Letter/Observer Forum

A Word of Caution About Comparing Different Populations

Jan_cover_webThe lead article in the January 2015 Observer, “Talkin’ About Your Generation,” is about purported generational differences in attitudes, values, and self-perception that were found in several large-scale studies. The studies’ results were noted to be often inconsistent.

This should not be surprising. The studies are based on different populations. The fundamental basis of statistical inference is that we are allowed to generalize to populations from which we have randomly sampled. This is wildly untrue of the studies cited. They are not random, or even “sort of random,” samples of any population. Rather, the studied groups are themselves rather specialized subpopulations. Only vague, suggestive conclusions are therefore justified, and it is not surprising that different subpopulations differ.

It is quite depressing for one who has observed psychological research for over 60 years to continually find that current studies are no more rigorous than they were in the 1950s.

-Norman Cliff, APS Fellow


Leave a Comment

Your email address will not be published.

Required fields are marked*

This site uses Akismet to reduce spam. Learn how your comment data is processed.