Brown and Gore (1994), in another Monte Carlo study, com-
pared ten indices and created a new index (c), with a symmetrical
underlying distribution (as opposed to positively skewed distribu-
tions underlying most other indices). This study used simulated data
and explored not only the sensitivity of the index and its ability to
distinguish codes with the same three letters in different order but
showed that indices with normal curves underlying them were more
useful in congruence studies. This research clarified the reasons for
different findings with different indices, created a new index with a
symmetrical underlying distribution, and spawned an extremely use-
ful new computer program for calculating congruence (Gore, 1995).
Finally, Young and colleagues (1998) compared eleven indices
of congruence and found weak relationships between congruence
and job satisfaction using all eleven congruence indices with 483
employed adults in 170 occupations. The researchers found little
moderating effect for sex and a weak moderating effect for different
Holland types, with Investigative being the strongest relationship
with job satisfaction (r= .21). This study is not in accord with other,
more recent studies and deserves close scrutiny. Tinsley (2000) ar-
gues, however, that absent true commensurate measurement (of
environments and persons), indices of fit will be no better than a
“present status” model in which dimensions of persons, environ-
ments, and interactions are entered simultaneously (Edwards, 1991).
Clearly, this criticism of the index paradigms is worthy of careful
consideration.
The extensive research literature on person-environment fit in
Holland’s theory suggests several conclusions. Although all of the
outcomes have shown contradictory findings at one time or another,
the correlational studies of congruence are particularly susceptible
to methodological artifact and conflicting findings. Nonetheless
there is a consistent but modest (ranging from r= .15 to r= .54) rela-
tionship between person-environment fit and job satisfaction. This
relationship, though slight, appears to be moderated by type and by
the importance with which an individual views group interactions.
Further, evidence continues to mount that individuals will move
400 CAREER CHOICE AND DEVELOPMENT