A simple data set for in-class illustration about how to estimate and interpret interactive relationships. The data here are deliberately minimal for that end.
Format
A data frame with 5914 observations on the following 14 variables.
versionversion identifier from ANES
caseidtime-series case identifier from ANES
healthoppose/"NFNO"/favor abortion if pregnancy would hurt woman
fataloppose/"NFNO"/favor abortion if pregnancy would cause woman to die
incestoppose/"NFNO"/favor abortion if pregnancy was caused by incest
rapeoppose/"NFNO"/favor abortion if pregnancy was caused by rape
bdoppose/"NFNO"/favor abortion if fetus would be born with serious birth defect
finoppose/"NFNO"/favor abortion if having child would impose financial hardship
sexoppose/"NFNO"/favor abortion if the child will not be the sex the woman wants
choiceoppose/"NFNO"/favor abortion if woman chooses to have one
pidrespondent's partisanship (Democrat, Independent, Republican)
knowspeakerwas the respondent able to correctly identify the Speaker of the House (John Boehner)
addchoicean additive scale of the abortion scores
lchoicea continuous latent scale of pro-choice scores (from a simple graded response model)
Details
"NFNO" = "Neither Favor Nor Oppose". All abortion prompts are on a 0-2 scale where 0 is oppose, 1 is "NFNO", and 2 is favor. The respondent's party identification is on a similar scale where 0 = "Democrat", 1 = "Independent", and 2 = "Republican". The additive scale of abortion scores has a minimum of 0 and a maximum of 16.