Page 189 - Psychological Management of Individual Performance

172 analysis of performance potential
intelligence tests with assessment center (.52) hardly helps to increase the validity that
intelligence tests (.51) have when applied solely. These results show that the inclusion of
motivational aspects (as are raised in tests of integrity and of conscientiousness) improve
the prediction of performance.
In spite of these empirically distinct findings, intelligence tests in general are hardly ac-
cepted as instruments for potential analysis in Europe and the United States. A European
study by Schuler, Frier, and Kauffmann (1993) as well as an internationally conducted
study by Ryan, McFarland, Baron, and Page (1999), in which 959 companies within
20 world-wide countries were investigated, revealed that intelligence tests are not of-
ten used. In Germany, only 2% of the investigated companies apply intelligence tests
in manager selection. As both studies show, Spain, Belgium, and the Netherlands, for
example, make use of intelligence tests more often.
One of the most common methods to measure potential performance is the assess-
ment center (AC) (Baumann-Lorch & Lotz, 1996; Baumann-Lorch, Millermann, & Lotz,
1994; Huck & Bray, 1976; Jochmann, 1989; Klein & Scheffler-Lipp, 1989; Kliem, 1987;
Lowry, 1994; Sarges, 1996; Shore, 1992; Shore, Shore, & Thornton, 1992; Shore et al.,
1998; Tziner & Dolan, 1982). ACs are used for the prediction of performance in both con-
texts, personnel selection and human resources development. As the procedure is costly
(Schmidt & Hunter, 1998a), ACs are mainly used for potential assessment of higher
management levels (for more practical details see the chapter of Altink and Verhagen,
this volume).
Of course, the validity of ACs depends on their methodological standard. A study by
Schmidt and Hunter (1998b) shows that the predictive validity of AC is reasonably good.
Lievens (1998) and Kleinmann (1993, 1997a, 1997b) point out conditions under which
predictive and construct validity can be increased.
However, two major objections are raised against the use of classical ACs. First, most
of the traditional assessment centers measure status quo performance rather than perfor-
mance potential. That means, the measured performance is influenced to a large extent
by the experience of the candidate. For example, if candidates are not supervising any
other employees, their knowledge of management and supervision will be dependent
on their experience with the behavior of their direct supervisor. And if these candidates
are asked to act as a supervisor in a role play during a classical AC, the probability is
high that their behavior resembles the behavior of the direct supervisor. If we further
assume that this supervisor is not very skillful, the management potential of the candi-
dates in question will probably be judged to be rather poor. Thus, on the basis of status
quo performance evaluations it is impossible to draw valid conclusions concerning the
learning potential of a candidate in this performance domain (Obermann, 1996). In order
to omit these shortcomings, several authors (e.g., Kleinmann, 1996; Obermann, 1996;
Sarges, 1996) suggest the use of dynamic ACs that allow the candidate to show results of
learning processes during the AC. The second major objection against classical ACs is
that future job requirements are usually not known in such detail that the qualifications
of candidates could be measured appropriately. There are of course exceptions where
the position the candidate should fulfill is already known. Several researchers (e.g.,
Obermann, 1996; Sarges, 1996) argue that in times of a rapidly changing environment
and the increased dynamic of technological and market developments, the capability of
an individual to cope with these changes, to adapt to them and to structure them unerr-
ingly is most important. According to Sarges (1996) this capability is best represented