A Theoretical Model for Trust in Digital Information Systems
University of Münster, Germany
Relevance & Research Question: Trust has been an important topic in technology research and adoption of web-based services. Users’ trust is not only essential for acceptance of new systems but also for a continued adoption. As a first step of a systematic approach to design trusted information systems and enable trustful interaction between users and information systems, we develop a new comprehensive model of trust in information systems based on existing literature. This models includes both precursors (e.g., perceived trustworthiness) and consequences of trust in IS (e.g., actual use, forgetting of information stored in the IS for cognitive relief).
Methods & Data: Based on extant literature, we differentiate experienced trust in IS from perceived trustworthiness of an IS, intentions to use an IS, and actual usage of an IS. Moreover, exiting experiences with an IS as well as perceived risk and more general contextual factors are considered as moderating factors. Finally, the resulting model is reflected and refined in light of empirical findings on precursors and outcomes of trust in IS.
Results: Our new comprehensive model not only structures the existing research, but also provides a number of concrete propositions. Perceived IS trustworthiness should affect trust particularly when users already have experience with the IS. Experienced trust in IS should strengthen users’ intention to adopt an IS, but this effect should be counteracted by perceived contextual risks. Trust should be influenced by the users’ personality, e.g., their general disposition to trust. The translation of users’ intentions to use an IS into actual behavior should be affected by enabling (e.g., control) and inhibiting (e.g., existing work routines) contextual factors. Finally, the model contains feedback loops describing how actual use of an IS influences perceived trustworthiness of an IS, and following trust experience of users. Evidence from two experimental studies supports the validity of our model.
Added Value: We provide a new theoretical model that outlines a structured process chain of trust in IS. In addition to the theoretical contribution, this systematic approach provides various practical implications for the design of IS as well as for the implementation and maintenance process.
Factors Influencing the Perception of Relevant Competencies in the Digitalized Working World
Bielefeld University of Applied Sciences, Germany
Relevance & Research Question: Many companies see the lack of skilled workers as a central obstacle to the digital transformation. It is well-known that diverse workforces lead to more balanced decisions and more innovation. Nevertheless, women, for example, are still underrepresented in STEM-professions. The following research question arises: Are there any differences in the perception of relevant competencies for the digitalized working world according to gender, age, employment status and migration background?
Methods & Data: Following preliminary literature research and qualitative expert interviews [n=6], a quantitative study was conducted from Nov. – Dec. 2018. Participants [n=515] were recruited among students and companies using faculty email lists, paper form and social media. Participants were asked to assess a total of 14 competencies, knowledge resources and behaviours in their relevance for the digitalized working world on a 6-tier scale. Correlations were determined by calculating Chi-square according to Person and Cramer’s V. Means were compared using T-Test and Levene.
Results: Respondents consider openness to change (5.50), IT and media skills (5.40) and learning ability (5.36) to be the most relevant. Analytical skills (4.79) and empirical knowledge (4.56) are less in demand.
Men rate innovation competence (χ²=10.895, p=.028, V=0.146), decision-making ability (χ²=13.801, p=.017, V=0.164) and ability to think in context (χ²=14.228, p=.014, V=0.167) slightly higher than women. No correlation can be found regarding respondents’ migration background. Among company representatives, eight competencies are rated significantly higher than by students, especially communicative competence (+0.91) and interdisciplinary thinking and acting (+0.74). Moreover, it is noticeable that older participants (generation X, born 1964-1979) consider all competencies to be more important than younger ones (generation Z, 1996-2009), apart from IT and media competence. The items openness to change (T-Test p=.004, Levene p=.004), self-organisation (T-Test p<.001, Levene p=.020) and problem-solving competence (T-Test p=.011, Levene p=.019) show significant correlation between age and assessment.
Added Value: First, results reveal a ranking of needed competencies for the digital transition, which companies and educational institutions should address. Second, differences between the employee groups could be discovered which have to be considered in the further approach, be it education or research.
How to regionalize survey data with microgeographic data
Infas360 GmbH, Germany
Relevance & Research Question: Using an online access panel of 10,000 households to precise target groups and their area-wide and microgeographical prediction for all regions and addresses
Methods & Data: The basis for the estimation is an online survey with 10,000 participants for which we have postal addresses. The addresses are crucial to match the microgeographic data from infas 360. The process of matching is called geocoding. Geocoding validates and locates addresses and integrates the individual geo key to which the microgeographical data are also attached. This database contains nearly 400 microgeographic variables to each of the 22 million unique addresses, e.g. building typology, building use, number of private households, average monthly net income, number of foreigners per block, unemployment rate per city district. Through the combination of the survey data and the microgeographic data, we localize, describe and analyze in a first step the responding households geographically, socio-demographically and according to other relevant characteristics. In a second step, we estimate the target group at 40 million households in Germany using a multi-level model. We demonstrate the procedure by means of an example, which is the prediction of the number of dogs per city district.
Results: Among other things the analysis in this example shows that dog owners live in houses with large gardens and are located mostly at the town border. Usually they live in regions with a high purchasing power and with a low migration share. This information is used to predict the number of dogs for all addresses with private households. Besides the quality criteria for the estimation model, we compare the results with official data from the Statistical Office Berlin on the level of city district. The comparison shows that the estimated data reflect the official data very well.
Added Value: Linking the online survey data with micro-geographical data from infas 360 enables a comprehensive description and segmentation of target groups as well as their address-specific transfer to the overall market as a prediction with astonishing precision. The use of an online access panel is cost efficient, flexible in time – and it is regionalizable.
"Like me": The impact of following prime ministerial candidates on social networks on perceived public agendas
Academic College of Emek Yezreel, Israel
Relevance & Research Question: Agenda-Setting research has been performed for more than four decades now as a matter of routine, in both traditional and online media. Recent years have seen a growing political use of social media messaging, especially during elections campaigns. The current study analyzes the effects of prime ministerial candidates' online messages on their followers' perceived agenda, as a function of the following patterns and the voting intentions, on which these followers have reported.
Methods & Data: To answer these questions, a representative sample of Israeli voters (n=1600) have answered a detailed Questionnaire. Questions regarded voting intentions, patterns of following prime ministerial candidates' accounts on social networking sites, and the followers perceived agenda.
Results: 43% reported that they follow a social network account of at least one prime minister candidate. 88% of them follow candidates through Facebook, 19% follow through Twitter and 18% through Instagram. The prominence of issues on the agenda differed significantly between those who reported that they followed candidates to those who did not follow them. Among the voters of the ruling party, 80% followed the party candidate exclusively, and 18% follow the party candidate in addition to the candidate of the leading opposition party. Among leading opposition party's voters, 48% follow the opposition candidate exclusively, and 47% also followed the ruling party candidate. The prominence of the issues on the agenda differed significantly between the exclusive followers of each of the two candidates.
Added Value:The present study indicates that social networking sites have a substantial impact on followers' perceived agenda during election campaigns. As social networks turn prominent than ever, it has become highly essential to deepen our understanding of their unique role in the political arena.
Social-media based research: the influence of motivation and satisficing on empirical results
1University of Applied Sciences Upper Austria, Austria; 2Johannes Kepler University Linz, Austria
Relevance & Research Question: Frequently business decisions are relying on data collected through self-administered web surveys. The arguments derived from such datasets are based on non-probability convenience samples, provided by a self-selection of social media users completing online questionnaires. This kind of research is focusing on informative samples. Troubles caused by a lack of possibilities to test the representability are pushed into the background. However, projects rooted in such approaches have the potential to further increase undesirable respondent behaviors that include but may not be limited to speeding or straight lining. Not least because social media usage is connected to the users motivation, as it applies for survey participants. Based on this the present paper addresses two research questions: i) whether and to what extend does motivation (in connection with satisficing) influence the results of social-media surveys? ii) Are different effect detectable, caused by different accesses to the field?
Methods & Data: Analyses are based on three different data sets of social-media surveys. The first consists of 104 users of online forums for “gamers”, the second of 234 Facebook-users, which followed an invitation to evaluate the Facebook dialog with public utility companies and the third dealt with politics and was composed from a quota sample based on the Facebook network of students. In order to measure motivation and satisficing, individual indicators of motivation were included in the questionnaire, while satisficing strategies were measured by paradata (e.g. item-nonresponse, speeding). Correlation analyses and linear regressions were conducted.
Results: Indicators of motivation and of satisficing are just partly correlated. Unmotivated gamers are rather clicking though the survey (speeding), while on energy or politics interested, but unmotivated, Facebook-user, are rather causing item-nonresponses. The effect of motivation and satisficing on substantial results differs as well. Non-differentiation is of more importance following the results of gamers, while indicators of motivation has to be considered when the results of “unmotivated” Facebook-user are interpreted.
Added Value: The presentation raises awareness for the relevance of motivation and subsequently satisficing of social-media survey participants. Conducting such information is important in order to interpret the results of such informative samples appropriately.