- Commensurate measurement is essential when testing P-E
fit models. - Fit indices are unfit.
- Present status models may work as well as P-E fit models.
- Hexagonal congruence and satisfaction do not correlate
significantly. - Measures of Holland’s model lack commensurability.
- Most hexagonal congruence indices are invalid.
- Holland’s hexagonal model lacks validity.
Tinsley offers a number of suggestions for improving theory and
research on P-E fit, including broadening theories, using more
repeated measures designs, and adding polynomial regression and
hit-rate analyses as analytic strategies. Tinsley’s critical treatise
argues strongly for the commensurate measurement of persons and
environments, as well as design and analysis strategies appropriate
to the complexity of the issues under study (Tinsley, 2000).
A second comprehensive review (Spokane et al., 2001) reasons
that correlational studies, although crucial at the early stages of
congruence research, do not reflect the complexity of P-E theory as
well as do longitudinal, interactional, and experimental studies.
Conclusions based on correlational studies must, therefore, be in-
conclusive by definition. This review concludes that congruence is
a sufficient but not necessary condition for vocational satisfaction
and identifies six benchmark studies of congruence with appropri-
ate research designs that incorporate
- Experimental and social process elements (Helms, 1996)
- Longitudinal analyses (Elton & Smart, 1988)
- Multidimensional congruence (Melamed, Meir, & Samson,
1995) - Moderator variables (Meir, Keinan, & Segal, 1986)
- Qualitative data (Blustein et al., 1997)
HOLLAND’S THEORY 397