We factored the 236 items comprising the ‘‘work activities’’
measure using principal axis factoring and direct
oblimin rotation. The final solution was based on frequency
ratings data because there has been some controversy over
the use of importance ratings in factor analysis (i.e., there
may not be actual true score variability in importance
ratings in that importance should be a constant; Cranny and
Doherty 1988). Therefore, we constructed the scales based
on the factor loadings of the frequency data by averaging
the ratings for each item found to load on a factor. We
chose oblique rotation over orthogonal because dimensions
of interest to psychologists, as in the present study, are not
often dimensions we would expect to be orthogonal. In the
present study, some of the dimensions studied are conceptually
similar, although distinct from other dimensions
addressed. To the extent that they are somewhat related
conceptually in terms of job content or worker attributes
required, it should be expected that they will co-occur in
jobs (i.e., if one dimension is present, then a similar and
related dimension may also be more likely to be present). If
the latent variables are, in fact, correlated, then an oblique
rotation will produce a better estimate of the true factors
and a better simple structure than will an orthogonal rotation