Page 186 - Psychological Management of Individual Performance

methodological issues in the measurement of potential 169
e.g., from a subject matter expert type job to an executive type job. This difference with
regard to the selection procedures makes potential analysis more difficult because the
relation between present and future job content is not so close, and less certain because of
a lack of reliable predictors. Further, there is a prognostic uncertainty that the predicted
and achievable development of a person will actually occur. Whether predicted and in
fact existing potential is actually transferred into performance primarily depends on the
motivation of the individual. Additionally, if the predicted development is communicated
openly within the organization (i.e., to the assessee and the immediate superior), this
communication can provoke a self-fulfilling prophecy (Becker, 1992) in stabilizing and,
furthermore, enhancing motivational factors.
Prediction based on measurement procedures (which apply also to the measurement
of characteristics) is dependent on the reliability of the chosen method. As Schmidt
and Hunter (1998a) point out, measurements are never perfectly reliable but show a
measurement error. Because this inaccuracy holds also for predictions with regard to
personal features it decreases the possibility to predict future performance correctly.
The authors therefore recommend the careful selection of those predictors of future
job-related performance that can be measured more reliably.

RELIABILITY AND VALIDITY OF JUDGES’RATINGS

The quality of potential analyses strongly depends on the sources of assessment, i.e.,
on the raters who render the judgment of an individual’s potential. In the context of
performance appraisals Harris and Schaubroek (1988) found out that ratings by different
sources correspond only moderately. The major reasons for this are the differences in
perspectives on and knowledge of the incumbent’s job. Therefore, the judges who assess
the potential of employees should represent different stakeholders, i.e., different com-
pany functions and various levels of hierarchy. Thus, a higher amount of experience is
employed to identify the required qualifications and to evaluate the individual’s potential
(Neuberger, 1980b). Further, the involvement of different kinds of judges is favorable
as rater errors and effects of similarity and likability (Cardy & Dobbins, 1986; Schmitt,
Pulakos, Nason, & Whitney, 1996) tend to neutralize each other. Moreover, the involve-
ment of different raters might cause judges not to render over-optimistic predictions
(see Reece & Matthews, 1993). Campbell (1999) identified accountability of judges as a
critical aspect. Judges should justify their assessments of specific individuals in order to
establish greater accuracy and fairness (Mero & Motowidlo, 1995; Motowidlo & Schmit,
1999). In any case, Woehr and Huffcutt (1994) suggest that the training of judges might
be helpful in order to promote a common frame of reference.
The most important source are immediate superiors as they know best the employee’s
work performance and are able to rate the ability, knowledge, and technical proficiency
(Borman, White, & Dorsey, 1995). However, Klimoski and Brickner (1987) argue that
the judgment of the immediate supervisor might be biased in a potential analysis pro-
cedure. They fear that instead of assessing the performance the ratee shows during that
procedure, the judgment will reflect the supervisor’s opinion of the ratee’s daily per-
formance and thereby bypass the criteria of the potential analysis procedure. Moreover,
the capability of the superior to assess the actual work performance of an employee
is not necessarily in accordance with the requirements to correctly predict the future
achievement of the ratee.