to the American Mass Public Opinion in the 1930s and 1940s Dataverse.

This project began in 2005, as a collaboration between researchers at MIT, under the direction of Professor Adam Berinsky, and at UC Berkeley, under the direction of Professor Eric Schickler. The aim of this work is to clean, recode, weight and disperse survey data from the 1930s and 1940s, which because of problematic sampling procedures, has not thus far been usable for academic research.

Raw data (.dat) files are available through the Roper Center. This website provides STATA .do, and .dct files to be used in conjunction with this raw data. Corrected codebook files have also been provided, as many of the original PDFs do not correspond well to the data, are incomplete, or illegible.

Please use the menu on the top-center of the page to navigate this site.

If you are new to the Dataverse, you will need to create an account to access the data.