Over the last few days I have been going over some recently gathered data, as well as revisiting some previously collected data examining right-wing authoritarianism and attitudes toward torture. Based upon what I know of Altemeyer's work, as well as some of my own research on authoritarianism and attitudes toward violence (e.g., Benjamin, 2006; Benjamin, in press), I would be truly surprised if there were no positive relationship between RWA and attitudes toward torture. More authoritarian individuals should show higher favorability toward torture.
A couple years back, I had some of my methods students attempt a replication of Crandall et al.'s (2009) experiment on status-quo framing and attitudes toward torture. We included some individual difference variables that in theory should serve as significant predictors in their own right, and that would hopefully interact with the status-quo frame manipulation. The smaller, less heterogeneous sample didn't yield the significant framing effect or interaction that we had hoped for (although the trends were certainly in the predicted direction), but there was evidence of a significant positive relationship between scores on Altemeyer's (1996) RWA scale and scores on Crandall et al.'s (2009) attitudes toward torture scale.
Although an intriguing finding, questions were left unanswered. In particular, I had to wonder what specifically about RWA was responsible for our initial finding. Utilizing Altemeyer's theoretical work, and his empirical work examining such variables as vigilantism and so on, I suspected that ultimately, it was the dimension Altemeyer calls authoritarian aggression that was really driving that relationship, and not so much authoritarian submission or conventionalism. However, as Funke (2005) correctly notes, Altemeyer's questionnaire measuring RWA consists almost exclusively of double-barrel and triple-barrel items: ones that measure more than one of the construct's dimensions. With the data I had available, all I could do was speculate.
More recently, I conducted a study utilizing Funke's (2005) RWA scale, which was developed to correct Altemeyer's scale's psychometric difficulties, and includes questions that are specifically measuring authoritarian aggression, authoritarian submission, and conventionalism. I had an item from Weems et al.'s (2012) Conservatism/Liberalism questionnaire that tapped into attitudes toward torture (as part of that questionnaire's Intolerance subscale). With a sample of over 200 adults, I was able to examine partial correlations between each of Funke's (2005) RWA subscales and the attitudes toward torture item from Weems et al. (2012). I used partial correlations since the RWA subscales were highly intercorrelated, and I wanted to get as pure a reading of the relationship between each subscale and attitudes toward torture, while holding variability on the other subscales constant (in other words, statistically controlling for those other subscales). As Altemeyer would undoubtedly predict, authoritarian aggression was a significant predictor of attitudes toward torture. Higher authoritarian aggression scores were associated with more highly favorable attitudes toward torture. The partial correlations between the other two subscales and attitudes toward torture yielded no significant relationships.