CLAIRE LEHMANN: How a rebellious scientist uncovered the surprising truth about stereotypes.

When I went through University as a psychology undergraduate Jussim’s work was not on the curriculum. His studies were not to be found in my social psychology textbook. Nor was Jussim ever mentioned in the classroom. Yet the area of study Jussim has been a pioneer of – stereotype accuracy [PDF] – is one of the most robust and replicable areas ever to emerge from the discipline.

To talk about stereotypes, one has to first define what they are. Stereotypes are simply beliefs about a group of people. They can be positive (children are playful) or they can be negative (bankers are selfish), or they can be somewhere in between (librarians are quiet). When stereotypes are defined as beliefs about groups of people (true or untrue), they correlate with real world criteria with effect sizes ranging from .4 to .9, with the average coming in somewhere around .8. (This is close to the highest effect size that a social science researcher can find, an effect size of 1.0 would mean that stereotypes correspond 100% to real world criteria. Many social psychological theories rest on studies which have effect sizes of around .2.)

Jussim and his co-authors have found that stereotypes accurately predict demographic criteria, academic achievement, personality and behaviour. This picture becomes more complex, however, when considering nationality or political affiliation. One area of stereotyping which is consistently found to be inaccurate are the stereotypes concerning political affiliation; right-wingers and left wingers tend to caricature each others personalities, most often negatively so.