It is so smart of this paper to take advantage of financial data to predict new candidates' legislative voting behavior. But as you mentioned at the start of the paper, there're inborn limitations of ideology measures such as DW-NOMINATE. Sorry if I understood the paper incorrectly, the supervised models derived in the paper are still learning to fit these measures. Is it possible that the models can provide new insights in return to develop better measures of political ideology?
w4rner
commented
5 years ago
Thanks in advance for your presentation tomorrow!
It is so smart of this paper to take advantage of financial data to predict new candidates' legislative voting behavior. But as you mentioned at the start of the paper, there're inborn limitations of ideology measures such as DW-NOMINATE. Sorry if I understood the paper incorrectly, the supervised models derived in the paper are still learning to fit these measures. Is it possible that the models can provide new insights in return to develop better measures of political ideology?
Thanks!