The first results of the RSG work were successfully presented at the scientific conferences
Maria Rodionova and Tamara Mkhitaryan presented the results of the search and application of interaction effects in their reports at the XII Russian Scientific Conference on the Remembrance of Yuri Levada. While on the XIX April Conference Svetlana Zhuchkova and Alexey Rotmistrov told about the research of handling missing data with CHAID.
Traditionally, April is the most intensive month for scientific events in the academic year. The results of the first months of the RSG work were presented at two conferences immediately: on the XII Russian Scientific Conference on the Remembrance of Yuri Levada "Modern Society and Sociology" and the XIX April International Scientific Conference on Economic and Social Development.
During the Yuri Levada conference, Maria Rodionova and Tamara Mkhitaryan, student members of the RSG, reported about various options for finding interaction effects as well as about their experience using these in their personal researches.
Maria Rodionova explained the principles of log-linear analysis, connected with the hierarchical type of the obtained models and the process of excluding insignificant combinations of variables, to listeners. The thematical side of her work was devoted to predicting the violence level in a couple. The student proved that taking into account the interaction effects is able not only to increase the formal quality of the model but also to deepen the interpretation. Thus, she managed to identify and describe many patterns that lead to the demonstration or inhibition of violence in a couple. Questions of the listeners were turned to the theoretical aspect of the research.
Tamara Mkhitaryan presented an alternative way of selecting significant interactions – the use of decision trees. She described how the transition from the results of building a tree to using the effects found in the regression works, what are the restrictions of prediction with the help of the trees themselves, and what meaningful and methodological results were achieved by combining different methods. Tamara Mkhitaryan's report, which was devoted to the determination of the factors that influence on the choice STEM-disciplines by modern schoolgirls, resulted in a broad discussion of the changing values and perceptions in this choice.
The other methodological direction of the RSG work was presented at the annual April conference by the team leader Alexey Rotmistrov and his deputy Svetlana Zhuchkova. The presentation on the study of handling missing data with CHAID and the discussion that followed it touched upon the broadest methodological issues of sociological research: what choice and on what basis should a sociologist make while working with missing data? How does the ratio of the sample and population changes when there are missings in the database; who is represented by the available data? Are there any alternative ways of working with missing data that compensate for existing imperfections of the imputation and deletion of data?