,
6 tweets,
3 min read
Read on Twitter
1/n Percent of liberals who 'strongly disagree' that racial problems in the US are rare jumped more 10 points between the 2016 and 2018 waves of the Cooperative Congressional Election Study. 
2/n Also find a significant interaction between the # of reported political social media activities over the previous 24 hours and ideology on the probability of a strongly disagree response
3/n I also tested to see whether other media consumption has similar predictive effects. In the end, the positive relationship appears to be limited to *digital* and *national* news consumption.
4/n Really need to test whether these relationships are causal vs. selection-driven (though I did control for political interest, among other variables).
5/n FYI, the item in question is often used as a measure of racism (i.e. 'agree' = racist)
6/n Should also point out (for those that started following me only recently) that the above is but one of a series of similar findings
