Number of factors in the personality sphere: Does increase in factors increase predictability of real-life criteria?

Many researchers support a 5- 8-factor theory; Cattell urges a 16-factor system. Factoranalysis has been unable resolve this controversy. Hence, the degree which 6 and 16 factorsrelated a proposed new criterion (real-life data) was evaluated in 16 data sets. When factors wereincreased from 6 16, 88% of the studies showed a significant (p < .01) increase in R. The percent-age of variance accounted for, after shrinking the R&, was doubled. We concluded that the 5- 8-factor position has limited usefulness; use of more factors is strongly supported.There is a continuing controversy over the appropriate num-ber of factors. Theorists and researchers have pro-posed differing numbers of factors as being necessary andsufficient describe normal personality. Although attempts tointegrate or amalgamate the new factors with existing factorsare rare, Peterson (1965) felt that many factors in different sys-tems differed in name only and that only a few fac-tors had truly been denned. Others suggest that more effortshould be spent on a larger number of factors (Buss&Craik, 1985).An agreed-on minimum number of factors wouldenable systematic exploration of personality's relation othervariables. One must be able specify the universe of traitsthat define personality (McCrae & Costa, 1985, p. 711) dosuch systematic exploration. Without established personalityfactors, research efforts are fragmented.A well-defined taxonomy of would delineate therelations between agreed-on factors as the periodictable of the elements delineates the interrelationship of chemi-cal elements (Cattell, 1946). The periodic table allows chemis-try to express all compounds in a common vocabulary. Thatis, any chemical compound can be viewed as an amalgam ofbasic elements, and compounds that are similar in their ele-ments have other important similarities (Goldberg, 1982, pp.232-233). It is hoped that a taxonomy of would in-troduce similar clarity into the study of personality.In clinical practice, agreed-on factors are alsoneeded. Clinicians want be as specific as possible in under-standing their clients. Whether recommending ways increasea person's medication compliance or deciding how handle asuicidal person, clinicians want the maximum useful informa-tion. They wish avoid uselessly specific data on the one handand overly general information on the other. The lack of anagreed-on factor system is handicapping in clinicalas well as research settings.Correspondence concerning this article should be addressed Rich-ard L. Gorsuch, Graduate School of Psychology, Fuller TheologicalSeminary, 180 N. Oakland Avenue, Pasadena, California 91182.