This project contains the replication files for "Measures of political ideology and beliefs are not appropriate for measuring ideologies in people’s heads because they primarily measure stable traits" For both questions, the NL data obtained from LISS needs to be downloaded from them. See read me files in the respective data folders. Once the data is obtained... # Question 1 If you want to run the STARTS models yourself, run all of the code in the folder "code for blade" for all datasets. Combine the outputs for each data set from that code with the code file "starts.combineoutput.R". If you do not want to run the STARTS models yourself (which honestly, is the way to go; they take a long time), the output from the starts models are in the folder "blade results". In this case, you can use the "figs.R" file for the main analyses. Please note that if you just want to work with the output of the STARTS models the raw data (e.g., from LISS) is not needed. # Question 2 Run the scripts with "_models" in the names. Then run the analysisfigures.R code. If you do not need/want to rerun the models to get the correlations, you can run "analysisfigures.R" without the raw data. This set of code is also where we calculate things like age and gender distributions.