Comparison of the impact of online news and social media on the electoral preferences of political parties in the Czech Republic between 2017-2025

This bachelor thesis analyzed the influence of visibility and sentiment on social media and news websites on the preferences of Czech political parties in the period 2017-2025. The thesis aimed to analyze these influences and specifically to determine their existence, direction, and magnitude. Correlation and panel analysis were chosen for the purpose of analyzing the influence. Longitudinal data from the twenty most-read news websites and the three most-used social networks were used for the analysis. In total, data for nine parliamentary parties, 2 million news articles, and 1.5 million posts from Facebook, Instagram, and X were analyzed. Sentiment data was generated by a machine learning model. The results indicate a statistically significant positive influence of political party visibility on electoral preferences. The influence of visibility as a whole was confirmed for visibility on social networks, but this relationship could not be proven for news websites. The influence of sentiment as a whole had a negative effect on preferences. However, the sentiment variables of social networks and news websites separately were not statistically significant. The absolute strength of the influence of visibility was many times greater than the influence of sentiment on party voting preferences. However,...