toward congruent occupational choices when opportunities arise to
do so (Spokane, 1985; Oleski & Subich, 1996). Because Holland
researchers have generally not corrected studies for the attenuating
effects of range restriction (especially on job satisfaction) and unre-
liability of measures, these findings may constitute an underestimate
of the congruence effect. Experimental and social process studies,
as well as methodologically sound longitudinal analyses, offer more
compelling tests of the model and generally appear to be more sup-
portive of the fluid propositions in Holland’s theory (Spokane et al.,
2001).
Findings on the relationship of consistency and differentiation
with career outcomes have been less favorable to the theory, though
the methodology and quality of these studies has been spotty. Fur-
ther, only one or two studies have adequately examined the com-
pound effects of all of the diagnostic signs on career outcomes.
Interests and Personality as Unique Constructs Relations between
interests and personality are a topic of intense, ongoing debate
(Costa, McCrea, & Holland, 1984; De Fruyt & Mervielde, 1997,
1999; Gottfredson, Jones, & Holland, 1993; Hogan & Blake, 1999;
Schinka, Dye, & Curtiss, 1997; Tokar & Fischer, 1998; Tokar &
Swanson, 1995; Tokar, Vaux, & Swanson, 1995). The issue, stated
simply, is the possibility that measures of interests and measures of
personality differ only with respect to item content employed. The
implication of this position is that when we are measuring interests,
we are tapping the same underlying construct as when we are mea-
suring personality; interests and personality are simply two duplica-
tive aspects (for example, arms or legs) of the same individual and
therefore largely similar, despite one being the left (arm, personal-
ity) and the other being the right (leg, interests). Depending on the
perspective from which one views the individual (from a personal-
ity perspective or from an interest perspective), one is simply exam-
ining the same individual’s makeup from a different direction.
Holland types, then, would be broad personality-based dispositions
reflecting the full range of behavior, attitudes, and so on rather than
HOLLAND’S THEORY 401