Contextualising personality assessment : using meta-perceptions of the Big Five to improve the prediction of work performance
This thesis examined the effects of contextualising assessments of the Big Five
personality factors through the use of meta-perceptions. Meta-perceptions are
judgements about how an individual thinks others see them, and are therefore an
interpersonal approach to contextualising personality assessments. Eight research
questions were set, related to three areas. The first was the nature of meta-perceptive
ratings. Two questions were posed to examine patterns of high and low ratings among
sets of meta-perceptions. The second related to the prediction of work performance.
Three questions were identified which addressed the potential for meta-perceptions to
increase prediction over decontextualised and social role based ratings of the Big
Five, and the potential for meta-perceptions to identify negative- and non-linear
relationships between personality and performance. A fourth question was set in this
area regarding the design of a brief measure of the Big Five in order to concisely
measure meta-perceptions. The third area related to incongruence (differentiation)
across meta-perceptive ratings. Two questions were posed to identify the relationship
between incongruence and well being, and to explore the effects of incongruence for
different Big Five factors and across specific meta-perceptive rating pairs.
The thesis reports findings from seven studies, which address the research questions.
Study 1 reduced an existing Big Five instrument for subsequent use in Studies 2 and 3. Study 2 (N = 56) examined incongruence across meta-perceptive ratings of the Big
Five. Incongruence was calculated through factor analysis of the ratings, and was
found to correlate positively with perceived stress and neuroticism. Study 3 (N = 178)
reported systematic variation in mean Big Five scores across meta-perceptive ratings,
and further evidence for the negative relationship between incongruence and well
being. Study 4 reported findings from 21 expert judges' perceptions of the likelihood
that traits would be expressed in behaviour in 5 interpersonal interactions. It was
concluded that the Big Five are linked to particular `critical' interpersonal domains.
Study 5 (N = 79) applied polynomial regression analyses to data from Study 2, and
indicated that incongruence effects were not consistently predictive of negative well
being across the Big Five. Study 6 (N = 791) reported the design and validation of the
single-item measures of personality for use in concisely measuring meta-perceptions.
Study 7 examined the prediction of assessment centre competency performance from
meta-perceptive ratings of the Big Five, compared with decontextualised ratings in a
sample of 111 local government managers. Meta-perceptions were found to be more
predictive than the decontextualised assessments, and most predictive when rated
from the critical interpersonal perspectives identified in Study 4. In general discussion
of the findings, it was concluded that meta-perceptive ratings of the Big Five have
applied implications for both improving the predictive validity of personality
assessmentsa nd understanding the effects of incongruence on individual well being.