This R Shiny application implements the models implemented in my paper on estimating the partisan bias in survey questions. Input any free text and the application will produce an estimated partisan bias, the most partisan unigrams and bigrams, and the question’s location on the ideological spectrum.
