Saturday, April 16, 2011

Comments on SIOP 2011: Narrow vs broad traits

The tremendous interest in Big Five factors has not resulted in improvements of personality tests in the sense of increased validity. The reason is that the five factors all have no or only very modest validity in relation to job performance. Very extensive research, summarized in dozens of meta analyses, document this fact. There is now growing consensus about the need to develop and use measures of "narrow traits" in order to increase validity.

Some of these narrow traits are found among the facets of the Big Five, others not. The principle guiding the search for an improved basis of testing are seldom found in theory, but in common-sense thinking about what traits one could expected to be of importance with regard to important criteria, such as psychopathy in the case of counter-productive behavior. Other examples are emotional intelligence and affect. The Big Five cannot account for more than a minor share of the variance of any of the narrow traits.

In order to get improved personality tests there is a need for tests which complement the Big Five with a number of narrow trait scales which are focused on job functionality. Some can be found among the facets, others not. There is a need to limit the narrow trait scales in order to make test validation and test interpretation practically manageable. The illusory richness of tests having 30-40 subscales creates merely a feeling of understanding the tested person, a feeling which is a reflection of a well-known tendency to over-estimate the value of information, if there is more information.

The feeling that the Big Five somehow constitute a final answer to personality and personality testing is fading. The sooner that belief is abandoned, the better it is.

No comments:

Post a Comment

Free counter and web stats