drop scales which is reflected in a decreased susceptibility to systematic response tendencies. At the same time, however, results also showed that both drag-and-drop scales entail a higher extent of respondent burden compared to conventional radio button scales as indicated by increases in item missing data and longer response times.

Added Value: This study provides a comprehensive examination of the potentials and limitations of new drag-and-drop procedures as an interactive data input method for rating scales in Web surveys. In addition, findings contribute to a better understanding of the cognitive processing of rating scales in Web surveys.
Kunz-Dynamic Drag-and-Drop Rating Scales in Web Surveys-186.pdf

Positioning of Clarification Features in Open Frequency and Open Narrative Questions

Metzler Anke, Marek Fuchs

Darmstadt University of Technology, Germany

Relevance & Research Question: The lack of interviewer assistance increases response burden in Web surveys and enlarges the risk that respondents misinterpret survey questions. Clarification features and instructions are seen as an effective means of improving question understanding and response behavior. However, clarification features often suffer from limited attention. Thus, they need to be positioned exactly where they are needed (Dillman, 2000). In the past it has been suggested to place clarification features after the question text. By contrast, recent findings from an eye-tracking study indicated that clarification features concerning question meaning are particularly noticed when they are presented before the question text whereas formatting instructions should be placed after the response options (Kunz & Fuchs, 2012). This study aims to answer the question whether optimal positioning of clarification features depends on the cognitive stage addressed by the instructions.

Methods & Data: In four randomized field experimental Web surveys a between-subjects design was implemented to test the effectiveness of three different positions of clarification features related to open frequency and open narrative questions: after the question text, before the question text and after the response options. Clarification features concerning question meaning, the retrieval of a response and the formatting of the response have been tested.

Results: Results indicate that clarification features regardless of the processing stage of the question-answer process addressed are best positioned after the question or after the response options. Instructions placed before the question stem yielded the smallest effect.

Added Value: The use of clarification features in Web surveys has a positive effect on survey responses and helps improve data quality. The optimal position of instructions does not depend on the cognitive stage of the question-answer process addressed. Survey researchers should avoid placing clarification features before the question since this position seems to be least efficient. However, results also indicate that clarification features positioned after the input field are similar effective as compared to clarification features positioned after the question stem.

Anke-Positioning of Clarification Features in Open Frequency and Open Narrative Questions-155.pdf
10:45 - 11:45B 2: Citizen Science
Session Chair: Martin Welker, Technische Universität Braunschweig
Room 147 

Get Them Involved: Motivational Strategies of Citizen Science Platforms

Kaja Scheliga1, Benedikt Fecher1,2, Sascha Friesike1,4, Cornelius Puschmann1,3

1HIIG, Germany; 2DIW, Germany; 3Zeppelin Universität, Germany; 4Universität Würzburg, Germany

Relevance & Research Question:

Citizen science describes research activities (e.g., data collection) that are conducted by people who are not professional scientists. Thanks to the internet, citizen science has gained a new dimension: digital technologies offer possibilities to reach out to volunteer researchers, to pool together efforts and to make the results visible. Moreover, the collective power of the crowd provides scientists with data that they could not have collected on their own. Integrating citizens into research projects is apparently a valuable strategy for scientists. Motivating volunteers to invest time in citizen science projects appears challenging.

● What are common strategies to motivate volunteers to partake in citizen science activities?

● What do volunteers get out of it?

● What does it take to build up and maintain a citizen science community?

● How can the quality of citizen scientists’ contributions be ensured?

Methods & Data:

The insights are based on 12 case studies of German citizen science platforms. We have studied the selected platforms, reviewed related materials and conducted qualitative interviews. The interviews were transcribed and thematically coded.


Based on our case studies, we found that motivational strategies revolve around: the task, learning, results, and/or community.

Some projects tap into an existing community and provide it with an infrastructure so as to channel the community’s efforts towards the project. Other projects need to build up a community around their project. We have found that important motivational factors are providing feedback, facilitating a learning experience and turning a task into a game. Moreover, while citizen science also has its long tail of participation, substantial contributions come from a few dedicated citizen science enthusiasts who provide the best quality contributions. We argue that in developing successful motivational strategies for citizen science projects these “heavy contributors” deserve a special role.

Added Value:

There is still little research on citizen science. This research project aims to contribute to a better understanding of the field by providing an overview of citizen science in Germany and discussing different motivational strategies.

Scheliga-Get Them Involved-130.pdf

Creating deliberation online: Developing a Citizen Science Strategy 2020 for Germany

Lisa Pettibone

Museum für Naturkunde Berlin, Germany

Relevance and research question

Citizen science is an increasingly popular approach to science, one that brings professional scientists together with volunteer scientists, scientific clubs, or interested laypeople. Citizen science has recently gained the attention of science policymakers, increasing the importance of developing a strategy to promote and support citizen science projects. Following the approach used by Socientize at the European level, the GEWISS (BürGEr schaffen WISSen, or “Citizens create knowledge”) project seeks to conduct an online consultation process to create a 2020 strategy for citizen science in Germany. This paper will present the deliberative approach used by both projects, report on initial results, and discuss the possibilities and challenges of using online deliberation methods for strategy development.

Methods and data

Over 200 comments were collected during the online consultation conducted by Socientize in 2014, in addition to offline input through stakeholder roundtables, participatory workshops, and telephone interviews. GEWISS seeks to collect at least this many comments through a two-stage online consultation, which will run in parallel to a series of dialogue events conducted throughout 2015. The GEWISS consultation for its Citizen Science Strategy 2020 for Germany will be based on a draft strategy document developed during a think tank event in Berlin in July 2014.


The Socientize consultation sought to engage diverse stakeholder groups to submit policy recommendations for a European citizen science strategy. Its consultation process received input from several sources, including an online consultation process. However, organizations found that not all participants were versed in policy language or able to contribute through the online tool. Future endeavors may thus require training for participants, additional analysis by organizers, and various input mechanisms.

Added value

This paper uses empirical data to assess the real-life challenges related to online consultation. In this way, it adds to our understanding of the real possibilities of online methods of consulting key stakeholders, as well as current challenges, in order that work begin to address them and increase the efficacy of online deliberation tools.
Pettibone-Creating deliberation online-197.pptx
10:45 - 11:45C 2: Business Analytics with Social Media I
Session Chair: Axel Maireder, GfK
Room 149 

Combining Survey and Social Media Monitoring in an Airline Customer Satisfaction Study

Charlotte Nau, Melanie Arens

TNS Infratest, Germany

Relevance & Research Question: Traditional customer satisfaction research with its heavy emphasis on surveys has come under pressure. There is a strong demand for surveys to become shorter, be conducted less frequently and at lower costs. At the same time, an increasing number of people share their experiences with products and services via social media, making the internet a continuously growing source of customer feedback. In this project, we investigate how to combine customer satisfaction surveys with social media monitoring for greater insight and efficiency.

Methods & Data: We conducted a two-component study focussing on a large European airline. The first component was a survey among 2,000 recent customers of that airline. It included measures of overall satisfaction and loyalty, detailed evaluations of different service aspects, and brief ratings of the airline’s major competitors. The second component was a social media monitoring where we collected online conversations about the airline over a three-month period, leading to around 130,000 relevant documents. A sample of 4,500 documents was coded manually.

Results: There was a strong relationship between how often a competitor airline was mentioned on social media and its attractiveness ratings in the survey, indicating that social media can serve as an early warning system for customer defection. Also, we found the volume of social buzz around a service topic (e.g. check-in) to not be predictive of its impact on customer satisfaction, suggesting that survey research remains necessary to estimate the urgency of issues. Qualitatively, the social data were able to explain why service aspects were rated positively or negatively.

Added Value: The combination of survey and social media data led to insights which were more comprehensive and detailed than what both of the methods would have revealed alone. While surveys are able to provide a strategic overview of a company’s customer relationships, social media can provide qualitative explanations for the ratings. On the other hand, social media data – which pour in unsystematically and in large quantities – are given context and structure by the customer survey, hence making them manageable. Finally, social data can replace some survey components altogether, allowing for shorter and less expensive interviews.
Nau-Combining Survey and Social Media Monitoring in an Airline Customer Satisfaction Study-137.pdf

Social ratings as the new currency of marketeers? – A comparison of influences from Likes and test seals on product ratings

Christian Bosau1, Liesa Henriette Pöppel1, Ulrich Rieger2

1Rheinische Fachhochschule Köln (RFH), Germany; 2MW Research GmbH, Germany

Relevance/Research-Question: In the so-called web 2.0 customers became more active and social ratings, e.g. Facebook-Likes, got steadily more important. Due to social psychological theories like normative and informational conformity (Deutsch & Gerard, 1955) it could be assumed that social ratings might influence customers attitudes (Bak & Keßler, 2012), maybe even more than official test seals, like “Stiftung Warentest”. However, the almost uncritically claim that good social ratings are always positive creating favorable outcomes for any marketeer has seldomly been tested yet. The interaction effect with test seals has not been analyzed at all.

Methods/Data: Therefore, this experimental 3x2x4-study compared the influence of the number of Likes (little, medium, high) and the test seal “Stiftung Warentest” (positive, negative) on product ratings (measured by quality rating, purchasing intention & customer recommendation) using four different products (2x shopping goods: TV, Laptop; 2x convenience goods: water, chocolate; repeated measurement). Potentially influencing variables (existing Facebook-Account, actual purchasing purpose, sociodemographics) were also measured. The products presented realistically in a web-based offer were rated by a representative sample (N=1170) drawn via a panel-database.

Results: As expected, the 2-way-ANOVAs showed a significant positive main effect for the quality seal for all 4 products. A clear positive main effect of Likes could not be found. Instead, the influence of Likes differed depending on a combination of the rated product, the actual purchasing purpose, the rating of the test seal and an existing facebook account; e.g. for TVs and Laptops with a negative test rating more Likes had indeed a more negative effect, this effect was even stronger for people who had an actual purchasing purpose and a facebook account. Thus, Likes had – other than expected – a complete reverse effect. For products with a positive test rating more Likes had almost none effect at all.

Added Value: The study tested the believe that is often taken for granted that a high number of Likes has always positive effects and showed, that this believe has to be discussed more critically. By combining several different products in one study it also showed that dependence on the product category.
Bosau-Social ratings as the new currency of marketeers – A comparison of influences-158.pdf

Brand Engagement Topologies on Instagram

Simeona Petkova

Amsterdam University of Applied Science, Netherlands, The

Brand engagement on social media has been mainly studied through its communicative potential and difference from offline marketing strategies (Jensen and Jepsen, 2006), through content analysis of nongovernmental organisations’ Facebook profiles (Walters et al. 2009) and Facebook messages per specific industry (Kwok and Yu, 2012). The most comprehensive typology on brand engagement on Facebook pages across industries have been provided by Coursaris et al. (2013) that also appeals for the creation of more engagement typologies on other social media platforms. This research takes upon this appeal and aims to provide a typology of brand engagement on Instagram by asking which are the most prominent engagement types detected through the profile tabs and the branded hashtags of the top three brands on Instagram.

We have coded the emerging types of brand engagement through the content and the interactions on the three most prominent brands profile tabs (Nike, Starbucks and Adidas Originals) and the images generated by Instagram users and tagged with the corresponded branded hashtags (#nike, #starbucks, #adidasoriginals). Our dataset consisted of 91 500 images (captured 27 October - 7 November from the branded walls and the API) from which we took 1% sample and registered the images’ content, caption, hashtags, comments, and the number of likes. We used a descriptive registration, complemented with open and thematic coding.

Preliminary results showed that engagement types vary per branded profile and branded hashtag posts. While the most occurring typologies on the branded profiles are campaigns (49%), promotion (19%) awareness (9%), the user engagement per branded hashtag is dominated by products and the ways they are used (respectively 35% and 27% of the provisional sample). Starbucks profile, for example, focuses on building community, running various themed campaigns, giving appreciation of its customers while the posts generated by Instagram users for the #starbucks are related to the products (featuring the products, mentioning the flavours).

Our findings signify that marketeers need to have a deeper understanding of the way the users of a specific platform engage with their branded content that goes beyond the registration of likes and comments.
Petkova-Brand Engagement Topologies on Instagram-191.pdf
10:45 - 11:45D 2: GOR Best Practice Award 2015 Competition I
Session Chair: Alexandra Wachenfeld, LINK Institut
Room 158 

Climbing down from the conscious tip of the iceberg – Discovering the unconscious mind of crossdigital consumers

Lisa-Charlotte Wolter1, Maik Schönbach1, Sonja Knab2

1Hamburg Media School, Germany; 2TOMORROW FOCUS Media GmbH, Germany

Relevance & Research Question: In a fragmented media environment with diversified consumer preferences and missing cross platform insights, advertisers are challenged to plan effective communication strategies. Steele et al. (2013) could show in their study that unconscious processes influence multi-platform effectiveness. Additionally knowing that “consumers are no longer considered as completely rational, because emotions, unconscious and automatic processes, play a central role in generating behavior” (Bechara and Damasio 2005; Camerer et al. 2005), the present study aimed at creating a methodology, which helps explaining implicit effects of crossdigital communication. To generate additional insights, the cooperative project of academic researchers (Think Tank Media Management) and a digital content marketing company (TOMORROW FOCUS Media) examines:

(1) How is advertising effectiveness changing if consumers have contact to several digital devices?

(2) Which combinations of digital devices lead to a higher activation and emotional engagement of the consumer?

(3) Is there a mere-exposure effect for crossdigital advertising contact?

Methods & Data: The authors gathered data across four stages with three advertisement contacts (online, mobile, tablet) of an artificial brand using neuroscientifical approaches (eye tracking, EEG, an implicit association test (IAT)). Combined with traditional online survey and consumer choice simulation a more complete picture of crossdigital media effects could be gathered.

Results: The eye tracking technology generated data about the visual perception of the participants, e.g. fixation duration. By simultaneously using mobile EEG technology the authors collected data about the cognitive state of the participants. The following IAT added information about cognitive changes in brand preferences after campaign exposure. An online questionnaire including well-established brand and advertising effectiveness variables supplemented the results. Basically, it could be pointed out how crossdigital media combinations are engaging compared to mono channel communication. However, mono platform communication seems to trigger consumers emotionally. It discovers important facts about the assumption of mere-exposure effects in advertising displayed on multiple media touchpoints.

Added Value: This unique combination of explicit and implicit methods creates a more complete picture of how consumers behave in a crossdigital environment. Since people are often unwilling or unable to report what they feel, the cooperative study is designed to overcome the gaps between self-report and reality. The research project delivered implicit data, which helps marketers to better understand the impact of crossdigital combinations on the effectiveness of communication strategies.

Wolter-Climbing down from the conscious tip of the iceberg – Discovering the unconscious mind of crossdigital .zip

Creating New Grounds for Insight Generation in the Healthcare Market

Juliane Berek1, Claudio Hasler2

1EARSandEYES GmbH, Germany; 2MCM Klosterfrau Vertriebsgesellschaft mbH

Relevance & Research Question:

Thorough qualitative insight generation is the key success driver in the product innovation circle. The healthcare market (OTC) is particularly challenging in that respect, being a market with very complex needs and diverse stakeholders in a highly regulated environment. The task was to provide meaningful insights for the development of a new OTC product range for Klosterfrau.

Methods & Data:

We used latest qualitative mobile and online research tools and trend research in a unique combination.

In the first step, the app MindJournal and an online discussion board were applied. The app makes use of mobile specific advantages for qualitative research. It was used like a diary and documentation tool, ensuring ethnographic exploration of the target group. The online discussion board brought together respondents who had previously engaged in the app to identify specific product indications as well as barriers and limitations.

In the second step, trend research in global innovative cities gathered information on advanced products that overcome previously identified barriers for product development.

Finally, the proprietary TrendMap that figures mid and long term sociocultural trends was used to locate and condense results from the previous steps.


App and discussion board explored the target group’s motivations, attitudes and needs in terms of product relevant behavior. Furthermore, promising ranges of indication including product benefits as well as specific expectations and limitations of a potential product range were identified.

Trend reporting delivered rich data in terms of market understanding and product / range specific solutions. The material is used as stimulus material in the idea and concept generation phase.

In order to determine which of the gathered consumer and product insights match human needs in the long term and are thus most promising for product development, they were correlated to socio-cultural trends deployed in the TrendMap.

Added Value:

The combination of qualitative mobile and online methods allows raising very precise data resulting in meaningful and relevant insights. Trend research enables to evaluate chances of market success for product ideas and innovation platforms very early and delivers “eye openers” that are fundamental for the creative process.
Berek-Creating New Grounds for Insight Generation in the Healthcare Market-226.pptx

Exploring the Web of Sustainability – Understanding Consumer and Stakeholder Groups

Yannick Rieder1, Manuela Wille2, Elke Euler-Horn2

1Linkfluence Germany GmbH, Germany; 2Eckes-Granini Group GmbH

Relevance & Research Question:

Eckes-Granini is a global leader in the juice sector. Since sustainability has become a major topic for both consumers and companies, Eckes-Granini wanted to explore this subject from a pure consumer perspective, which is not biased by social desirability, and therefore decided to conduct a web research project. To adjust CSR activities and the communication strategy or for future product developments, it is crucial for them to understand, how consumers in their major markets Germany, Spain, Finland and France perceive sustainability and which parts of their everyday life are affected by it.

Methods & Data:

This study was a pure digital research project. We build a data set of relevant blogs, forums and websites for each country. This sustainability actors were gathered through a self-programmed crawler and then qualified technically as well as manually based on their connection to sustainability topics, their hyperlink structure and their relevance inside the web. After that we were able to cluster them into different sustainability tribes which can be seen as digital target groups. This sample was the basis for netnographic and qualitative content analysis.


We found out that in all four markets sustainability is a highly discussed topic, but with country specific differences. The selection of discussed topic by consumers ranges from pragmatic waste reduction to creative DIY fashion made from recyclable materials. This variety shows that sustainability has a lot of faces and is not limited to food from organic cultivation. In the context of food especially for Spaniards and Finns regionality drives the buying decision even more than organic labels. This is mainly reasoned by supporting the local economy in contrast to an industrialized production, which suffers from a lack of trust. In this study we gathered lots of consumer insights and inspirations on e.g. trends in functional (“super”)food.

Added Value:

This is the first exhaustive study on consumer’s conversations on sustainability in the web in four European countries. With this approach, we were able to answer Eckes-Granini’s research questions by identifying and exploring the “ecosphere” of sustainability without being limited by priorly defined hypothesis or biased responses. The study also validates the potential of netnography, which allows the identification of consumer insights and future trends.
Rieder-Exploring the Web of Sustainability – Understanding Consumer and Stakeholder Groups-227.pdf
11:45 - 12:00Coffee Break
12:00 - 13:00A 3: Mobile Web Surveys
Session Chair: Bella Struminskaya, GESIS - Leibniz Institute for the Social Sciences
Room 248 

Exploring Why Mobile Web Surveys Take Longer

Mick P. Couper1,2, Gregg Peterson1

1University of Michigan, United States of America; 2University of Maryland, United States of America

Relevance and Research Question:

Surveys completed on mobile Web devices (smartphones) have been found to take longer than surveys completed on a PC. Several explanations have been offered for this finding: 1) slower connection speeds over cellular or Wi-Fi networks, 2) the difficulty of reading questions and selecting responses on a small device, and 3) the increased mobility of mobile web users, who have more distractions while answering Web surveys. Our analyses attempt to disentangle the sources of these differences.

Methods and Data:

We use data from three iterations of a campus Web survey, administered to samples of students, faculty, and staff in 2012, 2013, and 2014. Of the 4,725 respondents who started the 2012 student survey, 10.8% used a smartphone. We have both server- and client-level times for all items (over 360,000 item-level observations) that permit us to disentangle between-page (transmission) times from within-page (response) times. We have detailed information on the types of questions asked (e.g., grid questions versus single-item questions) and client-side paradata on other behaviors (such as scrolling or backing up). Using these data, we plan to build multi-level models (items nested within respondents) to explore correlates of response time by device.


Our initial analyses on the 2012 student survey found that significantly more mobile than PC users broke off (28% versus 13%) and significantly more of those who completed the survey did so in multiple sessions (28% versus 10%). Mobile respondents also took significantly longer to complete the survey (median of 839 seconds) than PC respondents (median of 951 seconds). The majority of the difference is due to within-page times.

Added Value:

Understanding why differences in response times between PC and mobile users occur and attempting to minimize them is an important step in reducing the higher nonresponse rates and breakoff rates observed for respondents completing Web surveys on small mobile devices.

Couper-Exploring Why Mobile Web Surveys Take Longer-113.pdf

Device choice in web surveys: The effect of differential incentives

Aigul Mavletova1, Mick P. Couper2

1NRU Higher School of Economics; 2Institute for Social Research, University of Michigan

Relevance & Research Question: Previous studies have not found efficient ways either of discouraging or encouraging respondents to use smartphones to complete web surveys. We suggest that conditional differential incentives (different incentives depending on the device the respondent uses to complete the web survey) can increase the overall participation rates and the proportion of respondents who use a particular device to complete the survey.

Methods & Data: We conducted an experiment using a volunteer online access panel in Russia. In total, 2,086 completed interviews were collected with the average participation rate of 38%. Smartphone owners who are regular mobile Internet users were invited to complete the web survey. We varied the invitation mode (SMS vs. e-mail) and encouragement to use a particular device for completing the survey:

1. Typical incentives (50 roubles), no encouragement for device

2. Typical incentives (50 roubles), encouragement to use a mobile phone

3. Typical incentives (50 roubles), encouragement to use a PC

4. Higher incentives for mobile phone, typical for PC: 75 roubles vs. 50 roubles

5. Higher incentives for mobile phone, typical for PC: 100 roubles vs. 50 roubles

6. Higher incentives for PC, typical for mobile phone: 75 roubles vs. 50 roubles

Results: The participation rate in the control condition was 35.3%. SMS increased the proportion of mobile web respondents, while e-mail increased the proportion of PC web respondents. As expected, differential incentives increased the overall participation rates (by 8-10 percentage points) if higher incentives were offered for completing the survey via mobile phone. Contrary to expectations, offering higher incentives to PC web respondents did not produce higher participation rates compared to the control condition. Offering higher incentives to mobile web respondents also had an effect on sample composition. Significantly higher participation rates were found among females and those with higher education.

Added Value: Conditional differential incentives can increase the participation rates in web surveys.

Mavletova-Device choice in web surveys-123.pdf

Responsive Questionnaire Design for Higher Data Quality in Mobile Surveys

Frederik Funke, Carmen Borger

LINK Institut, Germany

Relevance & Research Question: In 2014, about 22% of German Internet users accessed the Internet with mobile devices on a daily basis (see van Eimeren & Frees, 2014). However, small screens on smart phones challenge questionnaire designers. Mainly grid questions on mobile devices seem to increase response burden. If all items are to be presented on a single screen, small font sizes and small response options are required. Or, if not all response options are visible on load, additional scrolling is needed which might bias ratings (see Couper, Tourangeau, & Conrad, 2004). To determine the impact of questionnaire customization on data quality, we tested a responsive design that adapts the questionnaire depending on the device used for participation.

Methods & Data: In a split ballot design N = 3929 respondents were recruited in five online panels. They were either directed to a classic design questionnaire (i.e., the same layout for all devices, not optimized for touch screens, and grids on a single page) or to a responsive design questionnaire (i.e., layout depending on available screen size, optimized for touch screens, and grid questions split into single questions). Two grids and two quality check items (e.g., “click the second response option“) were included. Analyses are based on N = 428 (10.9%) respondents who used a smart phone to access the survey.

Results: We found less straight lining in the responsive design (8.6% vs. 16.8% in the classic design), which is an indicator of a beneficial effect on data quality. Completion with the responsive design took some longer (M = 18.9 vs. 17.0 minutes). This, however, seems to be due to deeper question processing as items for quality checks were more frequently answered correctly (90% vs. 78%). Furthermore, respondents gave better evaluations of the survey. No considerable differences were observed, though, regarding item nonresponse, the number of characters in open questions, and the distribution of ratings (e.g., means and top-2 boxes).

Added Value: The results provide further evidence that the questionnaire design should preferably adapt to the type of device respondents use to participate. A responsive design meets mobile participants’ needs and expectations and helps improve data quality.
12:00 - 13:00B 3: Online Participation
Session Chair: Carol Scovotti, University of Wisconsin-Whitewater
Room 147 

Skill Divide in Online Participation: The Case of Wikipedia

Aaron Shaw, Eszter Hargittai

Northwestern University, United States of America

Relevance & Research Question: Millions of people have contributed to one of the most popular Web sites online, Wikipedia, yet research has shown that participation is much more likely by certain types of people such as men. Are there other characteristics that make Wikipedia editing more likely? To sustain contribution and achieve an encyclopedic resource that represents diverse voices and thereby diverse topics in depth, it is important that Wikipedia draw on different people’s contributions.

Methods & Data: We collected survey data of close to 900 adults of all ages in the United States to explore what factors explain likelihood of having edited Wikipedia. Our survey does not sample on Wikipedia editors thereby sidestepping one of the limitations of most existing work on Wikipedia contributions. Rather, our data set includes information about both contributors and non-contributors.

Results: Analyzing data about 898 adults, we find that men, younger people, and those with higher income are more likely to have edited Wikipedia. Unique to our study is the inclusion of several measures of self-efficacy and skill, factors that prior literature about other online engagement has found to be relevant to explaining participation. Perhaps not surprisingly, we find that confidence in editing Wikipedia relates to having done so. Additionally, having received help with editing Wikipedia is a significant explanatory factor. We also find that knowing more about how to edit Wikipedia is not related to editing experiences when we control for other factors. However, interestingly, more general Internet skills do matter. When looking at both demographic and Internet experience variables together in one logistic regression model, we no longer find gender differences in editing. That is, when we control for confidence in editing, having received support with Wikipedia editing, Wikipedia-specific skills, and general Internet skills, there is no difference in women’s versus men’s likelihood of contributing to Wikipedia content. Age, however, remains significant with older people less likely to edit Wikipedia.

Added Value: The study contributes to understanding the role of Internet skills in differentiated Wikipedia editing. Findings suggest that with support, more people may become editors in the future.

Shaw-Skill Divide in Online Participation-190.pptx

Open Education and Online Participation: Exchanging Practices inside a Research Project

Francisco Freitas

Centro de Estudos Sociais, Portugal

In this presentation, it is intended to offer an overview of the research practices established in ALICE Project, an Advanced Grant funded by the European Research Council currently hosted at the Centre for Social Studies of the University of Coimbra. This 5 year project on the field of social sciences and humanities encompasses a team of 22 researchers distributed by 4 main research domains. There has been an effort in order to produce not only traditional academic outputs, but also a myriad of non-academic deliverables, including multimedia files, art exhibitions and performances. For all, there is a strong commitment with open dissemination over the internet, social networks included.

Relevance & Research Question:

How can the internet and particularly how social media can foster and influence dissemination?

Methods & Data:

The provided answers will be grounded on data coming from Google Analytics, Facebook, Twitter, or MailChimp. The data illustrates the dissemination results across the project timeline, tracing the evolution in terms of the online participation and achievements. In addition explanation of the processes and routines will be highlighted.


With the data, we are able to obtain results in different directions namely in terms of tracing target groups, in identifying the sociodemographic characteristics of the audience, in pinpointing their geographic distribution, in evaluating the participation.

Added Value:

The value of this work is twofolded. On the one hand, it is possible to demonstrate how internet is populated by tech savvy tools that foster academic dissemination at little or no cost. In addition and considering this age of information overload, there are reasons to consider that the effort to communicate over the internet is becoming crucial to conquer the necessary audience. Secondly, the project data is generated in many different sources and data collections. In many respects, the produced information and outputs testify how the offline and the online worlds are interconnected, not establishing a clear divide between the two branches.
12:00 - 13:00C 3: Business Analytics with Social Media II
Session Chair: Christian Bosau, Rheinische Fachhochschule Köln (RFH)
Room 149 

A Longitudinal Perspective on the Social Media Usage by Retailers

Rianne van Lambalgen1, Jesse Weltevreden2

1Amsterdam University of Applied Sciences, The Netherlands; 2Amsterdam University of Applied Sciences, The Netherlands

Relevance & Research Question (longitudinal, social media use, retailers):

The past few years, many retailers have acknowledged the value of social media for their marketing strategy. However, not every retailer has been successful in maintaining their social media account(s), resulting in a decrease in the number of followers over the years. So far, little is known about the development of commercial social media strategies over time. This research focuses on the comparison between the social media usage of retailers in 2011 and that of the same companies in 2014 for various social media sites. Both the adoption of social media is investigated, as well as the factors that influence retailers’ success on social media.

Methods & Data (content analysis, web scraping, large samples, longitudinal analysis)

In 2011 an extensive research has been executed among more than 5,000 retailers in the Netherlands concerning their social media usage and their success on social media sites in terms of number of followers. In 2014, the same sample was used to analyze the changes in social media adoption between 2011 and 2014. It was investigated to what extent changes in adoption and success over the years are influenced by firm characteristics, online experience and online popularity. Data collection has been done from June 2014 to October 2014 by using Web scraping software as well as manual content analysis. Descriptive analyses, linear and logistic regression analyses were applied.


Results show that although social media adoption of retailers has increased, not all companies are as successful in 2014 as compared to 2011 (in 5% of the cases the number of followers declined). More specific, large and international companies have been significantly more successful over time as compared to national and small companies. In addition, influences of type of industry and online characteristics differ per social media site.

Added Value

This research provides insight into the development of commercial social media use over time, which so far received little attention in the literature. Furthermore, it gives retailers knowledge on how similar companies have been doing with regard to the use of social media.

van Lambalgen-A Longitudinal Perspective on the Social Media Usage-163.pdf

“Can Internet Searches Forecast Tourism Inflows?”

Concha Artola1, Pablo de Pedraza2, Fernando Pinto3

1Central Bank of Spain, Spain; 2University of Amsterdam, The Netherlands; 3University of Salamanca, Spain

Relevance and research question.-To improve the forecast of tourism inflows into Spain by use of Google -indices on internet searches measuring the relative popularity of keywords associated with travelling to Spain.

Methods and data.- Two models are estimated for each of the three countries with the largest tourist flows into Spain (Germany, United Kingdom and France): A conventional model, the best ARIMA model estimated by TRAMO (model 0) and a model augmented with the Google-index relating to searches made from each country (model 1). The overall performance of both models is compared.

Results.- The improvement in forecasting provided by the short-term models that include the G-indicator is quite substantial up to 2012, reducing out of sample mean square errors by 42%, although their performance worsens in the following years. Deeper study and conceptualization of sources of error in Google trends and data quality is necessary.

Added Value.- The paper illustrates that while this new tool can be a powerful instrument for policy makers as a valuable and timely complement for traditional statistics, further research and better access to data is needed to better understand how internet consumers’ search activities translate (or not) into actual economic outcomes.

Artola-“Can Internet Searches Forecast Tourism Inflows”-144.pptx

Phubbing because of FoMO? – “Fear of Missing Out” as a predictor for problematic mobile phone use – when being alone and in company

Christian Bosau, Max Kühn

Rheinische Fachhochschule Köln (RFH), Germany

Relevance/Research-Question: In recent month the new phenomenon “phubbing” (snubbing someone in favour of your mobile phone) emerged (Lobe, 2014). Additionally, it is argued that the use of modern internet services (e.g. social networks) can be seen as addicted behavior (Masur, 2013; Bosau, Aelker & Amaadachou, 2014a & 2014b). “Phubbing” could therefore be described as a consequence and manifestation of addicted habbits, i.e. addicted people can’t stop to use their mobile phones even when they are in company of friends. To explain this behavior the concept of FoMO (Fear of Missing Out) could be used since it was already successfully applied in similar research contexts (Przybylski, Murayama, DeHaan & Gladwell, 2013; Bosau et al., 2014b).

Methods/Data: The online-study (N=101) analyzed the influence of FoMO (measured by Przybylski et al., 2013) on the compulsive checking of mobile phones, operationalized by three different scales: a) general problematic mobile phone use (Güzeller & Cosguner, 2012), b) habitual checking tendencies (Bayer & Campbell, 2012), c) frequency of checking (Collins, 2013). To analyze differences of mobile phone use while being alone vs. being in company respondents had to answer scales b) and c) twice to disclose their behavior in the two different settings. Several stepwise regression analyses using bootstrapping were calculated to analyze the explaining power of FoMO.

Results: The results show that FoMO in general is a very strong predictor: people with more FoMO show more problematic mobile phone use. However, in social situations the problematic use is generally lower and the predictive power of FoMO is slightly smaller, i.e. being in company serves as a moderator and buffers to some extend the influence of FoMO on the checking behavior; although, FoMO is still highly correlated to all three behavioural indices. Interestingly, the influence of FoMO differs significantly between males and females and is commonly stronger for males.

Added Value: The study is a first step in explaining why people can’t stop to use their mobile phones, even when they are in social situations where they meet other people in person. It further highlights the explanatory power of the fairly new concept FoMO.
Bosau-Phubbing because of FoMO – “Fear of Missing Out” as a predictor-131.pdf
12:00 - 13:00D 3: GOR Best Practice Award 2015 Competition II
Session Chair: Oliver Tabino, linkfluence Germany & Q Agentur für Forschung
Room 158 

Native Ads - Ad Content in Context

Kerstin Hammen1, Michael Wörmann2, Eva Herzog1

1Yahoo! Germany, Germany; 2Facit Digital GmbH, Germany

Relevance & Research Question:

Digital display advertising is going through a seismic change. Reaching consumers with relevant advertising is becoming increasingly difficult. Nick Cohen, Managing Partner at MediaCom Beyond Avertising summed it up well: "The old models of display advertising are not as effective, for some jobs, as they used to be. We need to look at new opportunities to connect with people."

Those new opportunities are increasingly being built around two of the most disruptive elements in advertising today – content marketing and native advertising – yet the level of insight as to how consumers are responding to them is relatively low.

For this reason Yahoo! and Facit Digital examined the way native ads work with respect to visibility, their ability to convey content, user preferences and the impact on the perception of brands.

Methods & Data:

We conducted 40 individual interviews in which we also collected eye-tracking data (desktop and mobile) and validated the results by surveying 4,014 users, 14-49 years old - half of them on their mobile devices.

Participants were randomly assigned to one of 34 experimental conditions. We distinguished between the ad environment, mobile/desktop, type of advertising, brand and communication objective.

- High visibility with native ads: 93% of those surveyed on their desktop PC and 85% of those surveyed on their mobile device visually engaged with the native ad
- On a desktop native ad's specific content was recalled 13% more often than display advertising content. With mobile advertising, users likewise remember native ad's specific content 19% more often than they remember display advertising. Relevant content helped increase the recall performance both on desktop and on mobile devices.
- Users know that they will land on the advertiser‘s page after a click (>70%); The general rule is: The landing page must contain the content that the ad has promised
- Content driven approaches lead overall to the impression of a closer connection to customers – even if the content provided or the topic do not relate to the user‘s own interests.

Added Value:

The complexity of the experimental setting used for this study gave us the chance to deliver insights on a wide variety of aspects around native ads. It shows not only the strengths, but also the potential pitfalls - from an advertiser as well as from an user perspective. Therefore, this study can both help optimize campaigns and make the digital experience of users more meaningful.

Hammen-Native Ads - Ad Content in Context-229.pptx

Screentypes - representative typologies of smartphone- and tablet-users

Anja Manouchehri1, Wolfgang Dittrich2

1Axel Springer Media Impact GmbH & Co. KG, Germany; 2d.core GmbH, Germany

Relevance & Research Question:

“Screen-Types” describe in two typologies, representative for Germany, different types of users of smartphones and tablet computers. Both typologies provide comprehensive information about differences of users of digital mobile devices concerning demographic characteristics, interests, attitudes, motives, preferences, purchase behavior and media-usage. These typologies provide a deeper understanding of these fast growing target groups.

Methods & Data:

The typologies were calculated by cluster analysis. Active variable were items concerning usage of the mobile devices. We reduced the lists of about 40 items via factor analysis to 12 factors. The cluster analysis was calculated with the data of the market media study “best for planning 2013 III”, which is based on more than 30.000 interviews face to face, ADM sampling system. The reach of online media is partly technically measured.

The sample population are users of smartphones and tablets in Germany 14 years and older.


The result is two typologies that make clear, that users of mobile devices differ a lot as far as demographic and psychographic characteristics are concerned. For example one type is the “digital newspaper- and magazine-reader” who uses his smartphone intensively to read print media. He is male, wealthy, status conscious and he likes branded articles. Another type is the “mobile communicator”. He is very young and he belongs to a postmodern, hedonistic milieu. He permanently uses social networks, he posts and tweets a lot. It is important for him to experience a lot in his life.

Added Value:

The Screen-Types are the first representative typologies that provide information about users of mobile devices on this particular high level of detail and based on such ab big database and sample size.

The typologies were integrated in the actual market media study "best for planning 2014 I". So up to know all characteristics are variable combinable for individual analysis in "best for planning".

Manouchehri-Screentypes - representative typologies of smartphone- and tablet-users-224.pptx

Using Online Real-Time-Measurement for a Deeper Understanding of TV-Spot Impact

Angelina Brandenburg1, Michael Lenz2

1YouGov Deutschland AG, Germany; 2InnoGames GmbH, Germany

Relevance & Research Question: InnoGames, one of the biggest developers of online games worldwide, wanted to establish a reliable test design for TV-spots within the early stages of development in order to maximize their general appeal and success. The intention was to highlight the key strengths and weaknesses of a spot in a very detailed manner that allows concrete conclusions about the potential for optimization. In the presented case, two different animated storyboards of a TV commercial for the game “Grepolis” were analyzed.

Methods & Data: We conducted a quantitative panel based online-survey. The special focus of the interview was a real-time-measurement using mouse movements to identify the involvement scene by scene. This was combined with a classical ex-post advertising effect test. In one of the spots the focus was its powerful protagonist Zeus, in the other spot it was a female character reminding of peaceful times in the past.

Results: The results of the ex-post ratings showed hardly any differences between the spots: Both performed well on category fit, insight and credibility but weaker on activation. Rather surprising in this context was that when asked to rank both spots in direct comparison, the Zeus-spot was the clear winner. Real-time-measurement helps explaining this remarkable finding. In terms of involvement the Zeus-spot excels: starting at the very first second this spot pulls the viewer in and keeps the involvement at a constantly high level throughout the whole spot, especially during the scenes that focus on the character Zeus. The second spot reaches a high involvement level as well, however, only after one third of the spot is seen and at a less constant level. Here, only the part of remembering old times performs extraordinarily well.

Added Value: The real-time-response-measurement was able to deliver very detailed and highly informative differences between the two spots that would not have been revealed by using an ex-post survey method only. The study delivers a deep understanding of the impact and specific strengths and weaknesses of the spots and by this allows deriving concrete and targeted recommendations for action.
Brandenburg-Using Online Real-Time-Measurement for a Deeper Understanding-225.pdf
13:00 - 14:00E: Präsentation der Auftaktstudie zur Tour der Marktforschung 2015
Session Chair: Raimund Wildner, GfK-Nürnberg e. V.
Session Chair: Daniel Förtsch, Initiative Markt- und Sozialforschung e.V.

Eine Veranstaltung der Initiative Markt- und Sozialforschung e.V.

Room 248 
13:00 - 14:00Lunch
14:00 - 15:30C 4: Demos: Showcasing new technical developments
Session Chair: Alexander Wenz, University of Essex
Room 158 

nodeGame: a tool for real-time on-line behavioral experiments

Stefano Balietti

ETH Zurich, Switzerland

Relevance & Research Question: The last fifty years experimental research in economics, sociology, and social psychology has been conducted at university laboratories, with restricted subject pools, and limited numbers of participants per experimental session. This research has led to important scientific progress. However, the rapid development of Internet-related technologies makes it possible to overcome some important limitations of traditional experiments. This abstract introduces a methodological contribution, a software named nodeGame (

Methods & Data: NodeGame is a free, open-source library developed at ETH Zürich, which allows to design and implement behavioral experiments directly in the browser window. It is now a 3-year old project, and it is still under active development. NodeGame is a programming library written HTML 5 JavaScript, with support for Web-Sockets. It allows to handle thousands of simultaneous participants connected to the same or different experiments. NodeGame is also cross-browser, and cross-device, meaning that the same experiment can run smoothly on desktop computers, as well as on mobile devices. Moreover, it integrates smoothly with on-line recruitment web sites, such as Amazon Mechanical Turk. NodeGame helps solving the most common problems in running experiment on-line: browser incompatibilities, lagged connections, disconnecting players, creating waiting rooms, etc. Moreover, in nodeGame it is easy to add artificial computer-controlled clients (bots) into running experiments, preventing losing session-data in case of prolonged disconnections, or to test artificial intelligence algorithms playing against humans.

Results: Public goods games with and without assortative matching, ultimatum games, and creativity tasks has been conducted successfully on-line and in the lab. The results of one or more experiments will be presented.

Added Value: Given the lower cost of on-line behavioral research, the added value of software like nodeGame is the possibility to conduct calibration of experimental parameters to an extent that is not possible in the lab, due to budgetary and time constraints. NodeGame introduces a number of methodological improvements for conducting on-line behavioral research. It represents a step in the direction of what some researchers have described as “The Future of Social Experimenting”: rapid hypothesis generation and testing with the integration of computer simulations and social experiments.

Automatic analysis of open questions in market research

Johannes Hercher, Joao Filipe Baigger

Rogator AG, Germany

Relevance & Research Question: Rogator RogTCS is a system for semi-automatic identification, exploration and visualization of topics and sentiment in text collections mainly generated by open questions in surveys.

Methods & Data: The application combines a number of proven computational methods to an innovative, inductive approach. A key role is played by the automatic linguistic processing of the entries, the sentiment analysis, collocation identification, vector models of information retrieval, latent semantic indexing, text clustering methods, keyword analysis, and various visualization techniques. The typical text material processed by RogTCS consists of short textual answers and comments usually derived from open questions.

Results: The system has been designed to use as few language- and domain-specific resources as possible in order to enable the processing of of text in different languages, as well as to avoid the construction of ontologies, sentiment lexica, etc. Instead, it utilizes statistical methods that have been carefully tuned in order to give a faithful representation of the content of the input data, condensing them in topics and sentiments into a manageable amount of graphs, numerical indicators and tables.

Added Value: Rogator RogTCS, is intended for fast and cost-effective analysis of open questions in market research. RogTCS offers market researchers an alternative to the time-consuming and expensive manual coding of answers to open questions, which involves the creation and application of a specialized code plan for each question. RogTCS is designed to produce highly satisfactory results without any manual intervention, so that a rough analysis of a text collection can be generated within a few seconds. If necessary, the initial analysis can then be refined by users in an interactive, semi-automatic procedure. In theory, the application range can be expanded to the analysis of nearly any kind of electronic text collection that contains enough material for the statistical analysis. Possible applications, among others, include the analysis of Twitter tweets, online discussion boards, product reviews, or any other material obtained from diverse social media platforms.

A Near Real-Time Application for Twitter Data Analysis

Zinayida Petrushyna1, Alexandra Chueva2, Ralf Klamma3, Joachim Lanfermann4

1RWTH Aachen University, Germany; 2RWTH Aachen University, Germany, e-dynamics GmbH, Aachen, Germany; 3RWTH Aachen University, Germany; 4e-dynamics GmbH

Relevance and Research Question: Twitter is a source of huge amount of interesting and relevant data about nearly any topic in the world. Though democracy of publishing enables advertisers, spammers and chatters to soil information in Twitter. However, data analysis required for retrieving useful information from tweets is time-consuming. Therefore we have designed a software architecture that is able to provide near real-time data analysis solution.

Methods and Data: We base our architecture on a distributed real-time computational framework Apache Storm. The data are stored in the NoSQL document-oriented database Mongo DB that receives tweets from the Twitter Search API and the Twitter Streaming API. Tweets are analyzed according to given metrics in a near real-time manner. We visualize results applying d3.js JavaScript library.

Results: We design and develop a near real-time application for Twitter data analysis. After that we test our application on number of conferences and get promising results. For example, 500 tweets are analyzed in 2 minutes. Under analysis we understand text mining of a tweet content, ranking of a tweet according to its impact on the audience and its storage to MongoDB.

Added Value: The proposed software architecture can be used for data analysis of various Social Media data. It is especially useful for companies to build cloud-based solutions for monitoring and analyzing a big amount of Social Media data related to companies at a near real-time.

Petrushyna-A Near Real-Time Application for Twitter Data Analysis-160.pdf

The design for an advanced, multi-country, web-based survey and coding tool for occupations in ISCO-08

Kea Tijdens

University of Amsterdam, Netherlands, The

Relevance & Research Question: This paper discusses the design for an advanced, multi-country, web-based occupational survey and coding tool for ISCO-08. This includes the coding of open text questions as well as the use of closed format questions using look-up tables with text string matching or search trees.

Methods & Data: For an auto-coder applying machine learning algorithms a large volume, multilingual training set is required. This database falls apart into two components, one with individual level data, consisting of merged and harmonized survey data from as many surveys as available, including the raw text data. The second is an occupation level database and consists of coding indexes, occupation databases of, job titles and job descriptions.

Results: The explorative review of the possibilities to build such a survey and coding tool is positive. Autocoders have been applied in Germany and the USA (see Bethmann et al 2014; Cheeseman 2014). Web surveys have been using look-up databases for self-identification of occupations (Tijdens 2014). Some survey holders have already expressed their willingness to put their raw text plus codes at the disposal of the project. Multilingual occupational databases are available, and job titles and job descriptions will be collected from job boards. The project aims to cover 35 languages in 99 countries.

Added Value: A web-based survey and coding tool for so many countries and languages is currently not available. In addition the tool will be used for coding of industries in NACE2.0, educationand fields of education in ISCED_2011, and aims to apply social stratification indicators. This paper builds on work done on behalf of a proposal for such a tool, submitted for funding to the European Union’s Horizon2020 program. If funded, this tool will meet the demand of survey holders for cross-country harmonized, fast, high-quality and cost-effective surveying and coding of occupations.
Tijdens-The design for an advanced, multi-country, web-based survey and coding tool for occupations-184.pdf

Getting closer to the moment of truth with qualitative online and mobile diaries

Zacharias de Groote

Liveloop GmbH, Germany

Relevance & Research Question: Systematic diary approaches enable researchers to gain a better understanding of the lifeworld of research participants. Today, new methodological and technical developments promise to provide new opportunities to get closer to the moment of truth in online qualitative research projects.

Methods and Data: We will showcase a working demo of a qualitative online-mobile diary with a combination of one-time and recurring tasks, enhanced with picture uploads, user generated video recordings, creative exercises and individual webcam interviews.

Results: The practical example of a combined online and mobile diary provided will demonstrate how a research diary benefits from the flexibility and potential of most recent digital qualitative methods and technological advancements.

Added Value: The newly developed software showcased provides researchers with new opportunities to engage their target audiences in ethnographical, customer experience and usability projects – enabling them to dig deeper into the who, where, what and when with online and mobile diary approaches.
14:00 - 15:30D 4: Panel Discussion: Behavioural Economics: A new idea of man – a need for new methods?
Session Chair: Horst Müller-Peters,

Panelists include:

Dr. Florian Bauer (Member of the Board, Vocatus AG, Germany)

Prof. Dirk Frank (Managing Director, ISM Global Dynamics, Germany)

Jon Puleston (VP of  Innovation, Lightspeed GMI, United Kingdom)

Orlando Wood (Managing Director, BrainJuicer Labs, United Kingdom)

Room 248 

Behavioural Economics: A new idea of man – a need for new methods?

Horst Müller-Peters, Germany

New research in the fields of behavioural sciences and neurosciences have changed the view of human action and thinking. The works of Kahneman, Tversky and many other researchers have shown that our decisions are to a much lesser degree made consciously than we thought and are very much shaped by automatisms and situational influences.

Particularly our self-perception is not realistic: we are not really in a position to predict our behaviour. And memories of past behaviour and its sources are systematically distorted.

In economics this research has led to the establishment of a new discipline: behavioural economics. Also many market and social researchers are demanding new thinking and methods in their discipline. The key question is in how far verbal surveys can at all provide valid data on decision-making processes if respondents are not (or at least not completely) conscious of it. Do we need to reinterpret methods of empirical social sciences or even have to come up with completely new methods? Will new research designs, the use of gamification or implicit measurement be able to function as a bridge between human thinking and action or will we need to employ methods of neurosciences? Does this shift in methodology signify the real revolution in market research instead of questioning methods of data collection and de-anonymisation? Or is it a discussion of old issues which qualitative and psychological market researchers have addressed and solved already a long time ago?

These issues will be discussed by a group of experts in the field

• Orlando Wood, Managing Director, Brainjuicer Labs, United Kingdom

• Dr. Florian Bauer, CEO Vocatus AG, Germany, and author of the book „Der unvernünftige Kunde: Mit Behavioural Economics irrationale Entscheidungen verstehen und beeinflussen“

• Jon Puleston, VP Innovation, Lightspeed GMI, United Kingdom

• Professor Dirk Frank, Managing Director ISM Global Dynamics GmbH and Professor at Hochschule Pforzheim, Germany

• Moderation: Professor Horst Müller-Peters, Editor of

14:00 - 15:30Poster Session I: Online Methodology and Applied Online Research

Virtual Teams at Work: Do Attractive Interfaces Improve Performance?

Maria Douneva1, Russell Haines2, Meinald T. Thielsch1

1University of Münster, Germany; 2Old Dominion University, USA

Relevance and Research Question:

Online research has acknowledged the importance of website and interface aesthetics by examining its strong influence on users’ attitudes and reactions. However, evidence for effects on task performance is limited and mixed. By manipulating chat background colour in a within-subjects design, this study aims to investigate the effects of an unattractive green vs. attractive blue chat interface on task performance in a web-based collaborative setting.

Methods and Data:

During the study, participants communicated via chat and email using a custom-made browser application. Each of them performed the simulated role of a nurse, a doctor, laboratory technician, or specialist as a member of an emergency response team that had to diagnose patients within a given time. The data of 184 participants (53.3% male, mean age = 21 years, SD = 4.8) during three rounds of the same task were analysed for effects of the colour manipulation on mood, affect, and team performance (measured as number of diagnosed patients after each round). Chat background colour was randomly varied in round 2 and 3.


Although a t-test shows that participants clearly preferred the attractive over the unattractive version (p < .001, d = 0.5), analyses of variance reveal that neither their attitudes (mood or affect) nor the performance on the group level are significantly influenced by the colour manipulation. Still, participants in the unattractive colour condition have the lowest mood and affect scores. Likewise, there is a tendency for better performance after seeing the attractive colour in round 2 (d = .08) and in round 3 (d = .12).

Added Value:

While one can find several effects of aesthetics on attitudes and online user reaction measurements in the literature, none have directly examined the effects of aesthetics on performance, especially in a virtual team setting. The results here challenge the assumption “attractive things work better”, which also proposes mood to mediate between aesthetics and performance. Additionally, our study offers future directions for research on the link between web aesthetics and performance, namely varying the performance criterion, the nature of the task, and the aesthetics manipulation used.
Douneva-Virtual Teams at Work-236.pdf

Connecting Offline and Online Surveys: Reconsidering Respondent Determinants in Attribute Bias

Sae Okura1, Yohei Kobashi1, Leslie Tkach-Kawasaki1, Manuela Hartwig2, Yutaka Tsujinaka1

1University of Tsukuba, Japan; 2Free University of Berlin, Germany

Relevance & Research Question

While internet-based surveys are considered methodologically biased (Rasmussen 2008), previous studies suggest that bias involving political attitudes can be corrected partially through weighting based on respondent attributes (Taniguchi & Taniguchi 2008). However, the relationship between bias deriving from online surveys and attribute bias is unclear, giving rise to the question: Does mitigation accidentally arise due to respondent determinants in attribute bias?

To investigate further, in considering the possible universality of bias produced by online surveys, we demonstrate that bias could originate in the fact that respondents use online surveys (as Internet users), and that reassessment of attributes, including educational level, reveal the possibility of mitigation.

Method & data

Our data is based on (1) a 2014 online survey administered to respondents in Germany, Japan, the U.S., and Korea; and (2) a 2013 door-to-door survey undertaken in Japan, wherein most survey items were identical to the 2014 online format. We first compared the results of the survey with census data from each country to investigate the universality of the resulting bias. By correcting for inherent bias among Internet users drawing on each country’s current situation, we then tested the validity of bias correction in terms of political party support. Focusing on the combined results for the two Japanese surveys, we obtained our results by compensating for the propensity score.


According to our results, there are intrinsically common gaps among online respondents and the general public in terms of education (i.e., in Japan, college graduates at 66.0% versus 37.3%, respectively) and job status in the four countries. Consistent with previous research in Japan, correcting the bias for education is critical, and this tendency is supported. Henceforth, the correlation between education and political attitudes, and possible corrections for other attributes will be statistically investigated [Reference: Table 1].

Added Value

The strengths of our research lies in our approach to detecting universal bias and its contributing mechanism through our analysis of highly comparative cross-national data sets. Our results clearly suggest that the universal bias in online surveys is based on inherent social features, including education, in each advanced nation.
Okura-Connecting Offline and Online Surveys-196.pdf

Higher response rates at the expense of validity? Consequences of the implementation of the ‘forced response' option within online surveys

Jean Philippe Décieux2, Alexandra Mergener1, Kristina Neufang3, Philipp Sischka2

1Federal Institute for Vocational Education and Training (BIBB), Germany; 2University of Luxembourg, Luxembourg; 3University of Trier, Germany

Due to the low cost and the ability to reach thousands of people in a short amount of time, online surveys have become well established as a source of data for research. As a result, many non-professionals gather their data through online questionnaires, which are often of low quality due to having been operationalised poorly (Jacob/Heinz/Décieux 2013; Schnell/Hill/Esser 2011).

A popular example for this is the ‘forced response‘ option, whose impact will be analysed within this research project.

The ‘forced response’ option is commonly described as a possibility to force the respondent to give an answer to each question that is asked. In most of the online survey computer software, it is easily achieved by enabling a checkbox.


There has been a tremendous increase in the use of this option, however, the inquirers are often not aware of the possible consequences. In software manuals, this option is praised as a strategy that significantly reduces item non-response.

In contrast, research studies offer many doubts that counter this strategy (Kaczmirek 2005, Peytchev/Crawford 2005, Dillman/Smyth/Christian 2009, Schnell/Hill/Esser 2011, Jacob/Heinz/Décieux 2013). They are based on the assumption that respondents typically have plausible reasons for not answering a question (such as not understanding the question; absence of an appropriate category; personal reasons e.g. privacy).

Research Question:

Our thesis is that forcing the respondents to select an answer might cause two scenarios:
- Increasing unit non-response (increased dropout rates)
- Decreasing validity of the answers (lying or random answers).

Methods and Data:

To analyse the consequences of the implementation of ‘forced response’ option, we use split ballot field experiments. Our analysis focuses especially on dropout rates and response behaviour. Our first split ballot experiment was carried out in July 2014 (n=1056) and we have planned a second experiment for February 2015, so that we will be able to present our results based on strong data evidence.

First results:

If the respondents are forced to answer each question, they will
- cancel the study earlier and
- choose more often the response category “No” (in terms of sensitive issues).

Décieux-Higher response rates at the expense of validity Consequences of the implementation of the ‘forced re.pdf

A comparison of two eye-tracking supported cognitive pretesting techniques

Cornelia Neuert, Timo Lenzner

Gesis - Leibniz Institute for the Social Sciences, Germany

Relevance & Research Question: In questionnaire pretesting, supplementing cognitive interviewing with eye tracking is a promising new method that provides additional insights into respondents’ cognitive processes while answering survey questions. When incorporating eye tracking into cognitive interviewing, two techniques seem to be particularly useful. In the first technique (retrospective probing), cognitive interviewers first monitor participants’ eye movements and note down any peculiarities in their reading patterns, and then ask targeted probing questions about these peculiarities in a subsequent cognitive interview. In the second technique (gaze video cued retrospective probing), respondents are additionally shown a video of their eye movements during the cognitive interview. This video stimulus is supposed to serve as a visual cue that may better enable respondents to remember their thoughts while answering the questions. We compare gaze video cued retrospective probing with retrospective probing without any cue when it comes to identifying problematic survey questions by addressing the following research questions: 1) Do both techniques differ in terms of numbers of problems identified? 2) Do both techniques differ in types of problems identified? 3) Does using a gaze video stimulate the participants in different ways when commenting on their behavior?

Methods & Data: In a lab experiment, participants' eye movements (N=42) were tracked while they completed six questions of an online questionnaire. Simultaneously, their reading patterns were monitored by an interviewer for evidence of response problems. After completing the online survey, a cognitive interview was conducted. In the retrospective probing condition, probing questions were asked if peculiar reading patterns were observed during the eye-tracking session (e.g., re-readings of specific words). In the other condition, participants were shown a video of their recorded eye movements in addition to receiving probing questions about the questions displayed.

Results: Results show that both techniques did not differ in terms of the total number of problems identified. However, gaze video cued retrospective probing identified less unique problems and less different types of problems than pure retrospective probing.

Added Value: Our experimental study offers first insights in the usefulness of a gaze video cue in conjunction with the method of cognitive interviewing.

Neuert-A comparison of two eye-tracking supported cognitive pretesting techniques-136.pdf

Development and Validation of the "Participatory Market Communication Scale" (PMCS)

Stefan Beckert, Alena Kirchenbauer, Julia Niemann, Alexander Schulze

Universität Hohenheim, Germany

Relevance & Research Question: Recent media developments – especially the social web – paved the way for recipients’ active participation in advertising and market communication. Until now, this phenomenon of participative market communication (PMC) was mainly examined using qualitative approaches. To gain a better understanding of the motives for PMC, we developed a quantitative test instrument, the "Participatory Market Communication Scale" (PMCS). We draw on three dimensions identified by Berthon et al. (2008: 10f): “intrinsic enjoyment”, “self-promotion” and “change perception”. Based on the fact that most PMC campaigns offer a competition to stimulate participation, a fourth dimension “reward” was added.

Methods & Data: To develop the scale, items that represent the four mentioned dimensions were formulated and tested using qualitative and quantitative pretest techniques. The final scale contained 13 Items and was evaluated during a multi-thematic online survey. The sample was quoted by age, gender and formal education. It can be considered representative for the German online population (n = 448).

Results: The items show satisfying values for item difficulty and item variance. However the assumed multi-dimensionality of the construct could not be confirmed. Neither explanatory nor confirmatory factor analysis lead to results that indicate the four supposed dimensions. Especially the three dimensions “intrinsic enjoyment”, “self-promotion” and “change perception” that were derived from the qualitative study of Berthon et al. (2008) do not differentiate. All items that belong to these three dimensions load on the same factor, we called “action motivation” (α = .93). The second factor “reward motivation” (α = .77) consists of the items that represent the extrinsic motivation for participating due to the exposed reward.

Added Value: PMCS is a suitable instrument for advertising research. Due to its two-dimensionality the scale is flexible in its application. Advertising material that contains a reward for PMC-activities can be tested by adding the items for “reward motivation”, so these items can be seen as an add-on module of PMCS.


Berthon, Pierre/Pitt, Leyland/Campbell, Colin (2008). When consumers create the ad. In: California Management Review, 50(4), S. 6-30.

Beckert-Development and Validation of the Participatory Market Communication Scale-238.pdf

webdatanet & webdatametrics

Pablo de Pedraza

Univerity of Amsterdam, The Netherlands

WEBDATANET is a unique multidisciplinary European network bringing together leading web-based data collection experts, (web) survey methodologists, psychologists, sociologists, linguists, media researchers, Internet scientists, economists and public opinion researchers from 31 European Member States plus USA, Brazil and Russia.

By addressing methodological issues of web-based data collection (surveys, experiments, tests, non-reactive data collection, and mobile Internet research) and fostering its scientific usage, WEBDATANET aims to contribute to the theoretical and empirical foundations of web-based data collection, stimulate its integration into the entire research process (i-science), and enhance the integrity and legitimacy of these new forms of data collection.

The Master in Webdatametrics – web based data collection and analysis –offers a front of the line top quality programme in the area of Internet-based data collection methods and Big Data analysis. The programme was created by the Webdatametrics Academic Board elected from members of the COST network Webdatanet. It joins together academics and researchers from the best Universities and departments in Europe.

Who is Your Customer? A Data-Driven Approach to B2B Customer & Competitive Analysis

Carol Scovotti1, Ross Scovotti2

1University of Wisconsin-Whitewater, United States of America; 2NueMedia, LLC, United States of America

Relevance & Research Question: Customer and competitive information needed by firms in the business marketplace (B2B) is not as readily available as consumer data. Niches are smaller and lines of business more specialized. Customer preference data is difficult to collect and few materials, machinery, and product suppliers have extensive marketing analytics expertise. Nonetheless, businesses on the commercial side of the supply chain must understand customer needs and preferences. They must also be aware of alternatives that interest their prospects. Industry information disseminators like trade publishers, web portal managers, and associations can increase their value to members, subscribers, advertisers, etc. by providing customer and competitive intelligence that independently would be difficult to gather. This study examines an innovative use of customer and competitive data collected in a “State of the Industry” report conducted by a US-based site developer and manager of web portals for the commercial woodworking, finishing and countertop industries.

Methods & Data: Over 42,000 practitioners in these industries were emailed an invitation to participate in an online survey in August 2014. Almost 5000 responded (11.9%), providing information about the size, nature and scope of their businesses, and the media they use to gather company, product, and business process information. They also identified the equipment, software, and supplies used to produce their products, including current vendors and competitors they would consider in the future. These data were then grouped by product category to identify buying behaviors.

Results: Results vary by company within a product category. For example, the customer profile of a premium quartz countertop manufacturer is a fabricating specialist with 11-25 employees, $1-5 million in annual revenue and attends trade shows. The profile of a popular laminate maker is a cabinet maker or contractor that also makes countertops with 2-10 employees, less than $1 million in annual revenue and finds product information through search engines and email.

Added Value: These examples demonstrate that with a little planning and the right questions, B2B information disseminators can supply their clientele with enhanced demographics and media usage information so they can identify, communicate with, and ultimately convert viable prospects into customers.
Scovotti-Who is Your Customer A Data-Driven Approach to B2B Customer & Competitive Analysis-173.pptx

Research on Pilot Survey for Mixed Mode Effects: Face to Face Survey and Internet Survey

KyuHo Shim1, KyungEun Lim2

1Statistics Korea, Korea, Republic of (South Korea); 2Statistics Korea, Korea, Republic of (South Korea)

Relevance and research question: Nowadays, Household surveys such as the Labor Force Survey or the Household Finances and Living Conditions Survey face rising “response burden” in Korea. In order to reduce this response burden, Statistics Korea is using a mixed mode survey in the Household survey. But we have questions concerning how much difference exists in the effects of these various modes. Also, how can we design mixed mode surveys in order to reduce these differences. With experimental surveys for mixed mode effects, we are redesigning our mixed mode survey process and are trying to discover the best statistical method to estimate mixed mode effects.

Methods and Data: We conducted a pilot survey for research mode effect. We designed a pilot survey with two survey modes: a face to face interview and an Internet survey. And we added two selection modes with the Household survey, one allowing us to select the survey mode and one not allowing us to do so. Finally, we designed four survey modes, which allowed selecting the face to face interview survey or the internet survey. We sampled 1,600 household members randomly. 800 households selected the interview or internet survey, and the others did not select one particular mode. We provided the questionnaire which included general social survey questions The total number of questions is up to 60.

Results: Finally we collected data from 766 household members with response rates of almost 42%. In some questions, there were mode effects between the face to face interview and the internet surveys. For example, concerning the question “How many times do you contact with your parents in a week”, the respondents doing the internet survey less had less contact than those who opted for the face to face interview. We can show more results which contain mode effects.

Added value: Using the results of the pilot survey, we can redesign the process surveying with mixed mode surveys. Also, we can apply an estimate to the mixed mode data.
Shim-Research on Pilot Survey for Mixed Mode Effects-233.pdf

Not to Be Considered Harmful: Mobile Users Do Not Spoil Data Quality in Web Surveys

Jana Sommer, Birk Diedenhofen, Jochen Musch

University of Duesseldorf, Germany

Relevance & Research Question:

The number of respondents accessing web surveys using a mobile device (smartphone or tablet) has rapidly been increasing over the last few years. Compared to desktop computers, mobile devices have smaller screens, different input options, and are used in a larger variety of locations and situations. The suspicion that data quality may suffer when respondents access surveys using mobile devices has stimulated a growing body of research that has mainly focused on paradata and web survey design. The question whether there are mode effects on the validity of web survey data, however, has been examined in only a few studies. To add to this research, we compared the quality of responses produced by mobile and desktop users responding to a political online survey. To examine data quality, we determined the consistency of the participant’s responses, and validated them against various internal and external criteria.

Methods & Data:

We collected the data of a large sample of participants in a political online survey conducted on the occasion of the 2013 German federal election. The internal-consistency reliability of a political knowledge test was used as an indicator of data quality. As additional indicators, we determined the consistency between self-reported voting intention for the German federal election in 2013 and party preference, coalition preference, and self-reported voting behavior in the previous election. Moreover, we examined the agreement between self-reported voting intention and the actual outcome of the election.


We found no difference between mobile users and desktop users for any indicator of data quality. In fact, the agreement between self-reported voting intention and the actual election result was larger for mobile users than for desktop users.

Added Value:

The present investigation adds to the ongoing discussion whether there are mode effects leading to a poorer quality of data submitted by respondents using mobile devices. Our findings suggest that the participation of mobile users does not compromise data quality, and that researchers do not need to worry about the participation of mobile respondents in web surveys.

14:00 - 15:30Poster Session II: Social Media and Internet Research

The social network analysis of a public debate about the present referendum in Slovakia

Martin Boroš

Masaryk university, Brno, Czech Republic

Relevance & Research Question: Many authors had already shown that social networks are a great research tool for analyzing current topics (Sasahara et al. 2013). In February 2015, the referendum about certain legal rights for homosexual communities takes place in Slovakia. The actors of the public debate had divided into two major groups: supporters and opponents. The goal of the research is to analyze the network of these groups and their activities based on their presence on Facebook and Twitter.

Methods & Data: Both supporters and opponents have several Facebook pages and Twitter accounts. The analysis is based on posts and tweets from 2 biggest Facebook groups and Twitter accounts from both sides and a group of Twitter hashtags and keywords, which the users use in connection with the referendum. The research had already shown that social network theory is a good basis for an analysis of connections between organizations and people with social media data (Hogan 2008). The networks are visualized using the tool Gephi.

Results: The referendum takes place in February 2015. Therefore, the research has not ended yet. However, from the partial results, it is possible to segment users based on their activities. Moreover, data show many interactions within the groups, but much lower level of interactions between the groups. More results would be available after the primary public debate ends after the referendum.

Added Value: The analysis tests the differences between Facebook and Twitter data for an analysis of social networks of users. Moreover, it tests the suitability of social networks for an analysis of current trends and segmentation of users based on their social activity.

Hogan, B 2008 Analyzing Social Networks via the Internet. In: Fielding, N G – Lee, R M – Blank, G 2008 The SAGE handbook of online research methods. Los Angeles; London : Sage.

Sasahara, K – Hirata, Y – Toyoda, M – Kitsuregawa, M – Aihara, K 2013 Quantifying Collective Attention from Tweet Stream. In: PLoS ONE, 8(4).
Boroš-The social network analysis of a public debate about the present referendum-242.pdf

Open Science in Practice — Sharing Research Data in Academia

Benedikt Fecher, Marcel Hebing, Sascha Friesike, Stephanie Linek


Research Question: What Drives Data Sharing in Academic Research?

Relevance: Academic Data Sharing has the potential to use old data for new research questions and to verify research results. It is endorsed by funding agencies, policy makers, journals and researchers alike—nonetheless it is only practiced in moderation.

Methods & Data: We surveyed 1564 individual academic researchers across all disciplines on their dealings with data, their publication practices, and motives to share or to withhold research data in a web survey.

Results: We find empirical evidence for a system that is not driven by monetary incentives, nor the desire for scientific progress, but individual reputation—mirrored in (high ranked journal) publications. We labeled this system ‘Reputation Economy’. We conclude that policy that intend to foster research collaboration need to understand academia as a reputation economy and value intermediate products, such as research data, higher. Only when data sharing pays in form of reputation will it see widespread adoption among research professionals. Most respondents (74%, mean 4.10) agree that other researchers should generally publish their data. To the statement if making their own data available brings more disadvantages than advantages the majority of the respondents, 58% (mean 2.31), disagrees. 80% state, that making data available to other researchers benefits scientific progress (mean 4.34). For the respondents, the most important criteria when publishing results is “Reputation/Impact” (mean 4.15), followed by a “fast publishing process” (mean 3.46) and lastly “Open Access” (mean 3.08). It is noteworthy that those who favor a “Fast Publishing Process” are significantly less likely to make their data available than those who do not. We also present findings detailing on differences among gender and disciplines.

Added Value: For a long time now, scholars and policy makers have been saying how important it is to open up science. Making research data available to others is a prime example for practicing openness in research and fostering collaboration. With the study at hand, we aim to provide empirical evidence for science policies that meet the individual researchers’ demands.
Fecher-Open Science in Practice — Sharing Research Data in Academia-182.pdf

Facilitating Lifelong Learning by Social Media Tools

Oliver Bastian Tristan Franken, Matthias Heinz, Helge Fischer, Thomas Köhler

Dresden University of Technology, Germany

Relevance & Research Question: In our postmodern information society, where knowledge is a key factor for prosperity and welfare, lifelong learning is necessary. To realize a part of this approach, institutions of higher education offer scientific further education programs. Adults, especially fulltime workers, prefer time- and place-independent learning to reach this approach. The usage of social media provides a lot of advantages towards pure in-class-training, which led us to the following questions: (1) Do stakeholders use social media tools in their scientific further education programs? (2) Which social media tools they use? (3) Which strategies are behind the use of social media tools? (4) Why stakeholders do not use social media tools?

Methods & Data: Two different online studies within the ESF-project Q2P helped us to answering these questions. 2013 we used an online content analysis to characterize 404 identified scientific further education programs at institutions of higher education in Saxony. In 2014 a following online survey allowed us to ask the stakeholders (n=173) behind these programs.

Results: (1) 345 of 404 programs are declared as part-time programs. Surprisingly only 13,4% of all scientific further education programs are using digital media. (2) In general social media tools are used little. But there is a significant difference between such like e. g. forum (32%) and e. g. social networks (7%). (3) The main strategies behind the use of social media tools are responding to the participant needs, reaching new target groups, improving image and permitting time- and place-independent learning. (4) Two obstacle groups exist. The lack of knowledge and the lack of willingness have the biggest influence on the disuse of social media tools.

Added Value: The empirical findings show the use, strategies and obstacles of using social media tools in scientific further education programs. It is assumed that the situation in Saxony is representative for Germany and thus allows conclusions for the way how to support the implementation of social media tools in scientific further education at all institutions of higher education to facilitate lifelong learning.

Franken-Facilitating Lifelong Learning by Social Media Tools-162.pdf

Functions and vulnerability of political sphere in the Internet

Alexander Kosenkov

Freelance security analyst, Ukraine

Research Question: The aim of the research is the determination of interrelation and interconnection of political sphere in the Internet and in the real life. The research is based on study of the actual case of informational aggression with wide use of the Internet. The results of the research can be used for analysis of consequences and threats of further convergence of the Internet and the real world.

The research consists of two parts: theoretical and practical. In practical part information warfare applied by the Russian Federation in conflict with Ukraine and role of the Internet in domestic policy of the Russian Federation is researched. In theoretical part conclusions about the interrelation of political sphere in the Internet and in the real life on the basis of information obtained from the practical part of the research are made.

Methods: Case study, media content analysis, observations, unstructured interviews, narrative inquiry.

Results: It was discovered that the Internet is used as an essential part of modern warfare and domestic policy due to the fact that some of basic social processes (e.g. structuration of social groups, evaluation of public support to certain ideas etc) have relocated from the political sphere in the physical reality to political sphere in the Internet. In some cases the influence of the political sphere in the Internet can be compared with the one in the physical realm or can have even stronger impact. Some political groups, whose activity is primarily based in the Internet, even demonstrate certain autonomy from the political sphere in the real life. Such partial 'transfer' of political sphere is sufficient to enable violation of information sovereignty and to influence the society or certain social groups in other countries. In domestic strategy this can be used for 'extension of informational reality' e.g. covert cultivation of ideas, beliefs (e.g. nationalism), spread of which through the traditional media is considered as unacceptable.

Added value: An assumption about particular social functions of political sphere in the Internet and interrelation with the real life political sphere is made. Also potential of exploitation of political sphere in the Internet in informational conflicts is considered.

Lucky Users on Cloud Nine? Applying the Unified Theory of Acceptance and Use of New Technology (UTAUT) on Cloud Computing Usage Behavior with Focus on Perceived Trust and Security Risks

Heiko Moryson1, Guido Moeser2

1University of Giessen, Germany; 2masem research institute GmbH

Background & Research Question: Cloud computing is currently on everyone´s lips and has become a part of the lexicon in our digital landscape. In fact, usage of cloud computing services like Dropbox, Apple iCloud, Microsoft OneDrive or ownCloud are widely-used and have become common place in today's private and business related digital environments. Our research addresses the question of what determinants the usage is based on when viewed from a technology acceptance perspective. We answer questions what the perceived security risks and measures are to ensure trust in the protection of data privacy for cloud users in Germany.

Our poster shows what factors drive the intention to use cloud based storage and backup services. Herby, we quantify the impact on usage intention of determinants like performance expectancy, effort expectancy, facilitating conditions, social influence, perceived security risks and perceived trust. Theoretical frame of the study is the Unified Theory of Acceptance and Use of New Technology (UTAUT).

Methods & Data: Data was conducted via a web survey among internet users in Germany in October 2014. A total of 2135 panellists started the questionnaire with 2040 finishing it. Data basis for the analyses was the number of Cloud Computing users (n=1047). Operationalization was tested using confirmatory factor analyses and causal hypotheses were evaluated by means of structural equation modeling.

Results: Core result is that the UTAUT model generally holds in the case of cloud service usage behavior, explaining 79% of the observed usage intention. This intention is most importantly driven by performance expectancy, facilitating conditions and social influence. Effort expectancy and perceived trust also promote the usage intention while perceived security risks have a negative impact.

Added Value: The results of this research provide implications for cloud service suppliers and marketing. Customers seem to evaluate cloud services on a functional level mostly represented by performance expectancy and facilitating conditions as well as on an emotional level by trust and social influence. Furthermore, the high correlation of trust and social influence, as well as the strong direct impact of social influence to intention, implies application of viral and peer-to-peer marketing techniques.
Moryson-Lucky Users on Cloud Nine Applying the Unified Theory-235.pdf

Maintaining the Audience through Relationship Marketing on Facebook: The Case of Berlin – Tag & Nacht and Köln 50667

Julia Niemann, Michaela Streicher, Michael Schenk

Universität Hohenheim, Germany

Relevance & Research Question: Facebook is an important channel for social media marketing. This holds true even for traditional media like broadcasting shows. Among those, scripted-reality series (SRS) achieve outstanding success on the platform: For instance, “Berlin Tag & Nacht” (BTN) accounts more than 3.3 million Facebook fans by now. These users actively like, comment and share the posts BTN publishes. In their posts, BTN extends the media reality of the TV-show onto the second screen, simulating that fictional TV-characters share their experiences and feelings. This leads to blurred lines between reality and fiction – and probably extends parasocial interaction with the fictional characters.

RQ: Which features of posts on SRS Facebook pages increase user feedback?

Answering this question we intend to draw conclusions both, from a social-media-marketing as well as an audience research perspective.

Methods & Data: We conducted a content analysis, including 282 posts from two Facebook pages of SRS (BTN, Köln 50667). Our codebook included topic, emotionality and language-style, as well as number of likes and user-comments. Additionally, the reactions of the fans were analyzed, based on 1830 user comments related to those posts.

Results: Multiple regression analysis revealed that language-style, topic, emotional valence and a call for dialog of a post accounted for 17 percent of the variance in the number of likes gained but only 4 percent of the number of user comments. Furthermore, users engaged into a para-social interaction with the TV-characters, if the post contained a call for dialog.

Added Value: To conclude, this study gives insight in the new phenomenon of the cross-media connection between TV-series, social media and the user reactions to these additional offerings. From a marketing perspective the study shows how posts have to be formulated in order to turn users into active fans and to motivate them for an ongoing engagement with the media brand. From an audience research perspective the study offers insights in the user perspective and illustrates how parasocial interaction with media figures binds fans to both TV-show and social media offering. Thereby, the TV-show becomes an integral part of the users’ social web reality.

Niemann-Maintaining the Audience through Relationship Marketing-232.pdf

"Germans about Ebola"

Martha Posthofen, Alana Knode, Anna Schneider

YouGov, Germany

How do Germans talk about Ebola and its implications over a time period of six months? The YouGovQualitative team was interested in the particular influence of Ebola's presence in media and effects on our panelists’ perceptions as well as the way they talk with others about that topic.

Between August ‘14 and January ’15 we conducted five Online Focus Groups on Ebola (n=15 participants each). To assure the comparability of the outcomes each discussion was held with basically the same discussion guide. In addition nascent and actual developments of the topic were integrated in later survey sessions.

Our findings show that developments in the way people how talk about Ebola are accompanied by developments of infection rates, single occurrences like Ebola cases in Europe as well as the media presence. Our results demonstrate that the overall perception of Ebola changed over time. Furthermore the outcomes display a connection between certain aspects like the media presence and perceived danger of Ebola for oneself.

The approach using qualitative methods to track how people’s mindsets change over a period of time is innovative and now possible in using online research in a effective and efficient way.

Posthofen-Germans about Ebola-240.pdf

Understanding Scientific Conference Tweets

Steffen Lemke, Athanasios Mazarakis, Isabella Peters

Kiel University - ZBW, Germany

Relevance & Research Question:

Scientists mainly tweet during scientific conferences to share information. Previous studies analyzed tweets only on a daily basis (Ross et al., 2011) whereas we assume that tweets are not equally distributed over the day - instead: Twitter activity might peak due to particular events. The present study analyzes tweets on an hourly and half-hourly basis to show that tweeting peaks are valuable moments for conference organizers to promote or announce information because of high Twitter awareness among participants. We also classified tweets by purpose, the target of a web link (if embedded in tweet), and the tweet content itself to examine the relation between temporal and thematic tweeting patterns.

Methods & Data:

Our test environment for data collection is the conference “VfS-Jahrestagung” (Hamburg, Germany, September 7-10, 2014). The hashtag #vfs2014 was used to find tweets related to the conference. All tweets were collected with the tools TwapperKeeper and Topsy. Classification and content analysis of tweets were based on a validated codebook with three classes: tweet purpose, possible URL-target, and connection of tweet content to conference topic.


During the four conference days many remarkable peaks in the tweeting activity have been detected. The detailed analysis revealed a power law distribution for all conference-related tweets. Only few users tweeted very often, whereas most users tweet only occasionally (maximum number of tweets for an individual user: 78; M = 7.21; SD = 15.55). The applicability of the codebook was confirmed with an inter-rater reliability (Cohens Kappa, two raters) of 0.83 (91% agreement) for the classification of the purpose of tweets, 0.81 (94% agreement) for the URL-target, and 0.60 (65% agreement) for the content.

Added Value:

Despite that the chronological accumulation of tweets is important for conference organizers, also the information about the content of the tweets at different points in time is valuable. The validated codebook is of great benefit for the qualitative analysis of conference-related tweets.


Ross, C., Terras, M., Warwick, C., & Welsh, A. (2011). Enabled Backchannel: Conference Twitter Use by Digital Humanists. Journal of Documentation, 67(2), 214–237.
Lemke-Understanding Scientific Conference Tweets-141.pdf

Digital Strategy: Development of a concept how companies should implement stakeholder's perception and web research behaviour within their online communication

Tobias Looschelders

Deutsches Medieninstitut, Germany

"Never before have so many people engaged so intensively in practices of information search. Billions of searches are performed every day through the internet." (Graham, et al., 2013)

Web search is basically motivated by a user’s intents which are transformed through him to search queries which process websites and their content as the output (Marchionini, 1995; Croft, et al., 2010). The very basis of web search are keywords which equal the path to information in the web. As a result keywords can be seen even as the smallest element in online communication.

Furthermore there is potential for economical relevant data to be retrieved from users’ web search since it reflects intent, interest and goals of users (Jones, 2011; Strohmaier & Kröll, 2009). Under consideration of the fact that search traffic reflects demand for information, it may be even indicating a lack of information of a user (Croft, et al., 2010). There are plenty of examples which describe keyword analyses as part of a search engine optimisation for sales (Parkin, 2009; Jones, 2013).

In prior research it lacks of a transfer of those principles into user research for companies in the web. With such transfer new insights for digital strategy, especially the performance of online stakeholder communication could be gained.

The thesis therefore aims to develop conceptual and analytical approaches for classifying stakeholder groups from keywords and to derive strategic implications from the analysis of stakeholder keywords in the context of:

(1) Performance of communication by increasing findability

(2) Reputation management by securing the online reputation of organisations

In fact, the consideration of stakeholder-keywords within digital strategies could help to build up a company’s reputation and with that the company value. This coincide with the economic goals of the given research of protection and increase of company value. Especially regarding the fact that digital strategy acts as a mediator between a company’s stakeholders and its reputation.

The objective of the primary research is to find which strategic derivations should be made by analysing search queries of users and allocating them to specific stakeholder groups. Consequently, the thesis aims for developing an approach in digital strategy which includes the web search of a company’s stakeholders on a keyword-level.

Stakeholder research in the web was analysed with regard to the top30 FTSE companies in the UK. The primary research has been conducted through a keyword analysis evaluating n=7.545 keywords in combination with the FTSE30 company’s names (e.g. “Vodafone jobs”).The empirical research has been conducted based on the amount of monthly search queries which are retrieved through the autocomplete function of search engine market leader Google Inc. to cover the most user relevant keywords.

By a content analysis of the keywords themselves, in many cases there could be keywords assigned appropriately. In many cases the stakeholder intent is very clear, as for instance in the keyword “HSBC app” which represents a client’s search or in “Unilever share“ which represents a financial search. As a result, there have been n=1,525 unique keywords defined in all the dictionaries which represent a fundament for future stakeholder-assignment of keywords.

The most striking result to emerge from the data is that a major part of 44.86 % of the keywords could be assigned to specific stakeholders.

To further meet qualitative objectives, the underlying search engine results of the evaluated search queries have been analysed. Two items have been analysed within the second part of the primary research, a content analysis of relevant search results (n=899). Firstly, the share of digital media type has been evaluated qualitatively. The sovereignty of communication represents how strong a company is present with their own content in the search results containing the company’s name. It is derived by the share of owned media in search results. As a result, the share of owned media is significantly less compared to the company`s names as a keyword. Derived from that user generated content as well as media content is stronger represented and influencing the reputation of company in the web essentially.

Afterwards, the share of negatively annotated results has been evaluated. Based on the selective perception in the web, the assessment has been based on the short content which is displayed directly in the search engine. The result is that stakeholder-keywords are basically less positive represented than the company-keywords. In average the stakeholder-keywords are reaching 45.6 % positive content while the company-keyword is represented with a share of only 57.0 %. The same tendency occurs regarding the negative results.

In summary, it could be brought evidence that such stakeholder-keywords are significantly more vulnerable by bad reputation.

Practical application is ascertained by a newly developed portfolio matrix for managerial decision making, a four-step action plan and a concept for an integrative monitoring which focuses both on keywords as well as on content has been developed.

Finally, the thesis synthetises that corporate communication could benefit from integrating stakeholders’ research and reputation management in many departments: brand management, public relations, marketing communication as well as financial and science communication. The evidence from this study suggests that analysing and monitoring stakeholders’ research on companies inter-divisional entails a great potential for enhancing business decisions. Additionally, the research suggests that integration of stakeholder research in digital strategy could help to build up corporate reputation and therefore company value.
15:30 - 15:45Coffee Break
15:45 - 16:45A 5: Response Quality & Fraudulent Respondent Behaviour
Session Chair: Ines Schaurer, GESIS - Leibniz Institute for the Social Sciences
Room 248 

Are professional respondents a threat to probability-based online panels?

Joop J. Hox, Edith D. de Leeuw

Utrecht University, the Netherlands

Relevance & Research Question

A major concern about the quality of non-probability online panels centers around the presence of ‘professional’ respondents and several studies have addressed this (AAPOR-taskforce, 2010). Although the definition of professional respondent varies, common indicators are number of panel memberships and number of surveys completed in a specific period. In reaction to criticism on non-probability panels, probability-based online panels are established (e.g., LISS in Holland, GIP in Germany, ELIPS in France, and Knowledge Networks in the USA), reflecting the need for representativeness. However, probability-based panels suffer initial nonresponse during panel formation with the danger of selective nonresponse. R.Q: Are probability-based panels a safeguard against professional respondents? Or are they comparable to opt-in panels regarding count of panel memberships and survey behavior of their members.

Methods & Data

In the Netherlands a large study (NOPVO) of 19 (opt-in) online panels reports on professional respondents (Vonk, Van Ossenbruggen & Willems, 2006). We partly replicated their study in two Dutch probability based online panels: the first is a probability sample of the general population of the Netherlands (LISS panel), and the second a probability sample of the four largest ethnic minority groups in The Netherlands (LISS immigrant panel). We posed the same questions on panel membership, reasons to participate in online research, and the number of questionnaires completed in the last four weeks. We also asked questions on internet use and mobile devices.


In the two probability-based Dutch online panels, the number of panel memberships was lower than in the NOPVO panels: 84.5.% of the LISS members did not belong to other panels and 80.3% of the ‘minority-panel’ did not have multiple panel memberships, while in the NOPVO-panel study only 38% was not a member of multiple panels. In the NOPVO-study on average more than 80% of the respondents reported to have completed more than one questionnaire in the past 4 weeks, while in the two probability-based panels this was less than 40 %

Added Value

Assessing the prevalence of professional respondents in general probability-based online panels and in probability-based panels for special groups. Establishing profiles of professional respondents.

Hox-Are professional respondents a threat to probability-based online panels-127.pdf

PageFocus: Using Paradata to Detect and Prevent Cheating in Online Achievement Tests

Birk Diedenhofen, Stefan Trost, Jochen Musch

University of Duesseldorf, Germany

Relevance & Research Question:

Cheating participants threaten the validity of unproctored online achievement tests. To address this problem, we developed PageFocus, a JavaScript allowing to determine if and how frequently participants abandon test pages by switching to another window or browser tab. To validate whether PageFocus can detect and even prevent cheating, we conducted a parallel lab and web validation study.

Methods & Data:

115 lab participants and 194 members of an online panel completed a knowledge test consisting of 16 items that were difficult to answer, but easy to look up on the Internet. In the experiment, one half of the participants was invited to cheat by looking up solutions on the web. In the second half of the test, a popup message warned participants not to cheat whenever PageFocus detected that a participant was leaving the test pages.


As expected, test-takers invited to cheat abandoned the test pages more frequently and achieved higher scores. Presenting a popup warning successfully reduced cheating rates. With operating system data and the test-takers' self-reports as external criteria, a very high sensitivity and specificity was found for cheating detection based on the PageFocus script. Concurrent evidence from the lab and the web sample suggests that our lab results generalize to testing in online contexts.

Added Value:

Participants switching back and forth from a test to an unrelated application running in the background pose a huge problem to online achievement testing. Our experiment shows that PageFocus cannot only be used to successfully address this problem by detecting cheating participants; PageFocus also allows to reduce cheating by presenting a popup warning whenever cheating is detected. Our results suggest that the validity of all kind of online achievement tests can be improved by adding the PageFocus script to test pages, making PageFocus a promising new tool to improve data quality in online research.

Counting confusion: The role of attitude importance and item clarity in extreme responding

Anton Örn Karlsson1, Vaka Vésteinsdóttir1, Fanney Thorsdottir1, Nick Allum2

1University of Iceland, Iceland; 2University of Essex, United Kingdom

Relevance & Research Question:

Response sets can bias the results of surveys, leading to erroneous results. It is therefore important for survey researchers to assess the form and extent of response bias in order to ensure data quality. Extreme response set (ERS) is one form of response sets that has been identified as a possible threat to survey quality. In short, ERS refers to the tendency of respondents to simplify the response scale of survey questions by using the outermost scale points more frequently than the scale points near the middle. There are some indications that this might especially be linked to cultural differences, making comparisons between different cultural groups difficult.

This study is aimed at assessing to what extent respondents in an Internet survey display a pattern resembling ERS as a function of attitude importance as opposed to resorting to ERS because of the difficulties with the task.

Methods & Data:

The data used were from a 2011 study on immigrants in the Netherland which was a part of the Dutch LISS-panel, administered by CentERdata. ERS was assessed by counting the number of times each respondent selected the most extreme response option in a set of items aimed at measuring attitudes towards different ethnic and cultural groups. The main independent variables were to what extent the respondents felt the issue of immigration was important and to what extent they felt the questions asked were unclear.


First results suggested a significant interaction between attitude importance and the clearness of the questions. For respondents who felt that the questions were unclear, the level of ERS was higher when attitude importance was low, while the opposite was true in the case of respondents whose attitude importance was high, where the level of ERS was lower for respondents who felt that the items were unclear.

Added Value:

The study provides added understanding of the mechanism underlying extreme responding by separating responses of high intensity from responses that can be taken as an indicator of a response set.
15:45 - 16:45B 5: Social Media & Society
Session Chair: Katharina Kinder-Kurlanda, GESIS - Leibniz Institute for the Social Sciences
Room 147 

Social network sites and smartphone's news alerts as an alternative means for news consumption in Israel

Yaron Ariel, Vered Malka

Yezreel Valley College, Israel

Relevance and research question: This study examines the extent to which social network sites and smartphone's news alerts function as a new way of news consumption in Israel, either as an alternative to mass media, or as an additional means.

Methods and data: Data is based on a representative sample (N=508) of the Israeli population aged 15+ (maximum sample error 4.4%). Using an 80 questions- long questionnaire, this research examined various patterns of news consumption via traditional media, as well as via new media (mainly social network sites and smartphone's news alerts), among different age groups and other demographic variables.

Results: Findings reveal that despite of smartphones' ubiquitous presence in Israel, there is a negative correlation (r = -.32, p < .001) between the frequency of smartphone usages and respondents' age. Television viewing has been able to maintain its role as a central source for news (68% of the respondents reported television as their main source for daily news). Nonetheless, ANOVA tests reveal a more complex picture. For example, while adults (aged 45-55) are high consumer of news across all media platform - their daily news come mostly from traditional media such as TV (88%), radio (81%) and printed newspaper (57%). In contrast, younger audiences (aged 15-25) receive their daily news mostly from social network sites (57%) and smartphones' news alerts (40%). Thus, a negative correlation (r = -.14, p < .01) was found between the frequency of keeping updates with current events via social network sites and respondents' age. Findings also suggest that age correlates with the tendency of 'cross-platform consumption' when following important news events.

Added value: During the last few years, social network sites and similar smartphone applications (i.e. WhatsApp) became popular means for interacting and socializing. The findings of this study suggest that these media also serve as an alternative way for news consumption, especially among relatively young audiences. Furthermore, the characteristics of online media enable users flexibility on the time of consuming and the ability of sharing content on their own terms.

Social media in election campaigns: Different channels, different patterns?

Stephen Quinlan, Tobias Gummer, Joss Roßmann, Christof Wolf

GESIS Leibniz Institute for the Social Sciences, Germany

Relevance & Research Question

Social media is now a constant in the daily lives of many people. Accordingly, some forms of social media presence are now considered as a necessary component of modern election campaigns. Yet it is unclear which parties and politicians are using social media, and furthermore, if the patterns of use differ depending on the social media channel?

Methods & Data

We explore the use of Twitter and Facebook by candidates in the 2013 German national election combining data from the German Longitudinal Election Study (GLES) candidate survey (N=1,130) and unique data gathered by GESIS on the Twitter and Facebook use of all candidates that stood in the 2013 German election (N=2,707). Combining these two sets of data allows us to append socio-demographic, attitudinal, and behavioral data with social media data and provide a unique insight into politicians’ use of social media. We devise a series of logistic regressions to explore the determinants and the differences in the patterns of Twitter and Facebook use by politicians.


Our results show that use of social media by candidates is driven by both political and socio-demographic factors and that the patterns of use differ between channels. We find that: a) while Twitter is used by a highly selective subgroup of candidates, Facebook is more widely-used; b) the use of social media is generally the purview of younger candidates; c) Politicians affiliated to the AfD tend to use social media the least, while Pirate Party candidates are strongly predisposed to use Twitter in particular; and d) candidates with greater resources are more likely to use social media.

Added Value

First, the paper has implications for understanding modern electoral campaigning and contributes to the growing literature on social media’s impact on politics, particularly campaigning and the developing literature on the differences between social media channels. Second, our results rely on the total of candidates from the relevant parties and, thus, remedy shortcomings of more selective choices of data. Third, combining survey with social media data enables us to perform in-depth analyses and control for additional variables.
Quinlan-Social media in election campaigns-153.pptx

Users' best friend during a national crisis? WhatsApp and its roles in the lives of Israeli citizens in wartime

Vered Malka, Yaron Ariel, Ruth Avidar

Yezreel Valley College, Israel

Relevance and research question: This study looks at the roles that WhatsApp, the popular smartphone application, played in the lives of Israeli citizens, who were exposed to war menaces during July 2014. Research questions focused on the ways in which citizens used the application, the attributed effects of that usage on their lives, and the possible connections between users' profile, patterns of usage, and perceived implications.

Methods and data: data is based on a representative survey of 500 Israeli citizens aged 16-75, all of whom are smartphone users (maximum sample error 4.5%). The survey was conducted during the third week of the military operation "Protective Edge", which took place between Israel and Hamas in the summer of 2014.

Results: Our findings suggest that WhatsApp played a central multi-functional role in the lives of its users during the wartime, functioning as a mass, as well as an interpersonal communication channel. Participants used the application on a daily basis for various purposes: getting news and updates regarding the war (68%); checking on their loved ones (43%); delivering humorous-satirical messages (50%); spreading war-related rumors (17%); helping to promote voluntary aid initiations (27%). Users expressed their beliefs that the application enabled them to stay updated and 'in the know', helped them calm down, and deepened their communal and national sentiments. Findings also indicate that while using WhatsApp during wartime was common among all group ages, only young user (16-35) used it for news updates as an alternative to traditional new media (similar results were found regarding other social media, e.g. Facebook).

Added value: This research is among the first ones to look at the roles that a popular smartphone application played in the lives of citizens under threats of war. While findings regarding WhatsApp and similar applications usages have been collected for the last few years, no research has yet exposed its centrality under extreme circumstances. Further on, this work suggests that WhatsApp may be thought of as a unique combination between mass and interpersonal communication channels.

15:45 - 16:45C 5: GOR Thesis Award 2015 Competition I: Dissertation
Session Chair: Meinald T. Thielsch, University of Muenster
Session Chair: Frederik Funke, (1) LINK Institut (2)
Room 154 

Open-ended questions in Web surveys - Using visual and adaptive questionnaire design to improve narrative responses

Matthias Emde

Universität Hamburg, Germany

Relevance & Research Question:

One of the most significant decisions when designing survey questions is whether the questions will be posed as closed-ended or open-ended. Closed-ended questions require respondents to choose from a set of provided response-options, while open-ended questions are answered by respondents in their own words. Open-ended questions offer the benefit of not constraining responses and allowing respondents to freely answer and elaborate upon their responses. Narrative open-ended questions are especially useful when there are no suitable answer categories available for a closed-ended question format, or if providing response options might bias the respondents. Open-ended questions are also powerful tools for collecting more detailed and specific responses from large samples of respondents. However, open-ended questions are also burdensome to answer and suffer from higher rates of item-nonresponse. This thesis aims to improve narrative open-ended questions in Web surveys by using visual and adaptive questionnaire design.

Previous research on open-ended questions demonstrated that respondents react to the size and design of the answer box offered with an open-ended question in Web surveys. Larger answer boxes seem to pose an additional burden as compared to smaller answer boxes. At the same time larger answer boxes work as a stimulus that increases the length of the response provided by those respondents who actually answer the question. By varying the visual design of answer-boxes this thesis seeks ways to improve narrative open-ended questions. Despite the influence of different answer-box sizes, the effectiveness of a counter associated with the answer box is tested. In addition dynamic in size growing answer-boxes were compared to answer-boxes that where adjusted in size by respondents themselves.

Besides varying the visual appearance of narrative open-ended questions and the answer-boxes used, the interactive nature of the internet allows a multiplicity of ways to integrate interactive features into a survey. It is possible to adapt Web surveys individually to groups of respondents. Based on previous answers it is feasible to provide specific designed questions to engage respondents. This thesis puts two adaptive design approaches to improve narrative open-ended questions to a test.

Methods & Data:

Despite the influence of three different answer-box sizes, the effectiveness of a counter associated with the answer box that continuously indicates the number of characters left to type is tested. In addition dynamic in size growing answer-boxes were compared to answer-boxes that where adjusted in size by respondents themselves by a plus or minus button.

Further this thesis puts two adaptive design approaches to improve narrative open-ended questions to a test. The amount of information respondents typed into the response box of an initial open-ended question was used to assign them later in the survey to a custom-size answer box. In addition a follow-up probe was tested were respondents who didn’t respond to a narrative open-ended question were assigned to the same question in a closed format to get at least some information from them.

All experiments were embedded in large scale surveys among university applicants or students.


While larger answer-boxes were expected to pose an additional burden, we found no influence of the answer-box size on item-nonresponse. Using a counter indicating the number of characters left curtailed the response length if the default counter value was set low, and increased the response length when the default value was set high. However, a low-value counter limited the number of words used by respondents, but not the number of topics reported. Respondents always seemed to report what they intended to report. Automatically growing answer-box designs do not improve response length or the number of topics reported to narrative open-ended question. In the respondent-adjusted design, respondents were able to set the answer-box size themselves. Since they were aware of the box-size adjustment, the design corresponds better with the question–answer process, as compared to the dynamic growing answer-spaces. As a result, respondents reported more topics and produced longer responses with the self-adjusted answer-box design.

Adapting individually-sized answer-boxes increased only the length of responses to narrative open-ended questions. The answer-box size does not seem to pose a higher burden to respond. As in the visual design experiments, responses can be improved using the adaptive answer-box size assignment, but the willingness to respond is not affected by any of the designs tested. In order to improve response rates, the final experiment in this thesis used a closed-ended follow-up probe to combine the strengths of closed- and open-ended questions. Switching to a closed-ended question is not ideal, but the design accomplished the aim of getting at least some information from former non-respondents. In the initial open-ended question, respondents provided fewer topics but elaborated them. In the closed-ended follow-up probe the respondents checked more answer categories respective to topics in the open-ended question, most likely because they were at hand. Overall, the probe succeeded in getting information from those respondents who neglected to answer the same question in an open format.

Added Value:

Overall, the visual design experiments demonstrate that it is well worth paying attention to the visual and adaptive design of open-ended questions in Web surveys, and that well-designed open-ended questions are a powerful tool for collecting specific data from large samples of respondents.

Further results provide preliminary support for the effectiveness of a Web survey design that adapts the type and visual design of survey questions to the motivation and capabilities of the respondent. While previous studies in the design of open-ended narrative questions aimed to enhance the effectiveness of design features that were meant to influence response behavior (in particular of less-motivated respondents), the adaptive design changes the questionnaire in order to get the most out of the respondent, consistent with his motivation and capabilities.

Data quality in probability-based online panels: Nonresponse, attrition, and panel conditioning

Bella Struminskaya

GESIS - Leibniz Institute for the Social Sciences, Germany

Online panels offer the benefits of lower costs, timeliness and access to large respondent numbers compared to surveys that use more traditional modes of data collection (face-to-face, telephone, and mail). Probability-based online panels in which potential respondents are selected randomly from a frame covering the target population allow for drawing valid conclusions about the population of interest. With the increase in nonresponse over the last years in surveys using the traditional modes of data collection (de Leeuw & de Heer, 2002), probability-based online panels are an appealing alternative. However, whether the quality of data collected by probability-based online panels is comparable to the quality of data attained by traditional data collection methods is questionable. Probability-based online panels are expensive in their recruitment and maintenance, thereby raising the question: are the costs of constructing and maintaining them justified given the quality of data that can be obtained?

There are several sources of errors in online panel surveys: excluding non-Internet users may result in coverage error; if persons selected for the study cannot be reached or do not want to participate, it may result in nonresponse error; study participants may choose to stop participating in later waves (attrition). Furthermore, by taking surveys regularly, respondents can learn to answer dishonestly or answer filter questions negatively to reduce the burden of participation (panel conditioning). All these errors can accumulate in the survey estimates and the conclusions based on these estimates may therefore be misleading.

The goal of this dissertation is to study the quality of data obtained with probability-based online panels. The dissertation aims at advancing the understanding of the causes of errors and at guiding the design decisions when recruiting and maintaining probability-based online panels. This dissertation evaluates the overall quality of estimates from an online panel and focuses on potential sources of error: nonresponse during the recruitment interview, nonresponse to the first online survey, panel attrition, panel conditioning, and the effects of the survey mode.

This dissertation consists of five studies, which are theoretically integrated by the overarching framework of the Total Survey Error (Biemer, 2010; Groves, 1989; Groves & Lyberg, 2010). The framework is extended by including theoretical knowledge on two special types of panel survey errors – panel nonresponse (attrition) and panel conditioning (Kalton, Kasprzyk, & McMillen, 1989). The empirical analyses are based on the data from a probability-based telephone-recruited online panel of Internet users in Germany: the GESIS Online Panel Pilot.

The error sources are studied in connection to the recruitment and operating steps typical for probability-based online panels. Each chapter studies a different aspect of data quality. The chapters are written as individual papers, each addressing specific research questions.

Chapter 1 introduces the theoretical framework of data quality and the Total Survey Error. The goal of Chapter 2 is to evaluate the goodness of the final estimates collected in the online panel. The data from the online panel are compared to the data from two high-quality face-to-face reference surveys: the German General Social Survey “ALLBUS” and the German sample of the European Social Survey (ESS). Furthermore, since researchers may not only be interested in single estimates, I study to what extent survey results are comparable when the data are used for modeling social phenomena. The results show several differences among the surveys on most of the socio-demographic and attitudinal variables; however, these differences average to a few percentage points. To account for the design decision to exclude non-Internet users, post-survey adjustments were performed. However, post-stratification weighting did not bring the estimates from the online panel closer to the estimates from the reference surveys.

Chapter 3 focuses on nonresponse, studying the influence of respondent characteristics, design features (incentives and fieldwork agency), and respondent characteristics specific to the survey mode (Internet experience, online survey experience). The results indicate that participation in the panel is selective: previous experience with the Internet and online surveys predicts willingness to participate and actual participation in the panel. Incentives and fieldwork agencies that performed the recruitment also influence the decision to participate.

Chapter 4 studies why panel members choose to stay in the panel or discontinue participation. The main question is whether respondents are motivated by intrinsic factors (survey experience) or extrinsic factors (monetary incentives). The findings indicate that respondents who view surveys as long, difficult, too personal are likely to attrite and that incentives (although negatively related to attrition) do not compensate for this burdensome experience.

Chapter 5 focuses on panel conditioning due to learning the survey process. To find out if more experienced respondents answer differently than less experienced respondents, I conducted two experiments, in which the order of the questionnaires was switched. The findings indicate limited evidence of advantageous panel conditioning and no evidence of disadvantageous panel conditioning.

Chapter 6 studies mode system effects (i.e., differences in the estimates as the result of the whole process by which they were collected). The data from the online panel is compared to the two reference surveys: ALLBUS 2010 and ALLBUS 2012. Both face-to-face surveys were recruited by the same fieldwork agency and employed an almost identical design. Therefore, they together serve as a “reference mode” when compared with the online panel. I use questions with identical wordings that were present in both ALLBUS surveys and replicated in the online panel. The differences in sample composition among the surveys are adjusted by propensity-score weighting. The results show that the online panel and two reference surveys differ in attitudinal measures. However, for factual questions the reference surveys differ from the online panel and not from each other. Judging by the effect sizes, the magnitude of the differences is small. The overall conclusion is that data from the online panel is fairly comparable to the data from high-quality face-to-face surveys.

The results of this dissertation provide additional insight into the processes that contribute to the quality of data produced by probability-based online panels. These results can guide researchers who plan to build online panels of Internet users or of the general population. The results will also prove useful for existing panels that consider switching to the online mode.

Structure, change over time, and outcomes of research collaboration networks: the case of GRAND

Zack Hayat

The Interdisciplinary center, herzliya, Israel

In this dissertation, I study the interplay between the structure of a collaborative research network and the research outcomes produced by its members. To achieve this goal, I examined GRAND (an acronym for Graphics, Animation and New Media), a Canadian network with over 200 researchers funded by the Canadian government. I use the social network analysis (SNA) framework in order to aid in understanding the structure of the GRAND collaborative research network, as well as how it changes over time. I then look at the changes of the structural network’s characteristics over time and how they interplay with the researchers’ outcomes. Subsequently, I explain this interplay by discussing the changes of the structural network’s characteristics as conditions that can potentially affect a researcher’s social capital and that can, in turn, affect the researcher’s outcomes.

Using data collected through the two online surveys (the first conducted on September 2010, and the second on March 2013), and a research outcomes paper based survey (conducted on May 2013) I was able to capture the research networks (structure and changes over time) of GRAND researchers. These networks captured four types of interaction among GRAND researchers: co-authorship of scholarly publications, communication activity, advice exchange, and interpersonal acquaintanceship. I was also able to get the perceptions of these researchers about their research outcomes. My sample consistent of 101 GRAND researchers, a subset of these researchers (N=50) were then interviewed.

The research outcomes were evaluated based on the four dimensions presented by Cummings and Kiesler (2005): knowledge, training, outreach and collaboration outcomes. The network structure and change over time were measured in terms of: three centrality measures (degree, betweennessand and eigenvector); density; heterogeneity; and effective size (i.e. the number of researchers to which a researcher is connected, minus a "redundancy" factor). Using a combination of social network analysis and statistical regression analysis, I was able to examine the interplay between GRAND researchers’ research outcomes, and their network structure and change over time.

More specifically I found that: (1) Knowledge outcomes (i.e. gains in new knowledge) are positively correlated with the density of the co-authorship network, and with the betweenness centrality, heterogeneity and effective size of advice network; knowledge outcomes are also negatively correlated with the degree centrality of the co-authorship network; (2) Training outcomes (i.e. training of students and post-docs) are positively correlated with the size and density of the advice network, and with the density of the communication network; training outcomes are also negatively correlated with the effective size of the co-authorship network; (3) Outreach outcomes (i.e. the formation of new partnerships/relationships) are positively correlated with the effective size of both the acquaintanceship and advice networks, as well as with the degree and eigenvector centrality of the co-authorship network; finally (4) Collaboration outcomes (i.e. collaboration that has begun and will continue beyond the scope of GRAND) are positively correlated with the degree centrality of the advice network as well as with the degree centrality and density of the co-authorship network. I then distilled and interpreted these results and framed them within the context of related literature. This interpretation was largely based on the insight I gained through the semi-structured interviews I conducted with 50 GRAND researchers.

In this dissertation, I analyze GRAND as a case study of a research network. This enables me to provide a holistic, in-depth investigation of the range, population, and structure of an interdisciplinary, multi-institutional research network. My findings lend support to the argument that social capital and social networks, when combined, yield richer theory and better predictions than when used individually. The social networks analysis conducted in this research offered precise measures of the social structure as well as the changes in the GRAND research network. The social capital-driven findings helped move beyond the relations themselves and to understand how personal relationships or social structures can either facilitate or hinder the achievement of different research outcomes. These results offer to substantiate previous work, while drawing attention to the importance of analyzing interpersonal networks when studying factors effecting research outcomes. This direction for future study is especially relevant, as research collaboration continues to increase both in scope and in importance.

15:45 - 16:45D 5: No borders, no limits? Experiences in multinational market research
Session Chair: Holger Geißler, YouGov Deutschland


  • Dr. Martin Marganus (former Head of Customer Insight, Allianz SE)
  • Simon Kluge (Head of Business Unit BrandIndex, YouGov Deutschland)
Room 158 

The rise of online market research – experiences and challenges in global projects

Martin Marganus

formerly Allianz SE, Germany

Allianz is generating more than 80% of revenues under the Allianz brand. With over 83 mn customers in 70 countries, managing a global brand requires truly global market research capabilities.

Digitalization has a profound impact on consumer behavior and it therefore is little surprise, that online is the method of choice for most quantitative market research projects. The largest project Allianz is running, is the annual brand performance tracking carried out in 22 countries with over 30000 responses. More than two thirds are generated online. Better central steering and quality assurance as well as considerable cost savings for fieldwork and a shorter overall timeline are the main benefits.

There are, however, a number of challenges which need to be managed properly to ensure consistency and quality of data across geographies. Rules have therefore been established to mitigate potential risks.

Nonetheless, there are typical differences between results generated online versus offline. More brands are known and considered for purchase in online research, while image and loyalty results are lower. Open ended questions are answered by fewer respondents, yet the content is richer. Yet our experience shows, that online surveys produce more consistent results over time and across geographies.

The further growth of online research will depend on the availability of high quality access panels. Pure price competition resulted in quality issues especially in emerging markets. Panel provider need to invest in their recruiting processes, create an emotional connection with their members and consider closed expert platforms to stay relevant. Moreover, online research will increasingly take place on mobile devices especially to reach younger target segments and consumers in Asia. This requires responsive design and more concise questions.

Measuring brands globally - It's all about operational excellence!

Simon Kluge

YouGov Deutschland, Germany

Simon Kluge is Head of Business Unit BrandIndex at YouGov Germany and will share insights about the challenges online market researches meet when implementing, running and working with results of an international research brand performance tracking solution. The YouGov BrandIndex service is now available in 20 countries of the world and requires daily online interviews provided by either own or panel partner solutions in each country. In addition, he will share how Yougov deals with the different attitudes of people from various cultural backgrounds when answering scale questions to evaluate brands. That's an additional challenge online researchers need to be aware of in order to ensure reliability and comparability of results originating from various countries.
Kluge-Measuring brands globally-260.pdf
16:45 - 17:00Coffee Break
17:00 - 18:00A 6: Enhancing Survey Response
Session Chair: Anton Örn Karlsson, University of Iceland
Room 248 

How the Timing of Informed Consent on Paradata Use Affects Response Behavior

Sabine Sattelberger

Johannes Gutenberg University Mainz / Department of Political Science / German Politics and Political Sociology, Germany

Relevance & Research Question: Analyzing response behavior using client-side paradata raises ethical questions. In accordance with informational self-determination, German market and social research (ADM et al. 2007) require researchers to ask respondents for consent prior to non-reactive data collection. However, empirical studies indicate that any information about paradata use reduces the willingness to participate in the survey (Couper/Singer 2013). This study examines whether further negative effects of informed consent depend on the timing: Being informed at the end of the survey might provoke drop-outs and refusals of paradata use, whereas being informed at the start might build-up trust, but lead to biased response behavior.

Methods & Data: Cognitive interviewing (n = 12) was conducted to investigate whether the respondents' perception of paradata varies with the timing of informed consent. In a second step, a split-ballot online experiment (n = 754) was used to evaluate negative effects. Apart from an ex-post informed control group, the respondents were offered the option to reject paradata use – either at the beginning or at the end of the survey.

Results: Cognitive interviewing suggests that paradata are regarded as a borderline case of privacy intrusion. The fear of losing one's informational self-determination results from the inability to know what conclusions the researchers draw from paradata. Data misuse is more often discussed when being informed at the end of the survey. In contrast, the front placement seems to make respondents reconsider their answers and complete especially “simple” questions more quickly. However, these effects might weaken over time. The online experiment investigates whether the end position of informed consent increases drop-outs and refusals of paradata use, whereas the front position might affect the response distribution (e. g. item nonresponse, extreme or midscale ratings) and the response process (e. g. response latencies, answer changes, navigation).

Added value: Combining qualitative and quantitative methodology, this paper adds new results to the research on informed consent on paradata use by analyzing how the timing can assist in minimizing drop-outs, refusals of paradata use and biased response behavior.
Sattelberger-How the Timing of Informed Consent on Paradata Use Affects Response Behavior-133.pdf

A quasi-experiment on effects of prepaid versus promised incentives on participation in a probability-based mixed-mode panel

Ines Schaurer1, Michael Bosnjak1,2

1GESIS - Leibniz Institute for the Social Sciences, Germany; 2Free University of Bozen-Bolzano, Italy

Relevance & Research Question:

Research on cross-sectional surveys has shown that prepaid, unconditional incentives are more effective than postpaid incentives. However, no such evidence is available for self-administered longitudinal mixed-mode surveys of the general population. With the emergence of probability-based panels in the social sciences, identifying the optimal incentive timing strategy (prepaid, postpaid) to increase panel survey participation becomes paramount. In our presentation, we will address the research question whether prepaid monetary incentive is superior to postpaid monetary incentive in terms of panel survey participation.

Methods & Data:

We use data from the recruitment phase of the GESIS Panel – a probability-based mixed-mode panel (online, offline using mailed surveys) of the German population. Respondents had been recruited offine (CAPI surveys) and are invited to self-administered surveys every second month. Recruitment took place during May and December 2013. We conducted a quasi-experiment encompassing two groups. The control group (n=4340) consisted of respondents who answered the first regular self-administered survey in 2013. They were promised an incentive of 5 Euro for survey participation. To redeem the incentive, they had to provide their bank account data. The experimental group (n=589) consisted of respondents who answered their first self-administered survey in January 2014. They received a five Euro bill enclosed into a mailed invitation.

Results: First analyses show a substantial increase of survey participation from 77% in the control group to 90% in the experimental group (z = -7.21, p< 0.01). A separate analysis of online and offline respondents show that the increase is more pronounced in the group of offline respondents (66% control, 88% experimental) in comparison to online respondents (84% control, 93% experimental). Besides participation rates, the use of prepaid incentives send out by letter had a positive side effect for panel maintenance, because respondents were highly motivated to keep their address information up to date.

Added Value: The findings suggest that prepaid incentives increase panel survey participation rates compared to promised incentives in a probability-based mixed-mode context. Moreover, sending out prepaid incentives tend to increase panelists´ motivation to keep their contact addresses up to date, and hence, facilitating panel maintenance.

Schaurer-A quasi-experiment on effects of prepaid versus promised incentives on participation-180.pdf

On the Impact of the Presentation Form of Vignettes and the Choice of Response Scales on the Answering Behavior in Vignette Studies

Hawal Shamon

University of Cologne, Germany

Relevance & Research Question:

Vignette studies gained a lot of attraction in the field of social science in the past decade. Some authors emphasize that the use of vignette studies promises more valid and reliable measurements than survey items that are typically used in conventional surveys. Moreover, as vignette studies are based on an experimental plan, they allow to quantify for every respondent the magnitude for each of the presented factors on his or her ratings. Despite the attraction and the advantages of vignette studies compared to survey items, method effects among vignette studies have been rarely investigated.

Methods & Data:

In this split-ballot experiment, I examine in how far the choice of the presentation form of vignettes and the choice of the response scales affect the respondents' answering behaviors. For the purpose of the study, I designed an internet-based vignette study, in which every respondent was presented several vignettes that showed either a textual or tabular presentation. The interviewees were able to enter their answers either in an open or in a closed response format. The study is based on a quota sample for Germany and comprises five experimental groups.


The study results are discussed in terms of interview time (cost aspect) and information losses (quality aspect) and allow to pronounce recommendations for researchers interested in conducting a vignette study. The results suggest the use of a tabular presentation form with altering varying information and a closed response format if the goal is to reduce the interview time (and thus costs). If the goal is to reduce the number of missing observations, it is recommended to design vignettes as table vignettes in which varying information do not alternate, and to offer his interviewees closed response formats.

Added Value:

Reduce survey costs and increase data quality in vignette studies.

17:00 - 18:00B 6: e-Commerce
Session Chair: Rianne van Lambalgen, Hogeschool van Amsterdam
Room 147 

Investment, Purchase or Alms – The Motivation of Investors in Reward-Based Crowd-funding in the Video Gaming Industry

Norbert Steigenberger

University of Cologne, Germany

Relevance & Research Question: I present an inductive and explorative study of the emerging phenomenon of reward-based crowdfunding, where investors commit resources to product development projects in exchange for rewards. The unique risk and value appropriation structure, strongly favoring the developer over the investor, raises the question why investors commit resources to such product development processes.

Methods & Data: The paper is based on an internet survey of active crowdfunding investors, contributing to Kickstarter video game development projects. N=410 investment decisions.

Results: In a mixed-method study in the video game industry, I identify two sub-groups of investors, “buyers”, who are interested in acquiring a product from a specific developer, and “involved” investors, who additionally value insights into the development process and aim for an impact on the industry. The match between a focal project and an investor’s preferences strongly predicts per-capita investments for “involved” investors, while the funding goal and a time effect are relevant predictors for per-capita investments of all investors. I also find that reputation is the core mechanism of risk mitigation. Reward-based crowdfunding resembles a multi-stage game of repeated market contacts, where previous success builds reputation allowing future fundraising. Surprisingly, considerations of distributive justice do not play a role for investors in reward-based crowdfunding.

Added Value: My study advances our understanding of investor motivation in reward-based crowdfunding. I analyze an original sample of crowdfunding investors and obtain information based on statistical analyses as well as inductive text analyses. Reward-based crowdfunding is a dynamic phenomenon, developing rapidly and through identifiable steps. I discuss trends and identify research opportunities, in particular for stakeholder theory, justice and sensemaking research.

Steigenberger-Investment, Purchase or Alms – The Motivation of Investors-115.pdf

The Online Cross-border Shopping Experience

Sjoukje Goldman1, Rick Heldoorn2, Hester Van Herk2

1Amsterdam University of Applied Sciences, The Netherlands; 2VU University Amsterdam, The Netherlands

Relevance & Research Questions: Many e-retailers that are looking for opportunities to expand beyond the domestic market are challenged to develop customized marketing strategies. To be able to target customers in other countries, e-retailers need insight into the motivations and attitudes of consumers. This research adds to the literature on cross-border e-commerce by providing insight into the specific contribution of various factors that affect consumers’ purchase intentions when shopping online in neighboring countries.

Based on the literature on the comparison between online and offline shopping (e.g., Rohm and Swaminathan, 2004), the first factor considered is shopping motivation or consumers’ need for new sources of more variety. The second factor is consumers’ value-consciousness (Sharma, 2011). The third factor and expected to be the main driver is consumers trust in the online cross-border transaction being successful (Beatty et al., 2011). A main barrier may be risk perceptions, associated with a potential loss of time or money (Featherman and Pavlou, 2003).

Method: The survey method was used to collect data among a representative sample of Dutch consumers using an online panel of a marketing research agency. Questions pertained to web shops in the Netherlands and web shops in the neighboring countries, Belgium, United Kingdom and Germany. Respondents were randomly allocated to one of the 4 web shops. Structural equation modelling was used to estimate the model.

Results: Results show that Dutch consumers had lower purchase intentions when visiting foreign web shops compared with domestic web shops. When shopping abroad, Dutch consumers are most willing to purchase at German and Belgian web shops. Consumers’ e-trust is the most important factor explaining purchase intentions. Furthermore, e-trust even reduces the negative effect of consumers’ risk perceptions. Further, higher variety (than provided by domestic web shops) has a positive effect on cross border purchase intention in some, but not all countries. Last, high value consciousness has a negative effect on buying intention at cross-border web shops.

Added Value: This research further enhances our understanding of customers’ cross-border online purchase intentions. Enhancing consumers’ e-trust, and lowering their risk perceptions, will increase consumers’ purchase intentions at cross-border web shops.

Goldman-The Online Cross-border Shopping Experience-118.pdf
17:00 - 18:00C 6: GOR Thesis Award 2015 Competition II: Bachelor/Master
Session Chair: Meinald T. Thielsch, University of Muenster
Session Chair: Frederik Funke, (1) LINK Institut (2)
Room 154 

Sexist Comments in Online Social Networks. How the Degrees of Interpersonal Familiarity and Social Costs Affect the Targets’ Private and Public Responses

Anja Katrin Munder

Westfälische Wilhelms-Universität Münster, Germany

Due to the widespread popularity and various features of different social media sites such as Facebook and Twitter, simulations of these social media platform are promising for use in experimental psychological studies. Well-implemented, they could combine the advantages of high levels of standardisation, high validity and economic online implementation. Especially research on topics of social psychology can be conducted with simulations of social media platforms, as the interaction with other people is pivotal to social media (e.g. chatting, viewing profile sites, reading and reacting to postings).

The present study developed and utilized a simulation of the online social network Facebook to investigate the cognitive, affective and behavioural reactions of a person who is target of sexist discrimination. More specifically, the aim was to examine the impact of interpersonal familiarity between the target person and the discriminating person on the aforementioned reactions of the target person.

Previous research has determined different factors that influence attributional processes, affective responses, and public reactions of people who experience discrimination. However, the potentially important factor of interpersonal familiarity has not been examined yet. To adress this question, 220 participants completed a simulation of Facebook, where they were presented with different postings (status updates, pictures, links to videos etc.) that were published by fictional users, who were indicated as their friends or by unknown users. This was achieved by presenting constructed screenshots in an online survey platform. Cognitive reactions (attributions), affective states (psychological well-being) and behavioural reactions (commenting on a posting, clicking the „like-button“ etc.) were retrieved with the questions formats provided by the survey platform. The manipulation of the degree of interpersonal familiarity of fictional users was successful, as participants significantly rated the familiarity of fictional users indicated as friends higher than of fictional users indicated as unknown.

Participants who experienced discrimination from a friend were less likely to attribute this discrimination as such, but were more likely to confront their fictional counterpart about the discrimination in comparison to participants who experienced discrimination from a stranger. Furthermore, attributing the discrimination as such protected participants’ well-being when the discriminating person was a stranger, but not when the discriminating person was a friend.

Furthermore, it was found in previous studies that people tend to refrain from confronting discriminating behaviour, due to potential negative social consequences (social costs). However, the conducted simulation failed to replicate the inhibiting effect of social costs on public confrontation, likely due to methodological shortcomings.

Practical consequences and implications for further research on the factor interpersonal familiarity in regards to discrimination are discussed. The attempt to implement a simulation of the online social network Facebook in order to manipulate interpersonal familiarity and social costs was partly successful in its intentions. While interpersonal familiarity was manipulated successfully, the factor social costs was not. Participants indicated that they could imagine themselves passably well in the simulation and that their answers were quite similar to their real behaviour on Facebook. Developing and utilizing simulations of social media platforms for further psychological research, especially on topics of social psychology, seems promising. Important issues to consider are the effective operation of experimental manipulations, the technical options (e.g. creating a more interactive interface than screenshots), and the equivalence to face-to-face interactions.

Predicting Response Times in Web Surveys

Alexander Wenz

University of Essex, United Kingdom

Relevance & Research Question:

Survey length is an important factor that researchers have to consider when designing questionnaires. Longer interviews are assumed to impose greater cognitive burden on respondents, which may have a negative impact on data quality. In the course of long surveys, respondents may become fatigued of answering questions and may be more likely to use satisficing response strategies to cope with the cognitive demands. Furthermore, longer surveys lead to increasing costs for questionnaire programming and interviewing.

Despite the impact of interview duration on data quality and costs, survey designers are often uncertain about the length of their survey and only apply rules of thumb, if any, to predict survey length.

The research project presented in this article investigates how item properties and respondent characteristics influence item-level response times to web survey questions. The project builds on the response time analysis carried out by Yan and Tourangeau (2008) and other studies on response times and interview duration and examines whether their findings can be replicated using a different dataset. Finally, the development of a tool for response time prediction is discussed as possible use of the results.

Methods & Data:

The analysis is based on data of the GESIS Online Panel Pilot, a probability-based online panel which consists of German-speaking, Internet-using adults living in Germany. The survey is suitable for studies on response times because it contains a large variety of question types on multiple topics.

Response times to survey items were captured at the respondent’s web browser (client side) by implementing JavaScript code on each survey page. In contrast to response times collected at the web server, client-side response times represent more precise measures of the response process as they do not include downloading times.

Multilevel models are applied to take into account that item-level response times are cross-classified by survey questions and respondents. Assuming that the effect of respondent characteristics is constant over items and the effect of item properties is constant across respondents, a set of random-intercept fixed slope models is fitted. Starting from an unconditional model without any covariates, predictors on the respondent level and the item level as well as cross-level interactions are successively included as fixed effects in the model to account for the observed variation in response times.


The analysis shows that the respondent’s age, education, Internet experience and motivation are significant predictors of response times. Respondents who are younger than 65, have A-levels or Vocational A-levels, use the Internet frequently and are less motivated in survey participation need less time to complete items in web surveys.

Survey participants using smartphones or tablets to complete the survey have longer response times than participants who use desktop computers or laptops. Since the questionnaire of the GESIS Online Panel Pilot was not optimised for mobile devices, mobile respondents may have had problems in reading questions or selecting the appropriate response options.

Among item properties, the complexity of questions and the format of response options were found to have an impact on response times: Survey items with long question texts, many response options and open-ended questions are associated with longer response times, compared to less complex questions with closed-ended response formats.

When comparing the response times to a survey evaluation item across waves which was asked at the end of each survey, it could be shown that respondents become faster in the course of the panel study.

Several results were contrary to prior expectations: The response times of respondents with survey experience who completed at least one survey in the previous year do not differ significantly from the response times of inexperienced survey participants. The position of the item within the questionnaire does not significantly influence survey length, which implies that respondents do not speed up in the course of the survey. Factual questions, not attitude questions, induce the longest response times among all questions types. Moreover, it was found that sensitive items are completed faster than items not dealing with sensitive topics. This finding could be explained by the fact that survey participants may answer sensitive items less thoroughly or may tend to skip these questions. None of the cross-level interaction effects were found to be significant predictors of response times.

Apart from substantive variables, a set of paradata variables which describe the process of questionnaire navigation were included in the model to account for variation in response times. Respondents who scroll horizontally or vertically on the survey page need more time to complete the item on that page. Assuming that response times are an indicator of respondent burden, this finding implies that survey pages have to be adapted to the respondent’s screen size to avoid scrolling. If survey participants leave the survey window, for example to access another webpage in the browser, or revisit survey pages to edit their response, this also has an effect on response times.

Added Value & Limitations:

The present analysis could replicate many findings from previous studies on response times using a probability-based online panel where response times were collected at the client side. However, some results were not in line with previous research and need to be investigated by future studies.

Beyond replication, the study contributes to existing research by examining response times in the context of a panel study and demonstrating that respondents speed up across survey waves. Furthermore, the analysis indicated that paradata describing the process of questionnaire navigation are significant predictors of response times and should be collected and controlled for in future response time analyses.

The limitations of the present study are the very low response rate of the GESIS Online Panel Pilot and the small number of observations in the analysis compared to previous studies on response times. Although the sample sizes are sufficiently large to estimate multilevel models, the statistical power of the fitted models may be reduced. Moreover, the nested structure of survey questions within waves, additional to the cross-classified structure of survey items and respondents, has not been considered due to a small group size at the level of survey waves. To address these shortcomings, response time data from online panels with larger sample sizes and a larger number of survey waves have to be analysed.

Website Evaluation at Different Phases of Website Use

Leonie Flacke

Westfälische Wilhelms-Universität Münster, Germany

Relevance & Research Question: Every day, the average German Internet user spends nearly three hours (van Eimeren & Frees, 2014) on the Internet. Considering the myriad of websites the Internet contains, it becomes obvious that users are forced to make a rigorous selection of which websites to visit. Thus, if you were a website owner, you would need to instantly convince users to regard your website and not any of your very many competitors. But, how would you evoke your users’ interest? What makes them stay on your website or make them even revisit and recommend it at best? Which aspects of your website are imprinted in their memory?

Previous research detected the three core constructs content, usability, which can be further split into subjective and objective usability (Hornbaek, 2006), and aesthetics to influence the users’ perception and evaluation of websites (Cober, Brown, Levy, Cober, & Keeping, 2003; Hartmann, De Angeli, & Sutcliffe, 2008; Thielsch, Blotenberg, & Jaron, 2014). However, only little is known about their interplay. Indeed, Thielsch et al. (2014) developed a path model concentrating on the factors’ effects at four different phases of website use. They found aesthetics to be the critical factor influencing both the first and the overall impression, whereas the content of the website turned out to be decisive for the intention to revisit and to recommend the website.

The aim of this current study was to extend the results of Thielsch et al. (2014) and to enrich it by the measurement of the objective usability (as recommended by e. g. Hornbæk, 2006; Lee & Koubek, 2010) as well as by the distinction of the immediate and deliberate first impression (Leder, Belke, Oberst, & Augustin, 2004; Thielsch et al., 2014). A second time of measurement was implemented in order to investigate the long-term effect of the factors content, subjective usability and aesthetics. Despite the importance of long-term investigation (Mittal, Kumar, & Tsiros, 1999; Karapanos, Zimmermann, Forlizzi, & Martens, 2009) research in the field of website evaluation is still in the early stages. Therefore, the question of the importance of the factors content, subjective usability and aesthetics and how they change over time has been investigated in an explorative manner.

Methods & Data: As mentioned above, the online study consisted of two times of measurement. The participants were recruited via the German online panel ‘PsyWeb’. At the first time of measurement (T1), 306 participants took part. They were randomly assigned to one (out of ten) website. Their age ranged from 16 to 70 years (M = 44.19, SD = 15.09). Of the participants, 163 (53.29%) were female, 142 (46.41%) were male and one participant denied to give information about the gender. On average, the participants had used the Internet for 14.53 years (SD = 4.80). The different phases of website use were operationalized by different presentation durations (immediate first impression: 1 sec, deliberate first impression: 10 sec, overall impression: unlimited, intention to recommend and intention to revisit: website has already been faded out) of the website. The participants were asked to rate their impression of the website at the different phases of website use (dependent variables) as well as the independent variables content, subjective usability and aesthetics via different questionnaires, for instance: WWI + G (Content; e.g. Thielsch et al., 2014)., WWU (Subjective Usability; Moshagen, Musch, & Göritz, 2009, SUS (Subjective Usability; Brooke, 1996; translation of Rauer, 2011) and VisAWI (Aesthetics; Moshagen & Thielsch, 2010). By contrast, the objective usability was operationalized by two interactive tasks the participants were asked to answer, at which the task accuracy served as criterion.

The design of the second time of measurement (T2) was similar to T1, except for the objective usability which could not be measured again. At T2, 224 participants (drop-out rate: 26.80 %) continued with the study on average five days after T1 (M = 118.32 hours). They were not able to regard the website again. Thus, they were requested to judge the website’s content, subjective usability and aesthetics again from memory. The same questionnaires as at T1 were used.

Results: The impact of the factors at the five different phases of website use was calculated by a structural equation modeling (SEM). Apparently aesthetics was the major factor influencing both the immediate and the deliberate first impression, followed by the content. Regarding the overall impression, aesthetics also turned out to be the decisive factor, albeit more densely followed by the subjective usability and the content. By contrast, the content had the largest effect by far on the intentions to revisit and to recommend the website. The pattern was the same at T2 since aesthetics persisted to be the most important factor for the overall impression, whereas the content was the sole decisive factor for both the intention to recommend and to revisit the website.

Added Value: Website owners should definitely not underestimate the importance of aesthetics, since this is the factor which evokes the users’ interest at first. The significant role of aesthetics is emphasized regarding T2, at which it was still the decisive factor for the overall impression. Users mainly remember the aesthetics when judging their overall impression from memory. Thus, aesthetics is not only a nice-to-have but a must-have item since it does indeed have a long-term effect. If you wanted the users to recommend the website to others and to revisit it, you should pay special attention to the content of your website. The results of T1 draw a line to the Elaboration Likelihood Model (Petty & Cacioppo, 1986). Hence, people process the website on the peripheral route at first since they mainly take into account the aesthetics (a superficial factor) to form a judgment of the immediate and deliberate first impression. As the judgment of the overall impression demands a more complex dealing and an interaction with the website, the central route gains in importance, even if aesthetics is still the decisive factor. Regarding the intention to revisit and to recommend the website, the content is now the significant factor. Thus, people now assimilate the website on the central route. The results of T2 confirm the strong effect of the content on the two intentional outcomes. Additionally, this study showed further evidence for the need of a distinction between subjective and objective usability. Thus, if a website owner constructs a perfect objectively usable website, users may yet find the website difficult to use.
17:00 - 18:00D 6: Instant Research: Quick without dirty?
Session Chair: Olaf Hofmann, SKOPOS Institut für Markt- und Kommunikationsforschung GmbH & Co. KG


  • Andreas Woppmann (Managing Director, advise research)
  • Oliver Trabert (CTO, Questback)
  • Barbara Garau (Head of Global Solutions, Innovation & Forecasting,Ipsos InnoQuest)
Room 158 

Speed-Up! Tailor made solutions for urgent decisions

Andreas Woppmann

advise research gmbh, Germany

In recent years, time-to-market has become one of the most important criteria for companies to be one step ahead of competitors. But market research is often blamed on slowing down the product development process. Hence, in urgent cases, market research projects are often cancelled and decisions are based only on management’s gut feeling. This might be sufficient in some cases but lead to the question: is there a way to collect and analyze consumer data significantly faster? In addition to existing speed-tools that are mainly based on standardized questions, many companies already have their own research methods. As these are already well-known and accepted by all stakeholders, there is a great need to speed-up existing clients’ research methods.

Based on different projects, e.g. for SPIEGEL ONLINE with n=1.012 German WhatsApp user and an urgent multi-country study in Mexico, Czech and Russia, we want to show how non-standardized surveys can deliver quick results without being “dirty”. To achieve this, projects need acceleration in all research steps beginning with a focused test design, a high internal standardization by using innovative software tools and a speed-up of fieldwork. The latter is often discussed by the research society, especially for online research. But showing results of some online experiments will give an answer to this. To be quick not dirty, it is crucial to have market representative quotas and to calculate significant differences for example.

Value added: while this approach is very suitable for well-defined purposes with concrete stimulus-materials, e.g. concepts, packs or advertisement it is rather less recommended for strategic purposes. In general, it is not only possible for online research but also applicable for Studio tests or CATI interviews, quantitative as well as qualitative.

Market Research on the Fly – A Contradiction in Terms?

Oliver Trabert

Questback GmbH, Germany

Today an app exists for everything: chatting, cooking, healthcare, exercising and many more. The “DIY” trend is more prominent than ever. Why?

Because it is easy and technically feasible. But is it also appropriate for market research?

The answer: It depends. Not a very satisfying answer to be honest, but for single spot questions and polls free tools and apps can be a cost-efficient way to collect top-level or event-driven insights. However, for complex, sophisticated insightful questions, self-made research is stretched to its limits.

In his statement, Oliver Trabert, CTO of Questback will show when Do-It-Yourself-research can create value and when more profound research is needed.

Demanding Better – Not Just Faster – Research in the Digital Age

Barbara Garau

Ipsos InnoQuest, France

Our world is changing and the pace of that change is astonishing. Technology is developing at an exponential rate. Consumers want you to dazzle them –to give them the ‘wow’ – and they don’t want to wait for anything. Organizations are changing too. Marketers are being pushed to do more with less and to get to market faster. Innovating in this context is challenging and marketers are faced with a dilemma: get research results they can count on or take a chance and sacrifice quality for speed.

We will present a case study on how a restaurant client not only tested their ideas overnight and got validated results but also developed a new innovation process to accelerate speed to market moving forward.

For this client, the end result was not just faster results: the client actually walked away with more consumer insights than had been generated with a traditional approach to innovation. Based on this experience, along with similar experiences with clients who refuse to give up quality for speed, we will discuss how harnessing the power of the digital age shouldn’t be just about faster research, but about better research as well.
Garau-Demanding Better – Not Just Faster – Research-261.pptx
20:00GOR Party

Location: Gewölbe (Hans-Böckler-Platz 2, 50672 Cologne)

The GOR Best Practice Award 2015 will be awarded at the Party!

The GOR Party is sponsored by SSI.

Drinks and fingerfood included in the ticket price.
Please note: You need a ticket to attend the party. Your party ticket will be included if you have registered for the complete conference or if you own a day ticket for Thursday!

Date: Friday, 20/Mar/2015
8:30Begin Check-In
9:00Track A: Online Research Methodology and Internet Surveys

Sponsored by GESIS

9:00Track B: Internet and Society

Programme Partner: HIIG

9:00Track C: Social Media Research
9:00Track D: Angewandte Online-Forschung (Applied Online Research) in German

Programme Partner:

9:00 - 10:00A 7: Consumer Research
Session Chair: Holger Nowak, YouGov Deutschland
Room 248 

Online Consumer Search Behaviour: An International, Cross-Sector Analysis

Julia A Jacobs1, Christopher P Holland2,1, Stefan Klein1

1University of Münster, Germany; 2University of Manchester, United Kingdom

Relevance & Research Question: Online Consumer Search Behaviour, Consideration Set.

The Internet has revolutionised access to information and transformed consumer search and buying behaviour. (Hu et al. 2014, Dinner et al. 2014). The consideration set is an important construct in marketing to measure search. Contrary to economic theory (Stigler, 1961), existing Internet studies report relatively small online consideration sets between 2 and 3 (Johnson et al. 2004, Meyer & Stobbe 2010, Holland & Mandry 2013). Our research questions are: 1) Why are online consideration sets relatively small? 2) How do search strategies evolve? 3) What are the differences in search behaviour for constellations of consumers defined by: a) Product and market sectors b) Demographic variables, c) Differences in media use, i.e. computer versus mobile devices, d) Nationality and culture.

Methods & Data: Clickstream Analysis, Online Panel Data.

Clickstream analysis provides granular data on a very large scale, and can therefore be used to develop research experiments and insights into actual consumer behaviour compared with surveys, which rely on recalled behaviour and statements of future intent. International panel data from ComScore is used to investigate differences in consumer behaviour across six market sectors: banking, insurance, grocery, telecommunications, automotive and airline.

Results: International results, small online consideration sets

Initial results from the US, UK and German markets show that the online consideration sets are relatively small. This is contrary to economic and marketing theory, which predicted more extensive search patterns because of lower search costs. Online search also appears to be more narrow than similar pre-Internet studies.

Added Value: Empirical contribution, theory development, commercial data sets

The research builds on the few previous empirical uses of online panel data, and develops an ambitious international, cross-sector empirical database. Most research in this area, including recent survey based research, is typically a snapshot of activity in a single sector within one country, using sample sizes of between 30 and a few hundred. This research is based on longitudinal data samples with sample sizes of millions of users.

Jacobs-Online Consumer Search Behaviour-150.pdf

Modeling Online Hotel Choice: Conjoint analysis as a multivariate alternative to A/B-testing

Christine Hübner, Paolo Cordella

SKIM, Netherlands, The

Relevance & Research Question: How can researchers best investigate what drives online purchase decisions? Currently, A/B-tests of isolated features on two otherwise identical versions of a website are the gold standard for testing online purchase environments. Despite its simplicity, A/B-testing is tedious since only a single feature can be changed at a time and possible interactions between features are not accounted for.

In this study, we examine whether conjoint analysis can be used as a multivariate alternative to A/B-testing. Choice-based conjoint is a powerful method to analyze purchase decisions as it allows for optimization across various attributes and levels. While in most conjoint applications attributes and levels pertain to the product itself, the method can also be applied to optimizing the user interface (where elements are placed, how they are designed and executed, the use of rankings, reviews, and promotions). We apply choice-based conjoint to test which booking site features impact purchase decisions for hotel accommodation.

Methods & Data: We conducted an online survey with N=1492 respondents, asking them to choose hotels in a choice-based conjoint experiment as if on a booking site. We manipulated user interface features such as position of items, ranking options, reviews, and promotions using an orthogonal experimental design. To mimic the purchasing process and click-through in a live A/B-test, we assume that each respondent goes through one choice task only. We estimate the impact of feature changes on respondent preferences for hotels and assess model fit using out-of-sample prediction.

Results: Robust model estimation is possible at the aggregate level taking into account a single choice task for each respondent. Using the model estimates we arrive at robust predictions of hotel choices and click-through rates. The results give valid insights into which site features impact online purchase decisions.

Added Value: The study illustrates how conjoint analysis can be used as a multivariate alternative to A/B-testing of online shopping sites. Results of preference estimation are useful for optimizing the design and execution of online shopping environments and for playing what-if-games with a multitude of variables rather than just single isolated features.
Hübner-Modeling Online Hotel Choice-183.pdf

The role of market research online communities in qualitative market research: Increased situational validity for customer journey measurement concerning low-interest products

Stephan Groll1, Torben Tietz1, Zacharias De Groote2

1MSR Consulting Group GmbH, Germany; 2Liveloop GmbH, Germany

Relevance & Research Question: It is well known that prototypicality and situational validity are key success factors for any qualitative research. Qualitative online research offers new opportunities for gaining deep insights. Especially for low-interest products and target groups with little experience, market research online communities (MROCs) offer a better approach to understanding the customer journey. In our presentation, we will show that exercises during a MROC will – with the same or a lower investment - lead to a deeper understanding than face-to-face approaches.

Methods and Data: For our presentation, we will draw on data from an extended MROC vs. data gathered through face-to-face interviews and focus groups. The aim of the research was to find out how young people approach financial services. We will compare the findings of both approaches and show potentials and limitations for each method.

Results: Our results show that MROCs work especially well for low-interest products. Without prior experience in buying processes for a product, different tasks are used to track the information process over time. This makes the research process for each individual much more naturalistic as information gathered will influence the usage of sources to solve the different tasks.

Added Value: We will not only argue that MROCs are (for some target groups and research questions) advantageous with regard to the research validity but also in an economical sense: With MROCs researchers can answer research questions for costs that are often lower than for more traditional qualitative research designs.

Groll-The role of market research online communities in qualitative market research-164.pdf
9:00 - 10:00B 7: Technology Acceptance
Session Chair: Birgit U. Stetina, Sigmund Freud University
Room 154 

The reciprocity effect: how corporate transparency leads to voluntary sharing of personal data

Joris Demmers, Rosanne de Vos

University of Amsterdam

Relevance: Companies increasingly collect large amounts of personal consumer data from a multitude of sources, allowing marketers to target consumers with more precision than ever before. In the upcoming era of the “internet of things” and the “quantified self”, data-driven solutions will dominate the marketing landscape. Although effective use of personal data ensures that consumers are exposed to more relevant marketing content, consumers often react against the collection of their data. This reluctance is often attributed to a lack of understanding on when and why data is being collected, the perceived lack of control over the collection and use of the data, and privacy concerns. Leading scholars and practitioners agree that retaining consumer trust will become the key challenge for the effective use of personal data for marketing purposes. In the current study, we address this issue and investigate whether corporate transparency can trigger a process of reciprocity ultimately leading to an increased willingness to share personal data.

Method: We conduct an online experiment in which we ask 200 participants to share personal data they provided in the context of a study on shopping behavior and preferences with a leading food producer. To test the proposed reciprocity effect, we compare a condition (1) in which the firm proactively discloses information on its subpar performance with regards to sustainability to a condition (2) in which the firm explains how exactly it uses personal data and a control condition (3). To distinguish between the effect of mere information availability and the effect of corporate transparency, we added two conditions (4 and 5) in which information from condition 1 and 2 is provided by the researcher rather than the firm.

Results: Data is currently being collected.

Added Value: This study demonstrates if and how corporate transparency can trigger consumer reciprocity ultimately leading to voluntary sharing of personal data. If the results support our propositions, the study contributes to the debate on the collection and use of personal data by showing that by being transparent, firms can collect personal data and at the same time build rather than undermine consumer trust.

Evaluation of Technology Acceptance of Data glasses based on an application for Smart Ski Goggles

Bernhard Klaus, Daniela Glatz, Astrid Tarkus

evolaris next level, Austria

Research question: User-experience as well as the perception of a technology are playing a decisive part in the decision to adopt a new technology. Data glasses are becoming more and more popular and increasingly catch the attention of many researchers, developers and also the general public. In our study we aimed to evaluate the technology acceptance of a prototypical data glass application in the leisure and amusement sector: an application for smart ski goggles providing information in a built-in display for skiers and snowboarders on the slope.

Method & Data: We established a User Experience (UX) Model, based on the Technology Acceptance Model (TAM) by Davis. The model consists of the factors “Personal Value”, “Social Value”, “Task Value”, “Satisfaction”, “Attitude” and “Intention to Use”. After defining the target group based on an online survey representative for Austria (n=1000), a user experience test was carried out in an Austrian skiing region in April 2014 to capture the users’ perceptions on-site. In addition to testing the smart ski goggles and giving in-situ feedback through the goggles’ control element, the 54 respondents filled in an ex-ante and ex-post questionnaire designed to collect quantitative and qualitative data. Furthermore, real-time context data was tracked via GPS-sensors and a backend system. For the analysis we used PLS (Partial Least Squares) to test the model.

Results: According to the results, the “Personal Value” is the most important factor and neither “Task” nor “Social Value” have an impact on “Satisfaction” and “Intention to Use”. The quality criteria clearly show that the model is suitable to describe the user experience which is also mirrored by a high coefficient of determination regarding the factor “satisfaction”.

Added Value: The results provide insights into the specific values that influence the satisfaction with and usage intention of the new data glass technology applied in the leisure and amusement sector. The UX model can be applied by the scientific community and expanded according to the type of technology being reviewed. Moreover, the results deliver useful information for researchers regarding general needs of the target group of data glasses applications in the above mentioned sectors.

Klaus-Evaluation of Technology Acceptance of Data glasses based-152.pdf

Debunking the Diagnosis Internet Gaming Disorder: Motivational differences between high engagement and addiction in a German sample of World of Warcraft players

Mario Lehenbauer-Baum1,2, Zuzana Kovacovsky1, Armin Klaps1, Birgit U. Stetina1

1Sigmund Freud University, Austria; 2Vanderbilt University, Nashville/Tennessee

Relevance & Research Question:

The DSM-5 introduced Internet Gaming Disorder (IGD) as a condition needing more research. Proposed criteria include amongst others tolerance, preoccupation, deceiving, or continued excess despite psychosocial problems. However, some studies suggest differences between addiction (criteria like conflict, withdrawal symptoms, relapse and reinstatement and behavioral salience) and engagement (criteria like cognitive salience, tolerance and euphoria). Other studies highlight additional motivational factors as especially relevant in gaming such as immersion. The research questions of the presented study were if a difference between engagement and addiction is visible looking at motivational aspects and if there are gender related differences regarding motivation.

Methods & Data:

Using an online-based questionnaire, we surveyed 676 adult (18+) volunteers (mean age 23.26 years; 84.9% male) from German speaking parts of the world (Germany, Austria, Switzerland). In addition to demographic questions an adapted version of the “Asheron’s call” questionnaire (which covers six addiction criteria including salience, euphoria, tolerance) and the Gaming Motivation Scale (Yee, 2006) were used to answer the research questions. According to existing criteria only 127 of the participants could be grouped in engaged versus addicted. Statistical analyses for hypotheses testing included MANOVA, ANOVAs between groups and t-tests for gender differences.


Significant results (F(3, 123)=57.950; p<.001) show highly relevant differences between the groups regarding all motivational aspects. ANOVAs reveal that the relevant motivational factors to differentiate between engaged and addicted gamers are achievement (F(1, 125)=165,477; p<.001) and immersion (F(1, 125)=51,332; p<.001). Achievement (t(674)=-4.027; p<.001) and immersion (t(674)=3.602; p<.001) values are in addition significantly higher in male participants and women show higher values regarding social motivation (t(674)=2.559; p=.011).

Added Value:

Motivational aspects obviously play a huge role to differentiate between problem behavior (addiction) and engaged gaming. In addition gender differences need to be addressed even more carefully. Women tend to be less interested in achievements from a motivational point of view and immersion as motivational factor is also not as relevant as for men, but the social aspect seems to be the typical “female” motive. Gender sensitive models have to include that aspect in the future.
Lehenbauer-Baum-Debunking the Diagnosis Internet Gaming Disorder-200.pdf
9:00 - 10:00C 7: Social Theory in Social Networks
Session Chair: Katharina Kinder-Kurlanda, GESIS - Leibniz Institute for the Social Sciences
Room 149 

Minority Status and the strength of Facebook Ties: A test of Social Diversification perspective in the US

Gustavo S. Mesch

University of Haifa, Israel

Relevance & Research Question: Multicultural societies are characterized by social groups that hold different positions in the stratification system (natives, immigrants and disadvantaged minorities). This study investigated the role of race and ethnicity in the U.S. on self- reported composition of social ties of young adolescents on Facebook and tested the social diversification perspective that argues that in multicultural societies race and ethnicity are central social status that shape the nature of associations.

Methods & Data: A secondary analysis of the 2012 U.S. Pew Teens and Privacy survey was conducted (n=625). Respondents were presented with a list of potential contacts and were asked to indicate if these individuals are in their Facebook network. From the responses two mutually exclusive variables were created. Strong ties is a composite variable that was created from positive responses to the request to indicate if parents, siblings, extended family or friends at school are members of the youth Facebook network. Weak ties is a composite variable that was created from positive responses to the request to indicate if friends that are not at school, teachers or coaches, celebrities, musicians or athletes and other people you have never met in person are in the personal Facebook network. O.L.S. regression was used to investigate the effect of socio-demographic and Internet use variables on the strength of ties.

Results: The average number of strong and weak ties reported differs according to race/ethnicity. White adolescents report on average more strong ties , than African Americans while African-Americans report on average more weak ties than white Americans.

Added Value: We found support for the diversification hypothesis, which maintains that disadvantaged groups take more advantage of the Internet as a means of overcoming social inequalities in society to increase their social capital. The results provide strong support for the diversification hypothesis, suggesting that members of disadvantaged groups connect to weak ties rather than strong ties to increase access to non-redundant information.

Mesch-Minority Status and the strength of Facebook Ties-109.pptx

Self-disclosure on Facebook: Social capital and the match between appropriate media channel and type of disclosed information

Uwe Matzat1, Ruoyun Lin2

1Eindhoven University of Technology, The Netherlands; 2University of Tübingen, Germany


Self-disclosure (SDC) on Facebook is helpful for gaining social capital. The link between amount of SDC and social capital has been studied intensively. However, it is unclear what role different types of SDC in specific communication channels play for accessing social capital. We challenge the idea all types of SDC in all channels are equally helpful for social capital development. Rather, we test the claim that a “match” between the type of information (intimate vs. other information) and type of channel (private vs. public) facilitates access to (bonding) social capital. Our study rests on the idea that users regard self-disclosure of intimate information in public media often as inappropriate (Bazarova, 2012).


We use two experiments combined with survey measurements on users’ social capital (Ellison et al., 2011) of 380 Facebook users to test several hypotheses. Participants were randomly assigned to one of three Facebook channels (Chat/Group/Status Updates), and their patterns of disclosing intimate and non-intimate information in that specific channel were measured. The three Facebook channels were assigned between-subjects and the two categories of SDC (intimate vs. non-intimate) were assigned within-subjects. In experiment 2, the perception of the appropriateness of self-disclosure is analyzed. Here, we utilize a 2×2 factorial design, with disclosure channel (Chat vs. Updates) as within-subjects factors, and disclosure intimacy (high vs. low intimacy) as between-subjects factors. Similar to Bazarova (2012), Facebook pages with fictitious self-disclosure messages were created and their perceived appropriateness was measured.


For their own disclosure of intimate information on Facebook, users prefer private over public channels. Furthermore, users regarding public disclosure of intimate information as appropriate are more likely to disclose intimate information in public. Most important, while “matched disclosure” (disclosure of intimate information in private media) is associated with more bonding social capital, disclosing intimate information in public media is not.

Added Value:

The findings point to an important rule of self-disclosure that governs privacy management and characterizes media affordances. To the best of our knowledge, we are the first to show that users who disregard these affordances face barriers in accessing bonding social capital on Facebook.
9:00 - 10:00D 7: Mit dem Kunden statt über den Kunden: Customer Feedback und Customer Integration
Session Chair: Horst Müller-Peters,


  • Daniela Klos (Head of Customer Strategies, TNS Infratest)
  • Bernhard Keller (Director, MaritzCX)
  • Martin Grupe (ehemaliger Geschäftsführer Vision Critical GmbH)
Room 158 

Kulturwandel Kunde – wie man mit „Customer Feedback Management“ Unternehmen zu kundenzentrierten Organisationen transformiert

Daniela Klos

TNS Infratest, Germany

Die wichtigste Grundlage für Erfolg und Wachstum eines Unternehmens sind seine Kunden. Die Erfahrungen, die Kunden mit einem Unternehmen machen, können starke und damit gewinnbringende Beziehungen schaffen – oder auch die Geschäftsentwicklung nachhaltig gefährden.

Die meisten Unternehmen haben das schon lange erkannt. Kundenorientierung hat für sie höchste Priorität: ihre Geschäftsführer haben Customer Experience Strategien aufgestellt und Leistungskennziffern definiert; ihre Mitarbeiter verfügen über Kundenzufriedenheitsmessungen und plattformübergreifende Verfahren für das Beschwerdemanagement.

Damit verfügen viele Unternehmen zwar über strategisch und operativ relevante Informationen über ihre Kunden, sind aber oft kaum in der Lage, die „Voice of the Customer“ in die Organisation hineinzutragen – und damit zu ihren Mitarbeitern mit Kundenkontakt. Die Folge: sie reagieren auf operative Probleme nicht schnell genug und gefährden bestehende Kundenbeziehungen.

Wahrhaft kundenzentrierte Unternehmen haben dagegen einen wesentlichen Wettbewerbsvorteil: sie kennen die Bedürfnisse ihrer Kunden und die für sie entscheidenden Touchpoints. Und sie sind in der Lage, ihre Prozesse an diesen Kontaktpunkten konsequent am Kunden auszurichten. Ganz konkret heißt das: die Mitarbeiter an der Kundenfront können jederzeit handlungsrelevante Insights zu jedem Kontaktkanal abrufen und ihre Handlungen unmittelbar darauf abstimmen.

Untermauert durch weltweite Forschungsergebnisse (basierend auf 40.000 Interviews in 20 Märkten, überwiegend CAWI) spricht Daniela Klos in ihrem Vortrag darüber, was das Buzz-Word „Kundenzentrierung“ wirklich bedeutet und wie Unternehmen in einem zunehmend aggressiven Wettbewerbsumfeld profitables Wachstum erreichen können. Dabei zeigt sie, wie traditionelle Ansätze aus der Kundenbeziehungsforschung durch innovatives Customer Feedback Management ergänzt werden können – umso nachhaltige Kundenbeziehungen aufzubauen und einen Kulturwandel im Unternehmen zu erreichen.
Klos-Kulturwandel Kunde – wie man mit „Customer Feedback Management“ Unternehmen zu kundenzentrierten Organ.pdf

Customer Feedback Management hat zum Ziel, Unternehmenswerte zu stabilisieren und ausbauen. Der Weg dorthin: den Kunden in den Mittelpunkt des Handelns stellen.

Bernhard Keller

MaritzCX, Germany

Fragestellung / Problem:

Die Konzepte sind geschrieben, die Werkzeuge stehen bereit und der Wettbewerb ist hoch: CEM, EFM, CFM, VoC, CX – die Akronyme bezeichnen nichts weniger als einen Wandel in der unternehmerischen Ausrichtung. Was Tante Emma anno 1950 noch in ihrem Kopf speichern oder ein Vertriebsleiter 1980 in Excel niederlegen konnte, müssen Unternehmen heute über Cloud-Plattformen in Echtzeit stemmen: die 1:1 Kommunikation mit ihren Kunden. Denn heute geht es nicht mehr nur ums Messen von Kundenerfahrungen, sondern um das Management dieser Kundenerfahrungen. Und nur ein kundenerwartungsorientiertes Management garantiert den Erfolg. Dabei ist das nur die eine Seite – die andere umfasst die Auswirkungen der Managementmaßnahmen auf die Unternehmenskennzahlen: Kundenbindung, Cross Selling, Neukundengewinnung.

Doch sind die Unternehmen reif dazu? Wie können sie die internen Ansprüche und Barrieren und die externen Einstellungen Erwartungen zusammenbringen? Welchen Stellenwert haben die Kunden in diesem Prozess? Wollen diese überhaupt teilnehmen?

Methode zur Beantwortung der Frage:

Erfahrungen aus Workshops, Umsetzungen und Designs, also aus dem Alltag, kombiniert mit diversen Befragungen zum Thema Feedback.

Antwort auf die Fragestellung:

In vielen Unternehmen sind die Barrieren gegen die optimale Nutzung von Feedbacksystemen inklusive Analyse- und Simulationssoftware noch nicht abgebaut. Intransparenz und Nichtkommunikation über die Datenverwendung erschweren den Abbau abwehrender Meinungen auf Mitarbeiter- wie Kundenseite. Bedenken gegenüber individuellem Feedback können aber ausgeräumt werden, wenn Mitarbeiter- wie Kundenbedingungen proaktiv erfüllt werden. Dann wird ein perzipiertes Hindernis zum Motivator.

Die Feedbackbitte darf nicht nur als Maßnahme der Marktforschung (anonym) oder des Marketing (individuell) verstanden werden – sie muss als Kommunikationsmaßnahme auf Augenhöhe eingerichtet sein. Customer Integration erfordert, die Interessen der Kunden in den Fokus zu stellen. Und diese anschließend in Aktionen überzuführen. Das erfordert Mut auf Seiten der Unternehmen, sie müssen ihre Deckung hinter den Prozessen verlassen und ihr Gesicht bzw. das ihrer Mitarbeiter zeigen. Kunden im Gegenzug geben Informationen über sich preis, wenn Transparenz, Relevanz und Reaktion auf sie zugeschnitten sind.

Was geschehen wird, denn der Markt bewegt sich bereits in diese Richtung.
Keller-Customer Feedback Management hat zum Ziel, Unternehmenswerte zu stabilisieren-258.pdf

Die neuen Anforderungen an die Marktforschung oder warum sich die Branche ein Stück weit neu erfinden muss

Martin Grupe

formerly Vision Critical, Germany

More information will follow soon
Grupe-Die neuen Anforderungen an die Marktforschung oder warum sich die Branche ein Stück weit neu erfinden .pdf
10:00 - 10:15Coffee Break
10:15 - 11:15A 8: Innovative Topics in Web Surveys
Session Chair: Henner Förstel, MANUFACTS Research & Dialog GmbH
Room 248 

Teacher judgments in an online virtual classroom environment: the impact of gender-stereotyping on judgment validity

Markus Hörmann

PH Freiburg, Germany

While diagnostic competence is an important part of teachers' expertise, there is still an unsatisfying lack of research outcomes (e.g. Artelt & Gräsel, 2009). Previous results report deviations in the validity of teacher judgments in the classroom of about 20% from standardized assessments (Südkamp & Möller, 2009), a general central tendency (Fiedler, Walther, Freytag & Plessner, 2002) and a more accurate ranking of pupils’ performance compared to absolute measurements (Schrader & Helmke, 2002). Little is known about the reasons for inaccurate judgments as the personality trait of a student is a hardly controllable bias. This experimental online study prevents that bias by an online research paradigm. On the theoretical framework of dual-process models (e.g. Chen & Chaiken, 1999), this study investigates the impact of gender-stereotypes on the evaluation of the student’s aptitude for university studies, motivation and ability.

A web based, virtual classroom using PHP/AJAX-technology was developed based on an offline method (e.g. Fielder, Wöllert, Tauber, Hess, 2013). In this virtual, simulated classroom environment 59 female and 22 male participants interacted in the role of a teacher with three male and three female, virtual students who all had identical average ability (probability of correct answers) and average motivation (probability of raising hands). The participants asked predefined questions related to math or German. Based on the presented video answers, the participants gave evaluations regarding ability, motivation and aptitude of each of the virtual students. For measuring implicit gender-stereotypes as associative strengths, a web port of the Sorting-Paired-Feature-Task (SPF; Bar-Anan, Nosek & Vianello, 2009) was developed.

Significant gender differences following a traditional gender-stereotype regarding the aptitude for engineering (d-Cohen=0.44) and philology (d-Cohen=0.29) appeared. The independence of implicit and explicit gender-stereotypes was proofed.

Besides showing theory consistent results regarding implicit gender-stereotyping and judgments, an online tool to easily examine a large amount of research questions concerning educational psychology was developed. The virtual, online classroom was proofed to be fully suitable for online research. By porting the SPF another important assessment tool in cognitive research is now available to easily conduct online studies. Results of both developed instruments are comparable to offline alternatives.
Hörmann-Teacher judgments in an online virtual classroom environment-192.pdf

Using video to re-invigorate the open question

Pete Cape

SSI, United Kingdom

Relevance & Research Question: Open responses in online surveys are universally awful. Yet many people have a camera and microphone on the very device they’re using to take the survey, what happens if we ask them to video their response?

Methods & Data: An online research project using an “off the shelf” video capture and transcription service. Random assignment to video and non-video treatments.

Results: Our experiments reveal we get more response, more individual answers from each person and more depth, colour and emotion to work with. Judged on character count we obtain up to 6 times more data. These characters are additional substantive content as well as providing rich description and colour. Contrast: A video response: “Hi, I like gas because I like to cook with it, with a range top. It's fast, it's easy, you know it's on, you know what the temperature is based on how big the flame is. I like gas because it's quiet. I like gas because you don't smell it. I think it's cleaner than oil. I know it's less expensive than oil. I like gas, it's maintained better than having to call the oil company to come get it” A written response: “Clean, no smell, easy” Also a change in emphasis on the relative importance of issues; issues that seemed to only affect a minority in a written short answer become more important when people take time to speak their full mind. Data differences are substantial. E.g. “a reliable source” was mentioned by 8% of people who did not make a video, putting it 6th in rank. Almost 3 times as many people in the video group mentioned this (21%) making it a substantial, not a minority, answer.

Added Value: We don’t get videos from everybody so who chooses to answer? Our first experiment suggested that “extremists” want to speak out and so bias reduction is the topic for our second experiment. We conclude with some practical remarks and suggestions: How difficult is it to implement? How can you maximise response? How much does it cost, in Euros and in time?
Cape-Using video to re-invigorate the open question-122.pptx

New survey recruiting strategies: Online Panel vs. Mobile Adverting

Holger Lütters

HTW Berlin, Germany

Relevance & Research Question:

The idea of recruiting participants into online studies is about to radically change.

While the majority of the online market research industry is still inviting people over the email channel, some new approaches of using advertising networks to recruit people literally pop-up.

Besides the largely discussed approach of Google consumer surveys, using the Google advertising eco-system on desktop and mobile devices, a new and independent approach has been developed by a Berlin start-up company. This approach allows inviting unknown people over an advertising partner’s website to take a survey.

The market researcher’s question at this point of technology development is, whether this approach can be considered to create valid data compared to the classic panel approach. In order to find out about the opportunities and problems of this recruiting procedure the standard panel approach was used in competition with the advertising recruiting approach.

Methods & Data:

The comparison of the competing data collection approaches was supervised by independent academic researchers and supported by the commercial online panel GapFish and the advertising recruiting approach as developed by Dalia Research. Overall more than 3.000 people were contacted to create an online sample of more than 800 respondents for each approach equally covering desktop, laptop, tablet and smartphone devices allowing to compare the results.

The device agnostic questionnaire was dealing with respondent’s behaviour when answering questionnaires and their expectations in compensation for taking questionnaires in the classic and the mobile environment.


Overall this first experiment showed comparable results in the data of both approaches. The ad targeting approach can already be considered being an additional option for the researcher. The contribution will explain the effects of both approaches on the data by looking at:

• Different participation patterns on device categories, quota management and drop-outs

• Non-reactive data of the 8 minutes survey

• Reactive data of expectations of the participants in compensation for market research studies

Added Value: The new approach of advertising sampling will be explained during the presentation.
Lütters-New survey recruiting strategies-116.pdf
10:15 - 11:15B 8: Representation Online / Offline
Session Chair: Bernhard Keller, MaritzCX
Room 154 

Parliamentarians online – Members of the European Parliament and national parliamentarians in comparison

Jessica Kunert

Leuphana University Lüneburg, Germany

Relevance & Research Question: This paper compares how the Members of the European Parliament (MEPs) and their national counterparts (MNPs) present themselves on their personal websites. The focus of this paper lies on transparency measures (e.g. hyperlinks to social network sites, financial statements) and informational means (e.g. newsletter, speeches, explanation of parliament’s workings). With their personal websites, MEPs are not dependent on the channels of the mass media anymore, and can address the challenges of being of second-order importance to the voter and thus likened to the regional level (Reif/Schmitt 1980), as well as being subject to less media coverage than the MNPs (Franklin/Van der Eijk/Marsh 1996). Thus, this paper claims that MEPs are more engaged in using these online measures than their national counterparts, and address the European communication deficit this way (Anderson/McLeod 2004).

Methods & Data: The websites are analysed using quantitative content analysis, which was conducted with a standardised codebook for both MEPs and MNPs to ensure comparability. The dataset comprises 1818 German, British, Austrian and Irish MEPs and MNPs, out of which 1631 have a website. Aggregated indices for transparency and information measures are used for the analysis.

Results: The claim that MEPs show a higher amount of these measures than their national counterparts is confirmed. For the genuine online measures, this picture is uniform across nationalities. For the information measures, the picture varies widely inside and across nationalities.

Added value: I argue that MEPs utilize the communicative tools at their disposal differently from their national counterparts and should thus be recognized as individual parliamentary players with their own agendas when it comes to communication, and not just as an extended arm of their national parties or other European Union institutions. Moreover, personal websites are a prominent feature in campaigning and are widely researched during this timeframe, but their role in daily legislative life is only rarely acknowledged. This paper analyses the day-to-day presentation of MEPs and MNPs that often retreats into the background during campaigning, and allows an innovative analysis of the parliamentarians’ use of personal websites and their communicative offers.
Kunert-Parliamentarians online – Members of the European Parliament and national parliamentarians-138.pdf

The Influence of Urban vs. Rural Residence on the Use of Online Social Networks

Vlad Vasiliu

Haifa University, Israel

Relevance & Research Question: This study examines the relations of online and offline social circles and the extent to which they overlap. Focusing on the social network Facebook and two distinct offline social circles – the urban model and the model of the kibbutz (small collaborative community unique to Israel). The main contribution of this work is to search for the extent in which the offline and online social circles overlap and identify the possible impact of the two distinct offline social circles in relation to this overlapping.

Methods & Data: The research method used was an online survey questionnaire, which was distributed among 100 participants, half grew up in a Kibbutz and half in the city. The questionnaire was distributed via Snow Ball method due to the difficulty reaching kibbutz members, stemming from their scarce numbers.

Results: The findings demonstrated that the social circle on Facebook of people who grew up in a kibbutz manifest several properties that distinguish them from those of city dwellers. For kibbutz residents, the number of offline friends corresponds to the number of Facebook friends. A kibbutz resident's circle of friends is significantly smaller than an urban circle of friends, both offline and Facebook social circles. A person from a kibbutz interact offline with a significantly more Facebook friends than a person from the city. Moreover, a person from a kibbutz has more friends that he considers as close friends than a person from the city.

Added Value: These findings lead to the conclusion that the offline sphere and Facebook social circles of people who grew up in a kibbutz are more strongly connected than those of people who grew up in the city. This linkage demonstrates overlapping of traditional kibbutz social characteristics between offline and Facebook social circles, such as a large number of close friends and a small number of acquaintances. These findings may imply for the influence of offline community characteristics on online social circles.

Vasiliu-The Influence of Urban vs Rural Residence on the Use of Online Social Networks-211.pptx

Using Evidence from Vote Advice Applications to Estimate Issue Preferences of German Federal Electoral Districts

Dominic Nyhuis

University of Mannheim, Germany

Relevance & Research Question: Recent research on political attitudes has increasingly focused on the variation of preferences in small geographical units. However, traditional survey research is severely limited with regard to the estimation of small-area preferences due to data sparsity. To rectify this shortcoming, this contribution proposes a novel data source that exploits process-generated information based on online vote advice applications and asks: Are there preference variations between the German electoral districts? How are the preferences clustered?

Methods & Data: The evidence stems from a German vote advice application,, that provided users with information on the policy preferences of candidates during the 2013 federal election. All user inputs to the platform were recorded along with geographical markers, resulting in a data set of half a million users and their preferences on a range of policy proposals. Due to the self-selected sample, there are two imbalances in the data - geographical imbalances and an imbalance regarding the sociodemographic status of users. Therefore, users were invited to take an additional survey that collected various sociodemographic indicators. The user inputs are applied to estimate the policy preferences of the electoral districts using a hierarchical model with neighborhood clustering and poststratification in order to correct for the imbalances.

Results: There are substantial and systematic geographical preference clusters among the electoral districts. The estimated differences for the various policy proposals frequently cover a range of several dozen percentage points. The clusters also accord to a priori notions of the small-area preferences of the German electoral districts. The preferences even vary along historical border such that, e.g., issue preferences are more conservative in southern compared to northern Bavaria.

Added Value: This contribution provides a novel way for capturing small-area issue preferences. By incentivizing participation in the form of vote recommendations, it is possible to increase participation to numbers that are difficult to achieve otherwise. Moreover, the contribution highlights the value of process-generated evidence to capture relevant information for substantive research questions. Finally, the contribution features a technique to estimate small-area preferences by correcting for sparse and biased data sources.
Nyhuis-Using Evidence from Vote Advice Applications to Estimate Issue Preferences-119.pdf
10:15 - 11:15C 8: Social Media Research & Methodology
Session Chair: Brian Fabo, Centre For European Policy Studies
Room 149 

Development and Validation of a Scale for Social Exhibitionism on the Internet (SEXI)

Max Vetter1, Constanze Eib2, Sonja Hill-Kloß1, Philip Wollscheid1, Dirk Hagemann1

1Heidelberg University, Germany; 2Stockholm University, Sweden

Relevance & Research Question: The central aim of our research is to provide a reliable and valid scale to measure the degree of voluntary disclosure of private information (socially exhibitionistic behavior) on the internet and especially social media platforms. The relationship between self-disclosure on the internet and several measures of individual differences such as the Big Five, sensation seeking, depression, and social support is examined. We further examine the independence of online self-disclosure from offline self-disclosure. Correlations with objective measures of internet usage are also assessed.

Methods & Data: Three studies were conducted with heterogeneous samples of internet users recruited online and offline. Study 1 served to develop a reliable 15-item scale and an 8-item short version. This was followed by an extensive questionnaire validation study to investigate both discriminant and convergent validity (study 2) and a quasi-experimental study comprising extreme prototypes of socially exhibitionistic behavior (study 3).

Results: The findings strongly suggest that the new scale is an appropriate and reliable instrument for the measurement of socially exhibitionistic behavior in online environments. Moderate correlations with related constructs (e.g. extraversion, disinhibition, self-monitoring) were found while the independence of online self-disclosure from its offline variant was confirmed. Somewhat counter-intuitively, high self-disclosure on the internet did not correlate with depressive symptoms. Men scored higher than women on the newly developed scale.

Added Value: With the new scale, self-disclosure on the internet can be measured more reliably than with ad-hoc questionnaires. We provide first insights into the relationship of online self-disclosure and established measures of personality. We also make a number of suggestions for future avenues of research. For instance, we encourage further investigations concerning the causal direction of these relationships. The connection of self-disclosure with the hot topic of privacy on the internet also promises fruitful new opportunities for research. The newly developed scale of socially exhibitionistic behavior can contribute to this.

Hijacking the Hashtag: A Case Study of #BreakTheInternet on Twitter

Ognyan Seizov

University of Bremen, Germany

Relevance & Research Question: The "Discover" tab of the Twitter web and mobile application is an excellent way to gauge what topics are moving one's immediate and extended network at any given time. The most prominent way of identifying attachment to a certain topic remains the hashtag. However, as an annotation tool, the hashtag can (and is) misused for bandwagoning and displaying differently themed content in front of a sizeable audience which could not be reached otherwise. The present study asks: RQ1 - How often does hashtag hijacking occur? RQ2 - What alternative topics are displayed in "hash-jacked" tweets? RQ3 - How, if at all, do users react to "hash-jacking"?

Methods & Data: The study harvests all Top Tweets with the #BreakTheInternet hashtag at 10:30 (GMT+1) on November 14, 2014. The hashtag relates to the controversial photographs of Kim Kardashian for Paper Magazine. It then applies quantitative content analysis to the full sample, classifying tweets as on- or off-topic and assigning a sentiment value to them (positive, neutral, negative). The valence measure sheds light on the motivation of hash-jackers (i.e. mostly negative hashtags are less likely to be hijacked). The off-topic (i.e. hash-jacked) tweets are then analyzed qualitatively for verbal as well as visual rhetorical cues which reveal the poster's motivations for hijacking the hashtag.

Results: RQ1 - #BreakTheInternet was hijacked less than 10% in the Top Tweets sample. This can be attributed to the overall negativity (67%) of the associated posts. RQ2 - The most common hijacking scenario was a ploy to turn attention away from Kim Kardashian and towards the concurrent #CometLanding. A small sample of the hijackings were promotional posts for small local businesses trying to ride the popularity wave. RQ3 - Hijacking #BreakTheInternet did not initiate notable reactions or conversations.

Added Value: This mixed-method study delineates the hashtag-hijacking action in a high-profile Twitter event. It characterizes both the event's valence and the extent and nature of the hash-jacking action. It also sheds light on co-trending Twitter topics, as #BreakTheInternet and #CometLanding rivalled for supremacy on the microblog platform at the time.
Seizov-Hijacking the Hashtag-146.pptx

Social Map of Scientific Activities and Collaborations. A Network Study about Social Organizational Structures in Online and Offline Scientific Communities.

Cathleen M. Stuetzer1,2, Thomas Koehler1, Gerhard Thiem2

1TU Dresden, Germany; 2HS Mittweida, Germany

Relevance & Research Question:

The objective of this study is to analyze and visualize social behavior within scientific communities with the help of network analytics to detect social organizational structures in the online and offline context of scientists. The study focuses on answering following questions: (1) How do the social infrastructure of scientists look like? (2) What are the impact factors and gatekeepers’ strategies for evolving?

Methods & Data:

For exploration of social behavior within online and offline scientific communities the social network paradigm and key techniques as network analytics were used. Relational (quantitative) data about scientific behavior derived from activities of scientists in their scientific online and offline context. First, online data were extracted from the social media usage of scientists at an online research platform via tracking methods of integrated social media tools. Second, offline data were collected by listed project collaborations within a special research foundation between 2012 and 2013. The whole data set includes 119 scientists within 112 acquired projects within this research community.


The systematic analysis of structural data shows that during the project acquisition scientists keep in touch with each other to apply scientific projects and raise project funds together. The structural analysis demonstrates that social organizational infrastructures of scientists in both contexts based on the evolution of project elites in which scientists often co-occur and collaborate in different project teams. Approximately 80% of the scientists emerged within these project elites. About 10% of the scientists acquired isolated projects, operated independent from group processes, and acquired only a few projects over time. Scientists established in the project elite working in more than five projects. At the individual level the study shows that gatekeepers emerge online and offline. Thereby, central key actors differ less structurally more likely context-oriented at the individual level. They operates with different activity, intensity and diversity and emerge in the network. Furthermore, the study demonstrates evolution processes of social structures.

Added Value:

The results show that collaboration may be understood as strategy for success in the scientific business and evolution of elites and gatekeepers are features for functional scientific communities.

Stuetzer-Social Map of Scientific Activities and Collaborations A Network Study-208.pdf
10:15 - 11:15D 8: Social Media Research - Potenziale und Grenzen für die Marktforschung
Session Chair: Claas Lübbert,


  • Oliver Tabino (Geschäftsführer, linkfluence Deutschland)
  • Sabine Haas (Geschäftsführerin result) und Johannes Mirus (Teamleitung Forschung & Beratung)
  • Hans-Werner Klein (CIO Twenty54 Labs)
Room 158 

#socialmedia_lovers vs.#socialmedia_haters

Oliver Tabino

linkfluence Germany & Q Agentur für Forschung, Germany

Was die “Hater” sagen:

Auf einer Fachtagung zum Thema Social Media Forschung fasst der Moderator den Tag so zusammen: Ich bin froh, dass wir heute feststellen konnten, Social Media ist doch nur eine Ergänzung zu den konventionellen Methoden.

Oder Skeptiker entgegen immer wieder Folgendes:

Social Media ist nicht repräsentativ. - In den sozialen Medien wird nur gemotzt, die negativen Kommentare überwiegen. - Bei Twitter erhält man keine Insights, sondern eher Spam. – Kommt da überhaupt etwas dabei raus?

Was die „Lovers“ sagen:

Auf einer Marketing-Konferenz erzählt ein Head of Digital Marketing darüber, wie wichtig Echtzeit-Informationen für ihn sind, um Real Time Marketing Aktionen zeitnah zu evaluieren.

Social Media Forschung ist zwar nicht repräsentativ, aber das ist auch nicht der Anspruch. – Wer die digitalen Medien und die daraus gewonnen Daten außer Acht lässt, handelt verantwortungslos und hat offensichtlich die Zeichen der Zeit nicht erkannt. – Die Kritiker haben meist noch nie ein Social Media Projekt gemacht, wie kann man eine Methode dann schlecht reden?

Wir sind im verflixten 7 Jahr der Social Media Forschung. Zeit um mit Mythen, Legenden, Wunschdenken und Glaubenssätzen aufzuräumen.

Social Media Research – ein überschätzte Methode?

Johannes Mirus, Sabine Haas

result gmbh, Germany

Forschung ÜBER Social Media gibt es wie Sand am Meer. Wahrscheinlich hat schon jedes Institut das eine oder andere Mal eine Studie durchgeführt, in der Inhalte, Verhaltensweisen, Reaktionen, Interaktionen etc. von Social-Media-Kanälen untersucht wurden. Soziale Medien sind auch an wissenschaftlichen Einrichtungen ein beliebtes Themengebiet für Forschungen.

Fragt man im Kollegenkreis jedoch nach Forschung, die MITTELS Social Media durchgeführt werden, blickt man in stumme Gesichter. Soziale Medien eignen sich per se nicht für repräsentative Studien, die über die Nutzer dieser Medien hinausgehen. Quantitative Untersuchungen scheiden fast komplett aus. Qualitative Erhebungen sind in einem begrenzten Umfang denkbar, jedoch stellt sich auch hier die Frage zwischen Aufwand und methodischem Sinn.

Unstrittig bleibt Social Media als Möglichkeit zur Rekrutierung von potenziellen Befragten, wobei auch hier die Frage nach Aufwand und Repräsentativität nicht außer Acht gelassen werden darf. Auch für Sekundäranalysen, Marktbeobachtungen und Experimente kann Social Media ein interessantes Feld sein.

In einem 15-minütigen Vortrag möchten Sabine Haas, Gründerin und Geschäftsführerin result gmbh, und Johannes Mirus, Teamleitung Forschung & Beratung, Vor- und Nachteile von Social Media Research aufzeigen und Perspektiven darstellen.
Mirus-Social Media Research – ein überschätzte Methode-250.pdf

Kunden zu zu hören ist unser Geschäft!

Hans-Werner Klein

Twenty54Labs, Germany

Das Web ist voll von „nutzergeneriertem Content“ - ein großer Teil davon ist „Social Media“. Aber wer kann wirklich diesen wertvollen „Rohstoff Content“ aufbereiten und zu wertvollen Informationen veredeln?

Holen wir uns diese Aufgabe in unseren Beritt! Nutzen wir unsere Expertise und die vorhandenen Werkzeuge.

Qualitative wie quantitative Marktforschung ist im gesamten Lebenszyklus von Produkten und Dienstleistungen der größte Kanal zum Kunden. Gewesen.

Die vorletzte „Grosse Welle“ hat gar den Kunden als Controller im Bereich „Total Quality Management“ eingesetzt. Aber nur wenn er gefragt wurde.

Nun ist es laut geworden im Controlling: Kunden sind inzwischen gewohnt, zu Wort zu kommen. Ungefragt. Sie haben das Gefühl, gesehen und gehört zu werden. Nicht nur von Herstellern, sondern auch von anderen Kunden.

Die anderen Kunden und man selbst können als „Berater“ fungieren - verteilen fünf Sterne oder schreiben Produkte kaputt.

Auf Handelsplattformen wie Amazon, Bewertungsportalen wie Ciao, oder als „Stream of opinion“ in Twitter und Co. liegen Kundenmeinungen vor. In produktspezifischen Foren werden Erfahrungen ausgetauscht; Nutzer helfen Nutzern. Auf Facebook betreiben viele Unternehmen selbst eine Art Kundenportal.

War’s das dann für die Marktforscher? Ersetzen diese Kundenmeinungen in Echtzeit die Analyse der Kundenloyalität? Meiner Meinung nach nicht. Aber sie ergänzen unsere Analysen mit einer Unzahl an unstrukturierten Texten aus Nutzersicht. Liefern so wertvolle Inhalte.

Die Werkzeuge gibt es, um diese unstrukturierte Texte zu analysieren, und damit unsere Erfahrungen und Kenntnisse als Marktforscher zeitnah nutzen. Die neuesten Tools sind sehr einfach zu bedienen, setzen keinerlei „Anlernen“ des Tools voraus und funktionieren Sprachen übergreifend.

Also ran! Wer könnte besser als wir Marktforscher Informationen strukturieren, bündeln, bewerten, in einen Kontext stellen? Und vielleicht nebenbei eine 5-Sterne Mafia entlarven - und gekaufte Facebook Freunde des Wettbewerbs überführen.

Dies ist ein Aufruf, Kundenmeinungen auszuwerten, auch wenn wir sie diese nicht selbst gefragt haben.
11:15 - 11:45Coffee Break
11:45 - 12:30Keynote 2
Aula (Mainzer Str. 5) 

Quantifying human behaviour with Internet data

Suzy Moat

University of Warwick, United Kingdom

Our everyday usage of the Internet generates huge amounts of data on how humans collect and exchange information worldwide. In this talk, I will outline recent work in which we investigate whether data from sources such as Google, Wikipedia and Flickr can be used to measure and even predict real world human behaviour. I will provide case studies from the economic domain and beyond.

Suzy Moat is an Assistant Professor of Behavioural Science at Warwick Business School. In recent studies, in collaboration with Tobias Preis, H. Eugene Stanley and colleagues, Moat has provided evidence that patterns in searches for financial information on Wikipedia and Google may have offered clues to subsequent stock market moves, and that Internet users from countries with a higher per capita GDP are more likely to search for information about years in the future than years in the past.

Moat was awarded a Ph.D. from the University of Edinburgh and won a series of prizes during her studies. Since 2011, Moat has secured £3.3 million of funding from UK, EU and US research agencies. Her work has been featured by television, radio and press worldwide, including recent pieces on CNN and the BBC.

Moat has acted as an advisor to government and public bodies on the predictive capabilities of big data. She currently co-directs a data science research team working on these questions.

12:30 - 12:50Award Ceremony
Aula (Mainzer Str. 5) 
12:50 - 14:00Lunch
14:00 - 15:00A 10: Improving Questionnaires
Session Chair: Daniela Wetzelhütter, Johannes Kepler University
Room 248 

Coding Surveys on their Item Characteristics: Reliability Diagnostics

Frank Bais1, Barry Schouten1,2, Vera Toepoel1

1Utrecht University, Netherlands, The; 2Statistics Netherlands, Netherlands, The

Relevance & Research Question:

More and more surveys use multiple modes, which supplement or replace traditional interviewer modes by web. In multi-mode questionnaire design, usually some consideration is given to mode-specific measurement error. Despite this consideration, however, these measurement effects are frequently unexpectedly large and hamper publication. For this reason, there is a strong incentive to better predict measurement effects. Measurement effects are determined by the interplay between characteristics of the questionnaire and characteristics of the respondents. In our research, we investigate the existence and utility of so-called questionnaire and respondent profiles, in which these characteristics are summarized, for predicting measurement effects. As a first research question, we ask whether questionnaires can be coded reliably on item characteristics that are suggested in the literature as influential in mode-specific measurement effects.

Methods & Data:

We constructed a typology of item characteristics from the literature and applied it to a wide range of surveys; the Dutch Labour Force Survey of Statistics Netherlands and the core studies of the LISS panel of CentERdata. For all surveys, 16 item characteristics are coded by two main coders, while 7 of these 16 item characteristics that are assumed to be relatively influential in evoking measurement error are also coded by a third coder. Reliability diagnostics are derived for the various item characteristics.


Analyses of the survey coding scores indicate a relatively low reliability for characteristics that literature suggested as influential in mode-specific measurement effects: Sensitivity to social desirable answers, potential presumption of filter question, emotional charge, centrality, and language complexity.

Added Value:

It is investigated to what extent coding of questionnaires on its item characteristics is reliable and to what extent questionnaire profiles can be constructed based on this coding. Along with process data and register data that are linked to individual respondents who have filled out multiple questionnaires, the questionnaire and respondent profiles might shed light on the occurrence and scope of measurement effects for specific respondents and specific questionnaire characteristics over different survey modes.
Bais-Coding Surveys on their Item Characteristics-185.pdf

Approaches for Evaluating Online Survey Response Quality

Nils Glück1,2

1Cologne University of Applied Sciences, Germany; 2QuestBack GmbH, Germany

Relevance & Research Question: Online questionnaires are a common utility for research companies. While survey software products offer many features, respondent fraud and indifferent or inattentive respondent behaviour remain critical issues. How can responses with such bad quality be identified in an automated process?

Methods & Data: The author proposes a post-fieldwork approach which is based on behaviour pattern detection and does not rely on control or trap questions. Using response quality indicators as well as discriminant analysis, logistic regression and an optional flag variable, responses are classified with regard to their quality. The 17 indicators focus on aspects such as response differentiation to open-ended questions, the time spent for answering the survey and monotonous behaviour in response to matrix questions. For the procedure to work, the survey should include open-ended questions, several matrix questions as well as a minimum of ten questions overall. An incentivized survey containing quality-related trap questions and other control measures is sent out to a Facebook river sample (n = 134) as well as a commercial panel sample (n = 1,000). This survey is used to generate a standard classification. Another five survey data sets from past real-case projects are finally used to examine the effectiveness of the procedure developed (157 <= n <= 2,603). R is used for calculating indicators, SPSS for discriminant and regression analysis. The automation process is specifically designed for QuestBack EFS software.

Results: Depending on the data, the procedure identifies between 2.5 and 5.2 per cent of all respondents as low-quality respondents. Judging from their indicator values, their behaviour is clearly suspected to indicate bad quality. Therefore they should be considered to be removed from the sample.

Added Value: The approach offers straight-forward ways of judging whether survey responses should be considered trustworthy in comparison to one another. This knowledge supports post-fieldwork data cleansing and reduces effects of distortion by low-quality data. The procedure is ready for implementation in the EFS software.
Glück-Approaches for Evaluating Online Survey Response Quality-193.pdf

Deep impact or no impact, evaluating opportunities for a new question type: Statement allocation on importance-performance-grid

Sebastian Schmidt

SKOPOS GmbH & Co. KG, Germany

Relevance & Research Question:

Standardized grid question types in online questionnaires can be regarded as the backbone of modern quantitative research. Grid questions allow comparability among different survey waves. Furthermore, it is well proven that varying the way by which grid questions are displayed strongly affects response behavior. Despite its importance in everyday research, this type of questions is lacking the ability of taking full advantage from the opportunities the web currently offers, such as using media elements to express opinions or visualizing certain aspects e.g. by using pictures.

Against this background, the authors will explore an innovative approach of the importance-performance-analysis (IPA) whereby respondents are able to position particular customer satisfaction aspects directly on a grid which is divided into four so called action areas: “concentrate here”, “keep up the good work”, “low priority” and “possible overkill”. We will assess whether this approach allows identifying critical performance factors as already known from the traditional IPA. Benefits of this approach could be increasing respondent engagement, time savings and a more distinct prioritization of aspects. Pitfalls like a lack of comprehensibility and satisfaction patterns need to be considered to assess future usage of this approach.

Methods & Data:

The authors conducted a customer satisfaction survey via an Online-Access-Panel, applying a split-half-design to examine effects of this new grid replacing question type. Respondents were assigned at random to rate their customer satisfaction statements, either using the traditional Importance-Performance-Analysis (IPA), or the new approach explained above, whereby statements are directly assigned to the 4 field grid. A comparison of both designs will reveal differences among satisfaction and importance ratings. Analysis will also assess the comprehensibility of the grid approach while differences of the interview duration and respondent engagement are being examined as well.


Available by the end of January.

Added Value:

The authors will show to what extent a direct assignment of the customer satisfaction statements on the IPA grid can be considered as comprehensible and valid for the purpose of analyzing effects of customer satisfaction, highlighting possible advantages and potential risks, ultimately concluding whether future usage is reasonable and beneficial.

Schmidt-Deep impact or no impact, evaluating opportunities-205.pdf
14:00 - 15:00B 10: Bridging Generations
Session Chair: Birgit U. Stetina, Sigmund Freud University
Room 154 

Understanding Differentiated Internet Skills among the Elderly

Eszter Hargittai1, Merrie Ringel Morris2, Anne Marie Piper1

1Northwestern University, United States of America; 2Microsoft Research, United States of America

Relevance & Research Question: Many studies of Internet users focus on young adults ignoring information about older populations. Even in national samples, users above a certain age (e.g., 65) tend to be grouped together thereby assuming that the elderly are homogenous when it comes to their online behavior. Research on digital inequality has found across different national contexts that those from more privileged backgrounds are not only more likely to be Internet users, but also tend to be more skilled at using the Web. Such research rarely contains data on the elderly, however. As people from diverse backgrounds live longer, it may be that differentiated skills exist even among older age groups. This paper looks at whether there are Internet skill differences among older adults.

Methods & Data: We collected survey data from an online panel about 505 Internet users ages 60-80 in the United States focusing on their Internet skills.

Results: We find that the more highly educated and those with higher incomes exhibit higher skills. Once we control for autonomy of use, or the freedom to use digital media when and where one wants, we no longer see a relationship between Internet skills and income. This suggests that financial circumstances are influencing online abilities at the level of access even among older adults who are all Internet users. Those with the lowest levels of education continue to be less skilled regardless of user autonomy. Countering findings for other age groups, we find no gender difference in skills among the elderly. We also find that among a group of older adults, age still matters. Those in their early 60s are more skilled than those in their later 60s who are more skilled than those in their 70s.

Added Value: Our unique data about a diverse group of older adults’ Internet skills contributes to digital inequality literature on a population that tends to be ignored by most existing studies. We know of no other study that focuses on older adults’ Internet skills.

Parent-Child Connections on Facebook and Cyberbullying

Gustavo S. Mesch

University of Haifa, Israel

a) Relevance & Research Question:One understudied area of research is the extent that the presence of a parent in the SNS children’s friend list might serve as a protective factor from online negative experiences and cyberbullying online. This study fills a gap in the literature and investigates the association of parent-child connection on SNS on teens’ participation in online risk activities and exposure to online negative experiences and cyber-bullying.

Methods & Data: The 2011 Teens and Digital Citizenship Survey sponsored by the Pew Research Center’s Internet and American Life Project obtained telephone interviews with a nationally representative sample of 799 adolescents aged 12 to 17 years old and their parents living in the continental United States (Pew and American Life, 2011). A secondary analysis of this survey was conducted.

Results: We found that consistent with the argument of routine activity theory, participation in online risk activities is associated with the exposure to negative experiences in SNS and exposure to cyber-bullying. Controlling for the children's participation in risky online activities and socio-demographic factors, children reporting having a parent as a Facebook friend reported fewer negative experiences on social networking sites and were less likely to be victims of cyberbullying.

d) Added Value: The protective effect of the parent-child connection on Facebook is explained mainly by a reduction in the number of social networking site activities, not by a reduction in the children's participation in risky activities online. Furthermore, the parent-child connection on Facebook apparently has a specific protective effect might result from child voluntary disclosure of information to parents.

Mesch-Parent-Child Connections on Facebook and Cyberbullying-110.pptx
14:00 - 15:00C 10: Corporate Social Media
Session Chair: Anna Schneider, YouGov Deutschland AG
Room 149 

Social media and co-creation with the crowd – Success factors of idea contests in the consumer goods industry

Fabian A. Geise

Unitymedia KabelBW GmbH, Germany

Relevance & Research Question: The expansion of the internet offers companies the chance to realize open innovation strategies in collaboration with consumers. Crowdsourcing strategies can be carried out e.g. on intermediary innovation platforms (innocentive, atizo etc.) or on company-owned platforms (tchibo-ideas, mystarbucksidea etc.). While intermediary and company-owned platforms are focused on a constant generation of ideas from the users, another development has been seen over the last few years, especially in the consumer goods industry: so-called idea contests. This kind of web-based collaboration with consumers offers a further option for co-creation to generate ideas for new product development. Idea contests in the consumer goods industry are temporary projects/campaigns, which are carried out on a company-owned platform including social media (especially Facebook). The question arises in this context: how widespread are social-media based idea contests in the consumer goods industry and which success factors can be derived from idea contests that have already been realized.

Methods & Data: A content analysis of idea contest websites and Facebook sites of a dozen companies (McDonald’s, Beck’s etc.) was conducted. In addition open innovation literature and crowdsourcing blogs were analyzed.

Results: The results show the typical attributes of idea contests (scope of the task, types of incentives, idea generation process, tools for idea generation etc.). As success factors, three main factors with some underlying single success factors were discovered: (1) task content (two underlying single success factors: relevance of the idea generating task, simplicity of the idea generating task), (2) user orientation (two single success factors: definition of the right crowd, incentives to encourage participation) and (3) interaction competence (four single success factors: organizational structures and strategies to support interaction, credibility, transparency, fairness).

Added Value: Although there is a variety of studies with open innovation and co-creation with consumers as topics, no detailed study about social-media based crowdsourcing in the form of idea contests with special focus on success factors is available. The success factors show how to implement a successful social-media based crowdsourcing project/campaign with the aim to generate ideas for new product development in the field of consumer goods.

Social media as a driver of brand transparency: the effects of proactive disclosure on consumer behavior

Joris Demmers, Anne Sophie Erbé, Jet van Strijp, Claire Wientjes

University of Amsterdam, The Netherlands

Relevance: In today’s digital era marketers no longer control the flow of information available to consumers, who increasingly rely on fellow consumers. As a results, companies, being used to presenting their products in a predominantly positive light, are pressured to be more transparent and proactively disclose both more and more balanced information. Indeed, surveys show that consumers prominently consider a brand’s level of transparency when making purchase decisions. However, disclosure also poses risks of exposure when information is not unambiguously positive. In a series of studies, we investigate how brands’ proactive disclosure of negative information through social media affects consumer perceptions and decision-making.

Method: We conducted two online auctions, one social media experiment, and a field study. For the auctions and the experiment, we exposed 896 participants to different conditions regarding information availability (present versus absent) and source (brand versus consumer review). For the field study, we asked 85 supermarket shoppers to taste one of two displayed products after having presented them with moderately negative information disclosed by varying sources.

Results: The auctions showed that people were willing to pay more when negative information was disclosed by the brand, both in comparison with a condition where the same information was disclosed in a consumer review and a condition in which the information was not disclosed. The experiment showed that people perceived information as both more positive and relevant when it was disclosed by the brand (versus a consumer review). Finally, in the field study people were more likely to choose the focal product when negative information was disclosed by the brand. This effect was mediated by perceived information valence and relevance, hence connecting results from the previous studies.

Added Value: This study demonstrates that proactive disclosure of relevant information positively affects consumer behavior, even when the information is negative. Moreover, the results show that information is more influential when it is proactively shared by brands, and offer insights into the mechanism underlying this counterintuitive effect. To companies, this study shows that transparency is not just an inevitable consequence of consumer empowerment in the digital era, but poses interesting new opportunities.
14:00 - 15:00D 10: Die Sicht der betrieblichen Marktforschung: Online, Mobile und Co. im Methodenmix
Session Chair: Marco Ottawa, Telekom Deutschland GmbH


  • Dr. Ik-Sun Kim-Backmann (Senior Expert Customer Insight, Market Research & Competitive Analysis bei & Audience Products, Deutsche Telekom AG)
  • Kerstin Tober (Projektleiterin Marktforschung, Deutsche Bahn AG)
  • Ricardo Rubio González (Head of Research New Media, SevenOne Media)
Room 158 

Online Surveys und (sozio)demographischer Wandel: Erste Ergebisse auf der Basis von Beta Labs

Ik-Sun Kim-Backmann

Deutsche Telekom AG, Germany


Wie kann das Alterssegment 50+ in Deutschland vor dem Hintergrund des in den Industriestaaten stattfindenden (sozio-)demographischen Wandels am besten befragt werden? Im Jahre 2012 lag das Medianalter in Deutschland bei 44,2 Jahren bei Männern (arithmetisches Mittel: 42,8), bei Frauen bei 46,5 Jahren (arithmetisches Mittel: 45,5). Das Medianalter teilt die Bevölkerung nach dem Alter in zwei gleichgroße Gruppen: 50 % sind jünger und 50 % sind älter als das Medianalter. Im Ergebnis der steigenden Lebenserwartung und des niedrigen Geburtenniveaus ist das Medianalter kontinuierlich angestiegen, dieser Trend wird sich weiter fortsetzen.

Forschungsmethode und Daten:

Das Portal der Deutschen Telekom AG hat hierzu ein Online Access Panel mit dem Namen Beta Labs aufgesetzt. Die Panelisten bestehen aus den Es befindet sich momentan im Aufbau mit ca. 2.000 Teilnehmern, von denen über 60% über 50 Jahre alt sind. Alle Teilnehmer haben an einer umfangreichen Stammdatenbefragung teilgenommen. Die Ergebnisse basieren auf dieser Initialbefragung.


Das Segment 50+ im Panel ist am besten über die Screens Desktop und Laptop erreichbar, wobei 70% über ein Smartphone, 50% über ein Tablet und 32% über ein Smart-TV verfügen. Somit werden diese Endgeräte auch diejenigen jenseits des Medianalters für Befragungen immer relevanter.

Added Value:

Ein Online Access Panel ist für eine effiziente Erhebung im Sinne der Forschungsökonomie sinnvoll. Nun kann dieses Instrument auch für das Segment 50+ erfolgversprechend eingesetzt werden, wobei die Qualität weiter verbessert wird.
Kim-Backmann-Online Surveys und (sozio)demographischer Wandel-248.pdf

Von PaPi über CATI zu Online und Mobile - die Geschichte der Kundenzufriedenheitsforschung im Schienenpersonenverkehr bei der Deutschen Bahn

Kerstin Tober

Deutsche Bahn AG, Germany

Der drängende Wunsch nach immer schnelleren Ergebnisse führt dazu auch die Befragungen schneller durchzuführen, Feldzeiten zu verkürzen.

Eine Möglichkeit diesen Herausforderungen zu begegnen, sind neben klassischen Trackingstudien an Bahnhöfen oder im Regionalverkehr z.B. QR Code Befragungen in Zügen und Bussen der Deutschen Bahn.

Moderne Methode verknüpft mit einem intelligenten Reportingtool liefern fast schon Echtzeitdaten. Aber wie sieht es mit der Stichprobenqualität aus? Wer nimmt an diesen Befragungen überhaupt teil? Wie kann man trägere Kundengruppen von einer Teilnahme überzeugen? Und wie geht man mit nichtrepräsentativen Daten um? Erweiterte Zeitreihenanalysen ermöglichen Stichprobenfehler zu erkennen und helfen betriebliche Leistungen zu optimieren.

Vom Forschungsgegenstand zum Forschungsinstrument – Wie lassen sich Smartphones sinnvoll und effektiv in der betrieblichen Marktforschung einsetzen?

Ricardo Rubio González

SevenOne Media GmbH, Germany

Vier von fünf Onliner besitzen mittlerweile mindestens ein Smartphone. Neben zur Kommunikation werden die mobilen Alleskönner vor allem zur Information und zum Entertainment genutzt. Und sie haben sich zum beliebten Shoppingassistenten entwickelt: Bereits mehr als die Hälfte der Smartphone-Besitzer verwendet das Gerät zur Planung von Einkäufen und während des Einkaufs im Geschäft, zum Beispiel um sich über Produktalternativen zu informieren, Preise zu recherchieren oder Freunde und Bekannte nach ihrer Meinung zu einem Produkt zu fragen (Vlg. Media Activity Guide Mobile 2014 / SevenOne Media).

Dadurch ergeben sich neue Forschungsszenarien und das Smartphone entwickelt sich vom Forschungsgegenstand zum Erhebungsinstrument. Event- und Local-Based-Umfragen können dabei die Lücke zwischen Befragungen und Trackingdaten sinnvoll schließen. Welche Implikationen ergeben sich dadurch für die Durchführung von Umfragen am POS, für die Datenqualität und nicht zuletzt für den Datenschutz? Das sind Fragestellungen für die wir uns aus der Sicht der betrieblichen Marktforschung Impulse und Antworten von Seiten der Universitären- und Instituts-Markforschung erhoffen.
Rubio González-Vom Forschungsgegenstand zum Forschungsinstrument – Wie lassen sich Smartphones sinnvoll-256.pptx
15:00 - 15:15Coffee Break
15:15 - 16:15A 11: Respondent Behaviour and Data Quality
Session Chair: Stephan Groll, MSR Consulting Group GmbH
Room 248 

Media Diary Reloaded: Deep Dive into Multiscreen Behavior

Diana Livadic, Alexander Bohn

Ipsos GmbH, Germany

The digitalization massively changed the way we use media, content and devices. As consumers we are able to use, interact with and share digital content via various new devices every time and everywhere. Research can utilize these devices and modern lifestyle of “always on” to get closer to consumers. We used this trend to gather representative data about the current media and screen usage in German TV-households. Main research objective was to define the role of TV in our multiscreen world.

A mixed method approach combining traditional face-to-face interviews with an innovative mobile media diary app was applied: The representative media and device usage and consumers’ attitudes were measured via CAPI (n=1500 respondents). During the interviews respondents using other screen based devices in parallel to TV (multiscreeners) qualified for the mobile media diary. The diary captured media usage in the moment of experience: n=72 respondents documented their media and device usage every two hours for two consecutive days. They provided information regarding type of media and device, place and time of usage, parallel media usage and pictures of usage situations. Data collection was distributed equally to cover one whole week. Data of both collection methods was then combined and holistically analyzed.

The research enables us to gain a deep and precise understanding of current media and multiscreen usage as well as the role of each device. Furthermore complex usage situations could be investigated and valuable insights discovered, enriched by pictures taken by the respondents: Screen usage strongly depends on context. There are different peaks for each device used alone or in parallel with other screens during the day.

Instead of a classical paper and pencil media diary or an online diary that is completed only once a day for past usage situations and experiences the data collection via app perfectly fits the lifestyle and behavior of multiscreeners. Using apps to share and document experiences is part of their natural behavior. The mobile diary app takes advantage of this and makes research much more attractive, engaging and also reduces a potential bias of recall.

Livadic-Media Diary Reloaded-195.pdf

Deficient data quality as a consequence of respondents’ mobility? Response strategies according to mobility in terms of smart-phone usage and location

Daniela Wetzelhütter

Johannes Kepler University, Austria

Relevance & Research Questions: The internet has become an integral part of our lives and takes on an increasingly important role in survey research. The strength of this process is, as Dillman (2007: 400) puts it in a nutshell: “there is no other method of collecting survey data that offers so much potential for so little cost as Web surveys”. Despite and/or because of this progress, at least three related issues of online-data collection (apart from problems with sampling frames) are (still) challenging: i) response rates, ii) motivation and iii) mobility of respondents. The presentation aims to explore their relationship and its consequences for data quality. The main research question is: Does mobility decrease the participants’ motivation (to respond) and subsequently compromise data quality? For answering this question, the theoretical framework of “satisfizing” is applied to measure motivation, taking theoretical considerations about mobility into account.

Methods & Data: The analyses are based on an online-survey dealing with student participation at the university. Motivation strategies in terms of experimental arrangements were included to facilitate the identification of ”satisfizing”. The survey population was randomly divided into groups differentiated by the presence of specific “triggers”. 14.2% of the students (n=17,491) reacted to the invitation, 1916 (11%) answered at least one question but just 7.3% (n=1282) reached the final page.

Results: Mobility affects response rates in terms of significantly higher drop-out rates of mobile respondents. Unexpectedly, the percentage of item non-responses does not differ between mobile and non-mobile respondents, when drop-outs are excluded. However, drop-outs do not significantly influence sample composition concerning study related variables and demographic data. But, the respondents’ involvement in issues related to the survey-topic reduces drop-outs as well as satisfizing strategies does. Both variants have consequences on substantive results, whereas the latter is expected to be connected to mobility.

Added Value: The paper aims to emphasize the importance of record the setting of data collection and satisfizing strategies in case of self-administered online-surveys as it has consequences which are otherwise difficult to verify.

Literature: Dillman, D.A. (2007). Mail and Internet Surveys. The Tailored Design Method. New York: Wiley.
Wetzelhütter-Deficient data quality as a consequence of respondents’ mobility Response strategies ac.pdf

Online Eye-Tracking of Dynamic Advertising Content in (Mobile) Web-Surveys

Sebastian Berger

Dr. Berger Consult, Austria

Relevance & Research Question: Online eye-tracking by use of the patented spotlight-viewer methodology is an academically renowned (e.g. Psychology & Marketing, Vol. 29, Issue 6) and in applied research well established (in Germany e.g. mo’web research, MW Research, Interrogare) way to measure visual attention in online-surveys. However, so far its use had been limited to static advertising material. Due to the popularity of dynamic content in digital advertising the goal of the current research was to also allow rich media and video content to be tested with the spotlight-viewer in (mobile) web-surveys.

Methods & Data: Two main challenges had to be overcome in order to satisfy this demand. First, the data volume had to be minimised to keep download time for study participants within an acceptable limit. Second, it had to be ensured that independent of desktop or mobile device, browser type and internet speed all participants are confronted with the identical visual content at each point of time. Only then data could be jointly analysed. Following this, a video is captured of the screen content and frame rate is reduced if necessary. The (remaining) video frames are then fed into the spotlight-viewer and presented to the participants while simultaneously recording their viewing behaviour.

Results: On the basis of the first empirical results generated with the enhanced spotlight-viewer it is presented in how far, when and which rich-media advertising drives attention of Internet users. As a result the significance of pre-attentional bottom-up processes as well as parafoveal and peripheral vision in digital advertising research is discussed.

Added Value: The added value of the presented enhancement is that all types of dynamic advertising content can be tested on any desktop or mobile internet device concerning the ability to attract and hold attention and this (unlike infrared eye-tracking) in a simple, unobtrusive, low-cost, and time-efficient way. This allows to make an effective choice on which digital ads to run and with what weight in overall ad rotation.

Berger-Online Eye-Tracking of Dynamic Advertising Content in-204.pdf
15:15 - 16:15B 11: Global ICT Challenges
Session Chair: Olaf Wenzel, IPF Wenzel
Room 154 

Online health information and use of online health services: A test of the digital divide paradigm among the chronically ill in Israel

Rita Mano

University of Haifa, Israel

Relevance & Research Question: Studies on the use of online information and the social diversification hypothesis (Mesch et al., 2012) provide ample evidence about the consequences of lack/low use of exposure to online information has for individuals. The study examines the relationship between (a) access and use of online health information and (b) use of online health services and focus on individuals with a chronic illness. We combine internet use theories and the digital divide paradigm with health empowerment hypotheses. We ask how the relationship between access of online health information and use of online health services materializes among individuals with a diagnosed chronic illness relatively to those who are in good health.

Methods & Data: A telephone survey was conducted in 2010 in Israel among 4000 individuals. 2008 individuals completed interviews (54% response rate). 70% were Internet users (n = 1371).

Results: We show that for most individuals suffering from a chronic illness online health information is a viable way for health empowerment similarly to non-ill individuals but not regarding use of online health services. Distinguishing between type of online health services is important too. Hypertension decreases the use of online health services and diabetes has no effect whatsoever. A heart condition increases the use of online services for updates but not for making appointments.

Added Value: Access and use of online health services is a quickly developing area of health prevention and health promotion in affluent societies. The situation where individuals with chronic illness do access online health information and yet do not use online health services raises questions about the contribution of chronic illness to social disadvantage that has been seldom approached as such in existing literature. Chronic illness carries the characteristics of “second level” digital divide effect since a wide range of health services is now delivered online and health insurance firms presume that this type of services will increase over time raising the possibility that correlates of chronic illness, e.g. age and education are important in designing online health services

Mano-Online health information and use of online health services-128.pptx
15:15 - 16:15C 11
15:15 - 16:15D 11: Berufsbilder Marktforschung 2020 - Sozialforscher, Marketingberater oder Big-Data-Analyst?
Session Chair: Matthias Fank, Cologne University of Applied Sciences


  • Marco Ottawa (Marktforschung Telekom Deutschland)
  • Tilman Rotberg (Global Lead Technology; Technology Key Account Manager Group, GfK)
  • Henner Förstel (Geschäftsführer Manufacts)
  • Patrizia Trolese (Inhaberin pt profession)
Room 158 

Contact and Legal Notice · Contact Address:
Conference: GOR 15
Conference Software - ConfTool Pro 2.6.76
© 2001 - 2014 by H. Weinreich, Hamburg, Germany