Sina Chen
Currently, I’m working as a research assistant in the “Political Predictors of Polling Errors” project. I started my dissertation at the Graduate School of the Social and Behavioural Sciences (GSBS) in autumn 2020, with a specialization in Information Processing and Statistical Analysis. Prof. Peter Selb is the principle investigator of this research project as well as the supervisor of my dissertation. Goal of this dissertation is to develop a broader understanding of random and systematic polling errors and their triggers across elections. I received a B.A in Politics and Public Administration and a M.Sc. in Social and Economic Data Science both at the University of Konstanz.
Publikationsliste
-
(2024): Contributions to Understanding - Pre-Election Poll Accuracy: A Cross-Election PerspectiveChen, Sina. 2024. “Contributions to Understanding - Pre-Election Poll Accuracy: A Cross-Election Perspective.” http://nbn-resolving.de/urn:nbn:de:bsz:352-2-189d8c4b81d070.
Pre-election polls play a crucial role in informing researchers, pollsters, candidates, and the public, serving as a significant empirical application of statistical principles. However, it has become increasingly apparent over recent decades that these polls do not consistently align with actual election results. Understanding the factors contributing to the success or failure of polls poses a considerable challenge. This dissertation seeks to enhance our comprehension of pre-election poll accuracy through a cross-election perspective. Using diverse data sources, recent political science advancements, and hierarchical Bayesian modeling techniques, I systematically explore numerous potential correlates of poll accuracy at the respondent-, poll-, and election-level. In collaboration with my colleagues, I employ and refine hierarchical Bayesian modeling techniques and elaborate theories on correlates of polling errors within the Total Survey Error (TSE) framework.
-
(2023): Bias and Variance in Multiparty Election PollsSelb, Peter, Chen, Sina, Körtner, John L., and Bosch, Philipp. 2023. “Bias and Variance in Multiparty Election Polls.” Public Opinion Quarterly 87(4): 1025–1037. http://nbn-resolving.de/urn:nbn:de:bsz:352-2-17m9x5yc2n00m2.
Recent polling failures highlight that election polls are prone to biases that the margin of error customarily reported with polls does not capture. However, such systematic errors are difficult to assess against the background noise of sampling variance. Shirani-Mehr et al. (2018) developed a hierarchical Bayesian model to disentangle random and systematic errors in poll estimates of two-party vote shares at the election level. The method can inform realistic assessments of poll accuracy. We adapt the model to multiparty elections and improve its temporal flexibility. We then estimate bias and variance in 5,240 German national election polls, 1994–2021. Our analysis suggests that the average absolute election-day bias per party was about 1.5 percentage points, ranging from 0.9 for the Greens to 3.2 for the Christian Democrats. The estimated variance is, on average, about twice as large as that implied by usual margins of error. We find little evidence of house or mode effects. Common biases indicate industry effects due to similar methodological problems. The Supplementary Material provides additional results for 1,751 regional election polls.