Friday, August 08, 2003


Psychologists normally obtain a score for how conservative a person is by adding together the number of allegedly “conservative” statements he agrees with and the number of allegedly “liberal” statements he disagrees with. Deciding which statements to present as representing “conservative” or “liberal” views is however a major issue which is commonly resolved with great deal of arbitrariness.

I have noted previously how the measure of conservatism principally relied on by the Berkeley group has shown fairly laughable characteristics on some occasions -- with people who tend to assent to allegedly “conservative” statements also tending to assent to allegedly “leftist” statements! The implication of that result is that the items in the questionnaire are very poorly chosen and do not in fact represent distinctively conservative and liberal views to the people answering the questions.

If psychology were a science, one would have thought that a result as anomalous as that -- particularly when published in a widely-circulated academic journal -- would have caused all subsequent academic users of the questionnaire concerned to at least examine the correlation between the liberal and conservative items in their data. As far as I can see, however, nobody did. I therefore wrote to a large range of other users of the questionnaire and asked for copies of their raw data so that I could carry out the requisite analyses myself. There were only six fellow-psychologists who responded to my request -- mostly people with whom I had some prior acquaintance.

The results are reported in my post of 7th. here (or here). I found that the questionnaire was not uniformly unsatisfactory -- it did work as it was supposed to on the small sampling of occasions concerned. Such a finding does of course constitute some warrant for continued confidence in the usefulness of the questionnaire. That highly unsatisfactory correlations (such as the significant positive correlation between supposedly opposed items mentioned previously) can also be obtained, even with students, does mean, however, that the many studies which fail to report the correlation between the “liberal” and “conservative” items continue to be of dubious meaning. Accepting their results at face value is therefore a matter of faith rather than of science.


No comments: