Our research group has been awarded a grant from the University Excellence Fund (Egyetemi Kiválósági Alap, EKA) for the project "A new method to improve the accuracy of estimates in empirical social research: accessibility-based post-stratification".

During the research, we use a new method to forecast elections from questionnaire data, using specific proxy questions, and then perform post-stratification based on these questions. Built on the specific questions, we divide the surveyed sample into the following groups.

Our main question is to what extent we obtain different results when estimating attitudes towards participation in elections by adjusting the responses not only for demographic characteristics but also for accessibility groups.

Before the European Parliament elections, we carried out two data collections. In order to do this, we contacted nine research firms, and finally the surveys were carried out in cooperation with the market and public opinion research company Ipsos. The first data collection took place between May 14-18, while the second query was timed closer to the elections, between May 29 and June 4. The time distribution of these two data recording waves is illustrated in the figure below.

In the current phase of the project, we are working with the data generated by the two surveys: among other things, we compare the results we measured with the forecasts of other polls, we examine the time differences between the two data collections, and we carry out weighting and corrections on the data aimed at improving the estimates. In addition, we organize the respondents into different contact groups based on how easily the operators managed to contact them. We do this using several different methods, which are currently being developed. From the resulting contact-related variables, we will also use cluster analysis to examine differences between each group. A possible version of the availability grouping is illustrated in the figure below.

The project will run until February 2025, a particularly timely research theme in the run-up to the European Parliament and local elections. The method builds on several international and national studies and could lead to a method that can be used by the social sciences.