From a reader:
Dear Jonah,
I just read your piece on NRO about the Stanford-Berkeley study on conservatism, downloaded the document and read the crucial sections quickly. About myself: I am a social-cognition psychologist (academic) doing research on the situational conditions that increase or decrease stereotyping. I should be analyzing some data right now, but instead I’m writing to you.
The Berkeley authors have started from the work of Adorno in the 1950s on authoritarianism. This work has been heavily criticized, as the authors admit on the first page, as being heavily value-laden. Validity is the chief difficulty with value-laden research and it is not an incidental side issue in research. It is the very foundation.
The authors then go on (page 339) to derogate the traditional personality approach to studying this area and suggest that their approach is that of social-cognition. If this were true, it would be noteworthy. Social cognition is a rigorous, experimental area focusing on the link between thought (schemas, memory, emotion, etc.) and behavior. However, the authors do not in fact use these methods! Instead, they have done a meta-analysis (a correlational study) of a whole bunch of data obtained through the use of scales, each of which is of questionable validity. In short, they have used the same weak methods of the personality approach. That they have used weak methods in a highly sophisticated manner only obscures the fundamental issue of validity. A casual inspection of the actual scale items used will quickly reveal the bias that is obvious to conservatives, but is unquestionable truth to liberals…
Carole L. Bandy, Ph.D.
Assoc. Prof. of Psychology
Norwich University