I’m not convinced that it justifies your assertion that “many voters seem to make political choices using disturbing criteria”

The local election results described by Anthony Wells in an earlier post on this thread are another disturbing example.

Do they?

How do you know how many?

What “disturbing” criteria, and why “irrelevant”?

I know that psychometric tests are extensively validated. I also know there are issues in validating psychometric tests for different ethnic groups, for instance because the language tests are written in may contain cultural assumptions.

However, I stand by what I wrote. Consider the grief that has been caused to psychologists such as Richard Lynn and Chris Brand (neither of whose positions I agree with) who have published work on race and intelligence. Let me assure you that there are plenty of psychologists and psychology graduates who do not want to get caught up in similar controversy. Me included.

In addition, whilst I agree that the firms that develop psychometric tests validate them extensively, test producers often seem to avoid giving prominence to the results. Rather than race, consider sex differences, where the difference between the two sub-groups tested is very clear and there are therefore fewer validation problems. Now, men and women on average achieve similar scores overall on tests of ability (if you like, they get similar IQs). But this is because the psychometric instruments are carefully constructed to achieve this outcome by weighting different sub-scales. It is well-known that males score higher on certain types of ability (for the sake of argument, motor skills) and females on others (say verbal ability). That the differences can be large is shown by certain professions, such as professional footballer or nursery nurse, where there is a glaringly obvious imbalance between the abilities of the two sexes to do the job, although of course the occasional individual can always be an exception to the general results found statistically. But there are obvious problems in applying such statistical results for instance when using psychometric tests to select for jobs where there are political pressures to appoint equally. This is perhaps true not least for MPs, although let me hasten to add that it is probably the case than men and women tend to perform the role of MP differently, rather than that the “better” MPs usually come from one sex.

Interpreting results incompletely to avoid reaching uncomfortable answers is not the same as confusing different variables when interpreting results.

A short way of putting my reply to Peter is that psychometric tests are highly political in the widest sense, and that psychologists’ use of such instruments adjusts to political imperatives, even if not as many psychologists now highlight the politics of psychology as was the case in the late 1960s.

Andy JS, most of us agree with you. The problem is that many voters seem to make political choices using disturbing criteria which are irrelevant to the fitness of candidates to perform the elected role. The question is how we can use psephological knowledge towards encouraging people to make their voting decisions on grounds that trest candidates fairly and with respect.