The problem of response instability in survey measures of policy positions has been studied for over 20 years without any apparent resolution. Two major interpretations remain: Philip Converse's nonat
The data contain a weight variable (WEIGHT) that should be used in analyzing the data. The weights were derived using demographic information from the Census to adjust for sampling and nonsampling dev
