The study assesses the onset of covid-19 on people's political attitudes. It uses two data sources Nationscape: https://www.voterstudygroup.org/data/nationscape Year Long Longitudinal: https://openpsychologydata.metajnl.com/article/10.5334/jopd.54/ We share all of the relevant data from the YLL here, but only the cleaned data from the Nationscape. We do share our files for cleaning the data so that you can see our steps and clean the data yourself if you'd like (after you download the data at the link above). In both studies, the "models direction.R" and "models extremity.R" files will estimate the models. In both studies, there are "narrow" and "wide" versions of these files that will estimate the models using alternative definitions of covid onset. Some models/loops are commented out because they take a lot of time. Uncomment those that you want to run. Or make sure they are commented to avoid running them. The "analysis.R" file creates all of the figures and does some follow-up analyses on the models. Some of the figures from both studies are combined using the "combine_fig.R" file. We analyzed Google data for salience of covid. The data is in the "us_search_5.csv" file and the code is in the "search_fig.R" file. We analyzed CDC covid data to identify high covid states. The data is found in the "covid_cases_and_deaths.csv" and the code is "high_covid_states.R" We also include election result data that can be found in the similarly named folder.