Personality testing in employment settings. Problems and issues in the application of typical selection practices
| Published date | 01 December 2001 |
| DOI | https://doi.org/10.1108/EUM0000000005978 |
| Pages | 657-676 |
| Date | 01 December 2001 |
| Author | Winfred Arthur,David J. Woehr,William G. Graziano |
| Subject Matter | HR & organizational behaviour |
Personality
testing in
employment
657
Personnel Review,
Vol. 30 No. 6, 2001, pp. 657-676.
#MCB University Press, 0048-3486
Received February 2000
Revised November 2000
Accepted November 2000
Personality testing in
employment settings
Problems and issues in the application of
typical selection practices
Winfred Arthur, Jr
Department of Psychology, Texas A&M University, College Station,
Texas, USA
David J. Woehr
Department of Management, The University of Tennessee, Knoxville,
Tennessee, USA, and
William G. Graziano
Department of Psychology, Texas A&M University, College Station,
Texas, USA
Keywords Employee selection, Individual behaviour, Personality tests
Abstract Complex issues arise when personality variables are incorporated into traditional
approaches to personnel selection. Personality assessment and testing in employment contexts is
more complicated than it would appear. Rather than arguing against considering personality
variables, we focus on five problematic issuesassociated with their use in personnel selection. These
issues are: the appropriateness of linear selection models; the problem of personality-related self-
selection effects; the multi-dimensionality of personality; bias associated with social desirability,
impression management, and faking in top-down selection models; and the legal implications of
personality assessmentin employment contexts. Recommends that practitioners and researchersbe
cognizant of these issues in the use of personality tests in employment decisions.
Personality is receiving renewed attention in selection and employment
contexts. A search of PsycINFO abstracts using ``job performance and
personality'' as keywords and limited to 1990-2000 identified 248 journal
articles and 127 dissertation abstracts for just the past ten years alone.
Taxonomic advances such as the emergence of the five-factor model (FFM) of
personality structure (Goldberg, 1993; John, 1990; McCrae and Costa, 1990; Ozer
and Reise, 1994; Wiggins and Trapnell, 1997) as well as meta-analysis based
validity evidence (e.g. Barrick and Mount, 1991; Tett et al., 1991) have played a
major role in this resurgent interest in the use of personality variables as
predictors of job performance. This renewed interest is further evidenced in the
many other primary studies (e.g. Arthur and Graziano, 1996; Digman and
Inouye, 1986; Graziano et al., 1996; Hogan et al., 1992; Mount et al., 1994; Nolan
et al., 1994) that have demonstrated relations among specified personality
variables and real world criteria of interest.
The research register for this journal is available at
http://www.mcbup.com/research_registers
The current issue and full text archive of this journal is available at
http://www.emerald-library.com/ft
An earlier version of this paper was presented at the 11th Annual Conference of the Society for
Industrial and Organizational Psychology in St Louis, Missouri, USA, April 1997.
Personnel
Review
30,6
658
However, personality assessment for personnel selection purposes carries
with it several potentially problematic issues. The present paper identifies and
discusses some of the issues that arise when the assessment of personality
variables is incorporated into traditional approaches to personnel selection. We
do not argue against the importance of considering personality variables
in employment contexts. Neither is it our intent to provide an exhaustive
review of the personality literature. Rather, we focus on problematic issues
with the conceptualization and use of personality variables as they apply to
personnel selection and human resource management (HRM) researchers and
practitioners. None of these issues are new, but they appear in different forms
within the psychometric, personality, and industrial/organizational (I/O)
psychology literatures (e.g. Block, 1995; Ozer and Reise, 1994). The present
paper assembles and presents these issues in a manner that makes them salient
and accessible to both researchers and practitioners. Five major issues are
identified and discussed. These interrelated issues are:
(1) the appropriateness of linear selection models;
(2) the problem of personality-related self-selection effects;
(3) the multi-dimensionality of personality and generation of ``composite''
personality scores;
(4) the detection of bias associated with social desirability, impression
management, and faking and the use of top-down selection models; and
(5) the legal implications of personality assessment in employment
contexts.
The appropriateness of linear models
Personnel selection is characterized by the use of linear models to represent the
relationship between predictors and criteria (Chaplin, 1997). These models
assume that extreme (usually higher) scores on the predictor are always more
desirable. This assumption may be useful for many variables involving
knowledge, skill, ability, and aptitude, but the assumption is particularly
problematic for personality variables. In fact, one can generate several
conceptually and theoretically sound scenarios where the relationship between
personality variables and job performance is better conceptualized as being
nonlinear. For instance, one can envisage a situation where moderate levels of
agreeableness may be related to effectiveness in customer relations, with low
and high levels of agreeableness, on the other hand, being somewhat counter-
productive (Graziano et al., 1996; Graziano and Eisenberg, 1997). Is it possible
to be too conscientious to perform certain roles effectively or to be too open-
minded to reach decisions for action? We think so. Murphy (1996, p. 22)
comments on this possibility when he notes that an individual who is high on
conscientiousness ``might be so conventional and rule-bound that he or she
cannot function in anything but the most bureaucratic setting''.
To continue reading
Request your trial