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).
A6: Election Studies and Respondent Behavior
Tracking the decision-making process – Findings from an Online Rolling Cross-Section Panel Study
1University of Mannheim, Germany; 2YouGovPsychonomics AG
Relevance & Research Question:
One of the most important trends in electoral behaviour is the loss of stability in recent years. That said, it becomes increasingly important to study (and hence observe) short-term effects. But tracking decision-making processes has always been a challenge to academia as well as market researchers, since neither our traditional theories nor our tools are specifically designed to examine such short term effects. To overcome this deficiency we will conduct an innovative study design in the run-up to the 2011 Baden-Württemberg state election. It combines panel elements with the general idea of a rolling cross-section survey, yielding rolling cross-section panel waves. The rolling-cross section part provides us with the opportunity to track developments in the course of the campaign on a day-by-day basis, while the panel gives insight into intra-individual decision-making process. Our contribution will focus on the design of the study, but will also include findings from the first wave of our study that will be fielded in November 2011.
Methods & Data:
In the run-up to the 2011 Baden-Württemberg state election, we will conduct a five wave-panel. Respondents are recruited from YouGov’ access panel. The first wave (with n=4000) – providing a “pre-test” before the final phase of the campaign starts – will be fielded in November/December 2010. The following three waves will be conducted in the final six weeks before Election Day in order to get an extensive insight into the decision-making processes. Wave 5 will be a standard post-election wave.
Since the GOR-Conference will take place when our study is still in progress, we will only be able to present first findings mainly from our first wave. However, we will focus on outlining the concept and innovations of our Rolling Cross-Section Panel Study.
The paper presents a very innovative research study useful tools for academic and non-academic researchers that are interested in individual decision-making processes in future surveys.
A new approach to the analysis of survey drop-out. Results from Follow-up Surveys in the German Longitudinal Election Study (GLES).
1GESIS - Leibniz Institute for the Social Sciences, Germany; 2Mannheim Centre for European Social Research (MZES), University of Mannheim; 3University of Mannheim
(a) Relevance & Research Question
The proposed paper builds on findings presented by the authors at the GOR 10. High drop-out rates are considered a major shortcoming of web surveys and considerably threaten data quality. However, despite growing scholarly attention the knowledge on survey drop-out is still fractional. Previous research mainly addresses the impact of survey design, question wording, and characteristics of the respondents on survey drop-out via ex-post statistical methods. The research presented here is innovative in that the respondents are asked directly about the reasons for dropping out, the interview situation, and psychological predispositions in a follow-up survey.
(b) Methods & Data
Based on our previous research regarding survey drop-out, the principal investigators of the GLES granted funding for a series of short follow-up surveys of drops-outs. These surveys will be conducted subsequently to three consecutive online trackings of the GLES, beginning in December 2010. According to experience, a gross sample size of about 400 drop-outs per survey can be expected. Given an estimated response rate of 60 percent a net sample size of 210 to 240 per tracking is anticipated, thus providing a unique database of more than 600 interviews with drop-outs. Since the most essential items are also included in the tracking surveys, the design allows for comparisons between drop-outs and complete responders. Due to the explorative character of the research, the presentation will mainly focus on descriptive statistics as well as multivariate models illustrating our major findings.
First results will be available by mid-January 2011.
(d) Added Value
Follow-up surveys of respondents who dropped-out allow for an enhanced understanding of the complex processes underlying the phenomenon, especially with respect to the subjective reasons of the respondents as well as the situational influences and psychological predispositions, which cannot be studied applying ex-post statistical procedures. In this regard, our research will add to the knowledge on the reasons for drop-out in web surveys and amend both the theoretical explanations of and the prospects for reducing drop-outs.
Measuring Political Deliberation on Twitter: Forms and Functions of Digital Interaction through Microblogging
University of Bonn, Germany
Relevance & Research Question:
Partly driven by social media, the mediatisation of society has spawned new forms of political communication in the public sphere, accompanied by high hopes for more deliberative discourse along the Habermasian model of deliberative democracy. The paper explores how patterns of interaction and argumentation of political discourse on twitter evolve online discoursive practices.
Methods & Data: Material basis consists of 2500 tweets, which were collected during the 2010 North Rhine Westphalia state election. These tweets include: (1) tweets by selected candidates of each party, (2) by print media twitter portals, (3) by regular citizens posting about the election. These tweets were documented in a data basis three weeks prior to and one week after the election. By using the methods of content analysis, argumentation theory, and speech act theory, the focus of the analysis included semantic references (#hashtags), action level of the exchanges (@replies and @retweets) between the respective user groups, and cross media references in the journalistic brand media. The overall methodolocgial process can be characterized by triangulation, combining qualitative and quantiative measures.
Results: The quantitative analysis revealed that most tweets (80%) were single item, non-referential postings by media brands. The qualitiave analysis, however, showed, that closer to the elections, politicians engaged in an “in-group interaction”, by which members of the same or competing parties exchanged their views among each other. The broader public did not participate as actively as hypothesized.
Added Value: It will be argued that so far the idealized concept of the deliberative demoracy is only partly supported by the data, as the average citizen does not participate very actively. On the contrary, at least the data from 2010, shows excluding tendencies as the press/politican activists form an ingroup on twitter.The paper thus offers a critical insight into new ways of political deliberation via social media.