Celem artykułu jest określenie tego, jak dobrze sondaże przedwyborcze potrafią przewidywać wyniki wyborów parlamentarnych oraz od czego zależy trafność tych prognoz. Zmiennymi wyjaśnianymi jest poprawne wskazanie zwycięskiego komitetu oraz wyniku wyborczego poszczególnych opcji politycznych biorących udział w wyborach. Pierwszą zmienną wyjaśniającą jest czas między badaniem a datą wyborów. Drugą zmienną wyjaśniającą jest różnica wskazań między dwoma komitetami o największym poparciu. Dane empiryczne obejmują wyniki sondaży w okresie 12 miesięcy przed wyborami parlamentarnymi w Polsce od 1993 do 2015 roku. W analizie wykorzystano bayesowski model hierarchiczny i symulację Monte Carlo. Częściowo potwierdziły się hipotezy, które wskazywały, że zdolność sondaży do przewidywania największego poparcia i do wskazania wyników wyborczych poszczególnych komitetów jest tym większa, im bliżej do wyborów. W pełni potwierdziła się hipoteza, że sondaże tym lepiej wskazują zwycięski komitet, im większa jest różnica między sondażowym poparciem dwóch głównych oponentów.
EN
The aim of this article is to determine how well pre-election polls can predict the results of parliamentary elections, and what determines the accuracy of these predictions. The dependent variables are 1) the correct indication of the winning party and 2) the accuracy of election surveys in forecasting voters’ support. The first independent variable is the time between the poll and the date of the election. The second explanatory variable is the difference in results between the two parties with the greatest support. This study uses data from all publicly available polls that took place in the 12 months before every parliamentary election in Poland from 1993 to 2015. The analysis uses Bayesian hierarchical modeling and Markov Chain Monte Carlo simulation. The results show that the average probability that a pre-election poll will correctly predict the winning party is around 80%, whereas the probability that it will correctly predict the distribution of voters’ support (with 3% error margin) is around 50%. The evidence partially proved that the forecasting accuracy of an election poll is the better the closer the poll is taken to the date of the election. It was also proved that the ability of a poll to predict the winner is better the greater the gap between the survey results of the two leading parties.
Increasing numbers of citizens rely on social media to gather both political and non-political information. This fact raises questions about belief formation and belief updating in the social media setting. Using Facebook data on users’ behaviour in Poland in 2017, I test the hypothesis that individuals tend to like content that confirms their beliefs. I measure the political preferences of nearly 1.4 million users who were active on the main political and news media pages and classify them as being supporters of certain political organisations or as being politically unaffiliated. Based on the principles of analytical sociology, I construct a theoretical model that may explain the results. According to the model, users tend to like posts from only one source of information. There are also statistically significant differences in the news media preferences of supporters of different political organisations. They are prone to like posts published by sources that accord with their views. The model also correctly predicts that politically unaffiliated users choose media outlets that are considered unbiased or less biased. The results support the hypothesis that users of social media prefer exclusive or near-exclusive sources of information.
The article deals with the issue of voting overreporting in social surveys. The comparison of voter turnout reports in public opinion polls to official statistical data often shows great difference. Respondents answer that they voted or would vote whereas they actually didn’t. The culprit of this effect is social desirability bias i.e. reporting false attitudes and behaviours in a manner that is consistent with social norms. As a result analyses of voting behaviours have systematic error. The article presents methods for reducing social desirability bias which include: choice of research techniques which assure anonymity and confidentiality, question wording which lowers social pressure to give false answers, indirect questions, double-rating method, randomized response technique and making post hoc corrections.
JavaScript is turned off in your web browser. Turn it on to take full advantage of this site, then refresh the page.