AAU Students Help with Data in New Study

AAU SIRD researcher Alexei Anisin published a study based on data that were coded by an AAU student-led research team which included former SIRD students Anneliese Scalzo, Eulalie Bintein, Jessica Kazimirova, Mateja Zherajikj, Natalie Smutna, Savelii Nevzorov, Dijana Zherajikj, and Yuliya Vilgelm.

The study “Fraternisation and Repression during the 2020–2021 Attempted Revolution in Belarus” uses Video Data Analysis (VDA) to explore interactions between protesters and state forces during the largest protests in Belarus’s post-communist history. It seeks to understand why state repression and violence occurred in some instances but not others. Anisin statistically analyzes 144 video cases documenting various interactions, including verbal exchanges and attempts at fraternization—efforts by protesters to initiate a positive dialogue with police or security forces.

Data Collection Sample Size and Scope: The study analyzed 144 cases from video data that captured interactions between protesters and authorities, including verbal interactions, fraternization attempts, and the subsequent responses of the security forces. Source and Nature of Data: Videos were sourced primarily from platforms like YouTube and Twitter, providing real-time, unedited sequences of events.

Data Coding: Key variables such as fraternization, positive dialogue, and repression were coded from these videos. The coding involved determining whether fraternization attempts were made, if positive dialogue ensued, and if repression occurred.

Statistical Analysis Correlation Analysis: The study employed tetrachoric correlation analysis to examine the relationships between fraternization, positive dialogue, and the occurrence of repression. Tetrachoric correlation is used specifically to measure the association between binary variables, which is suitable for the dichotomous nature of the coded video data (e.g., presence or absence of repression). Logistic regression was used to predict the likelihood of repression based on the interaction variables. This model helped quantify how verbal interactions, fraternization, and positive dialogues influenced the probability of state forces resorting to violence. The models indicated that positive dialogue was a significant predictor of reduced repression, highlighting its potential as a strategic tool for de-escalating potential violence during protests.

Key findings include: Fraternization and the establishment of positive dialogue between protesters and state forces are negatively correlated with occurrences of state repression. This suggests that when protesters managed to engage positively with security forces, the likelihood of repression decreased. The study highlights the micro-dynamics of repression and dissent, revealing that certain types of interactions can reduce the likelihood of violent repression. The analysis also examines the role of Russian military-grade support to the Belarusian government, noting its significant impact in preventing the revolution’s success.

This inquiry contributes to understanding the nuanced interactions that can either escalate or de-escalate violence during protests and the strategic responses by state forces to manage or suppress dissent.