• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

News

Let the work begin: the first meeting of the Research and Study Group

The first meeting of the Research and Study Group (RSG) called «Adaptation of Methods for Interaction Detection in Social Data» was held on the 23rd of January and marked with two events. The meeting was opened by the introductory speech of the head of RSG Alexey Rotmistrov and continued with the discussion of participants` individual research projects.

Let the work begin: the first meeting of the Research and Study Group

© Photo: Tamara Mkhitaryan, 2018

The report of Alexey Rotmistrov was devoted to the general scheme of methods` application to search interaction effects in sociological data with the view to improve the quality of regression modeling. The information given in the report was illustrated on the data of cross-national European survey (ESS) and focused on variables that were possible determinants of trust in the institution of elections. Apart from participants of the RSG, the speech of the head of RSG was attended by other students and teachers from the HSE Department of Sociological Research Methods.

In the report, Alexey Rotmistrov underlined the importance of detecting reasons of the low explanatory ability of regression models in sociological research and also told the audience several tricks and ways to improve the situation with explanatory ability. As it turned out later, this speech helped the participants of the seminar to understand more fundamental ideas about the final goal of the Research and Study Group. 

Video recording of the report is available here (in Russian).
*The algorithm stated by Alexey is presented also in the form of the scheme: Regression_scheme_eng

After the report, the participants of the RSG discussed general questions that had risen during the lecture and then began to consider their individual research and technical tasks for the next two weeks. The discussion was ended with a repetition of the self-studied material on multidimensional log-linear analysis and regression analysis with deep interaction effects.