Cattell arrives at the number 16: Eysenck at 2 (or 3).Other investigators arrive at still different numbers. We encountered a similar situation in Chapter 12, where we saw that the number of factors defining the concept of intelligence could be 1 (Spearman's general intelligence factor, g),7 (Thurstone's primary mental abilities),or as many as 150 (Guilford,1982).
Some of the discrepancy occurs because different traits are initially put into the analysis; some discrepancy occurs because different types of data are being analyzer (for example, peer-ratings versus self-ratings);and some occurs because different factor analytic methods are employed-it is not as cut and dried a technique as it may first seem. But much of the disagreement is a matter of taste. A researcher who prefers a more differentiated or fine-grained description of personality will stop the factor analysis earlier and accept more factors, arguing that important distinctions would be lost if the factors were further merged. Another researcher, like Eysenck, will prefer to merge several lower-level factors into mare general ones, arguing that the resulting factors will be more stable (more likely to reemerge in other analyses).For example, when Cattell's 16 factors are themselves factor analyzer, Eysenck's 2 factors emerge as superfactors.