Page 188 - Psychological Management of Individual Performance

methods of potential analysis 171
METHODS OF POTENTIAL ANALYSIS

Research over the past 85 years has tried to identify valid predictors for future job-related
performance. Studies vary in the specificity of predictors (Cantwell, 1990: gender, needs
and motivation; Gadzella, Ginther, & Bryant, 1997: learning style and critical thinking;
Hojat, Vogel, Zeleznik, & Borenstein, 1988: psychological predictors such as stressful
life events, test anxiety, etc.; Love & O’Hara, 1987: work maturity; Locke, Frederick,
Lee, & Bobko, 1984; self-efficacy, goals, and task strategies; Stajkovic & Luthans, 1998:
self-efficacy) as well as performance criteria (subjective, such as grade point average,
Pringle, 1994; objective, such as job status, Martin & Bartol, 1985) and they usually
hold only for specific groups of professionals.
Further, attempts to measure potential and to predict performance differ in the methods
that are used. They range from appraisal interviews (Miner, 1970; Shahani, Dipboye, &
Gehrlein, 1991), cognitive tests (Gordon, Charns, & Sherman, 1987; Shore et al., 1998),
work sample tests (Mount, Muchinsky, & Hanser, 1977), and in-basket exercises
(Hakstian, Woolsey, & Schroeder, 1986) to personality factors (Cook & Emler, 1999;
Judge, Erez, & Bono, 1998; Linnehan, 1998), biodata (James, Ellison, Fox, & Taylor,
1974; Mael & Hirsch, 1993; McBride, Mendoza, & Carraher, 1997) and combinations
of different methods (Jones, Joy, & Martin, 1990; Mayberry & Carey, 1997; P¨ossnecker,
1992) (for practical issues see the chapter of Altink and Verhagen in this volume). For
each predictor there are a number of meta-analyses: Hunter and Hunter (1984), Hunter
and Hirsh (1987), Gaugler, Rosenthal, Thornton, and Bentson (1987), Tett, Jackson, and
Rothstein (1991), Robertson and Kinder (1993), Bliesener (1996), Salgado (1997). The
extensive findings by Schmidt and Hunter (1998a) may serve to exemplify the research
data (see Table 8.1).Their meta-analysis revealed that, with the exception of work sample
tests (validity of .54), tests of general mental ability (GMA) have the highest validity.
This predictor has a validity of .58 for job performance of managers and a validity of .51
for jobs of average demands. Assessment center show a validity of .37, biographical data
measures have .35 validity, personality tests as well as conscientiousness tests have a va-
lidity of .31, and integrity tests have one of .41. Structured employment interviews have
a validity of .51. The combination of two predictors shows that integrity tests in combi-
nation with intelligence tests make the best prediction (.65) whereas the combination of

TABLE 8.1 Predictive validity for overall job performance of general mental ability (GMA)
scores combined with a second predictor using (standardized) multiple regression (Schmidt &
Hunter, 1998b, p. 265)
Gain in validity
from adding % increase in
Personnel measures Validity (r) Multiple R supplement validity
GMA tests .51
Work sample tests .54 .63 .12 24%
Integrity tests .41 .65 .14 27%
Conscientiousness tests .31 .60 .09 18%
Employment interviews
(structured) .51 .63 .12 24%
Biographical data measures .35 .52 .01 2%
Assessment centers .37 .53 .02 4%