ArticlePDF Available

Predictive Validity toward Job Performance of General and Specific Mental Abilities. A Validity Study across Different Occupational Groups

Authors:

Abstract and Figures

There are two main views on the role of cognitive abilities in job performance prediction. The first approach is based on meta-analysis and incremental validity analysis research and the main assumption is that general mental ability (GMA) is the best job performance predictor regardless of the occupation. The second approach, referred to as specific validity theory, assumes that job-unique weighting of different specific mental abilities (SMA) is a better predictor of job performance than GMA and occupational context cannot be ignored when job performance is predicted. The validity study of both GMA and SMA as predictors of job performance across different occupational groups (N = 4033, k = 15) was conducted. The results were analyzed by calculating observed validity coefficients and with the use of the incremental validity and the relative importance analysis. The results supports the specific validity theory – SMA proved to be a valid job performance predictor and occupational context moderated GMA validity.
Content may be subject to copyright.
Business and Management Studies
Vol. 4, No. 3; September 2018
ISSN: 2374-5916 E-ISSN: 2374-5924
Published by Redfame Publishing
URL: http://bms.redfame.com
1
Predictive Validity toward Job Performance of General and Specific
Mental Abilities. A Validity Study across Different Occupational Groups
Jaroslaw Grobelny
Correspondence: Jaroslaw Grobelny, Institute of Psychology, Adam Mickiewicz University in Poznan, Poland
Received: May 24, 2018 Accepted: June 20, 2018 Online Published: July 4, 2018
doi:10.11114/bms.v4i3.3297 URL: https://doi.org/10.11114/bms.v4i3.3297
Abstract
There are two main views on the role of cognitive abilities in job performance prediction. The first approach is based on
meta-analysis and incremental validity analysis research and the main assumption is that general mental ability (GMA)
is the best job performance predictor regardless of the occupation. The second approach, referred to as specific validity
theory, assumes that job-unique weighting of different specific mental abilities (SMA) is a better predictor of job
performance than GMA and occupational context cannot be ignored when job performance is predicted. The validity
study of both GMA and SMA as predictors of job performance across different occupational groups (N = 4033, k = 15)
was conducted. The results were analyzed by calculating observed validity coefficients and with the use of the
incremental validity and the relative importance analysis. The results supports the specific validity theory SMA
proved to be a valid job performance predictor and occupational context moderated GMA validity.
Keywords: job performance, general mental ability, specific mental abilities, validity study
1. Introduction
Cognitive abilities are one of the most-often discussed predictors of job performance. Despite decades of research, some
uncertainty can still be found in the field. This study addresses this issue with a view to examining the relative validity
of both general (GMA) and specific mental abilities (SMA) toward job performance in a series of occupational groups.
This paper contributes to the literature dedicated to personnel selection by providing empirical data on the relative
validity of GMA and SMA for each of the featured occupational groups. The distinctive feature of this study is that it
simultaneously investigates both of these predictors in a systematic way, applying both ratings and quantitative criteria
of job performance and the varied statistical method, including the recently widespread and robust relative importance
analysis method.
1.1 Theoretical Perspectives
As Lang et al. (2010) pointed out, the relative importance of GMA and SMA is a “longstanding question in personnel
psychology” (p. 595-596). Two main viewpoints can be found in relevant literature. One of them can be called the
mainstream or unitarian perspective, while the other arose from the specific aptitude theory (Stanhope & Surface,
2014).
1.1.1 General Cognitive Ability
The main claim in the former theory is that GMA is the best predictor of job performance and so much evidence has
been collected in this field that it can no longer be debatable (Schmidt, 2002). Also, single result for GMA will always
be superior job performance predictor over two or more specific factors combined (Schmidt, 2002). Furthermore, the
proponents of this perspective argue that the only reason that SMA tests have validity is the loading of the g factor
(Carretta & Ree, 2000) and therefore the predictive validity comes solely from GMA. Finally, no validity in predicting
job performance would be gained from SMA and even if it were to be noted, it would be infinitesimal (Olea & Ree,
1994; Viswesvaran & Ones, 2002). Statements such as this are mostly based on a variety of meta-analysis studies,
which show the validity of GMA tests and studies applying incremental validity analysis based on hierarchical
regression models (e.g. Bertua, Anderson, & Salgado, 2005; Carretta & Ree, 1997; Hirsh, Northrop, & Schmidt, 1986;
Hunter & Hunter, 1984; Salgado et al., 2003; Schmidt & Hunter, 1998).
Business and Management Studies Vol. 4, No. 3; 2018
2
1.1.2 Specific Validity Theory
The second perspective, variously referred to as the differential aptitude theory, differential validity theory, multiple
aptitude theory, and specific validity theory (hereinafter referred to as SVT) (Lang et al., 2010; Richardson & Norgate,
2015; Stanhope & Surface, 2014) postulates that “job-unique weighting of several specific aptitudes (…) is greatly
superior to intelligence alone in predicting (…) performance” (Ree & Earles, 1992, p. 87). Broadly speaking, the
assumption here is that if prediction is supposed to be valid, it should be made for a specific context (e.g. a chosen
occupation) and based on specifically chosen predictors (which fit that context). These theories are considered by some
to have been disproved (see Schmidt, 2002), while others believe that the dispute about the impact of GMA and SMA
on performance has not been resolved. There are authors who have claimed that a composite combined out of SMA, i.e.
narrow cognitive abilities (or a job-specifically weighted composite) would be superior in terms of job performance
prediction to GMA (see Krumm, Schmidt-Atzert, & Lipnevich, 2014; La Grange & Roodt, 2007; Lang et al., 2010;
Stanhope & Surface, 2014).
1.2 Review
The mainstream view of the superior role of GMA in job performance prediction puts forth three major claims. These
are: 1) GMA is the single most valid predictor of job performance 2) GMA is valid regardless of the occupational
context 3) no incremental validity should be expected from SMA (Schmidt, 2002). It is reasonable to present and
discuss the matter thoroughly in order to justify interest in this topic, as there are a number of doubts regarding each of
the above statements.
1.2.1 Concerns about the Superior Role of GMA in Job Performance Prediction
The claim of the best predictive validity magnitude of GMA comes primarily from meta-analytic studies. These studies
summarized the results of the validity of different tests from hundreds of samples. Overall, they have shown validity
coefficients c.a. 0.3 0.5 (Bertua et al., 2005; Schmidt & Hunter, 1998). In most meta-analyses, one major coefficient
(or a few, at most) is reported, and it summarizes studies conducted in a variety of conditions and in many, sometimes
vastly different, contexts. As a consequence, an important amount of information may be lost because of such
generalization. Moreover, there is considerable uncertainty regarding the actual impact of meta-analytic correction
procedures on the estimated outcomes (Lakens, Hilgard, & Staaks, 2016; LeBreton, Scherer, & James, 2014;
Richardson & Norgate, 2015).
Given the important uncertainty presented above, if one decides to investigate the primary studies related to the topic,
some puzzling results could also be found. First, there is a major part of studies that presents no validity of GMA at all
in predicting job performance variance. Hirsh et al. (1986) reported no validity of GMA for law enforcement
occupations. Hogan, Hogan & Gregory (1992) found no relationship between GMA and measures of job performance
for salespeople. In addition, Verbeke et al. (2008) provided an extensive list of studies in which GMA were not related
to the performance of sales staff. There were no connections, either, between GMA tests scores and quantitative
measures in such occupations as recruitment consultants, bankers, insurance representative, and transit operators
(Barros, Kausel, Cuadra, & Díaz, 2014; Downey, Lee, & Stough, 2011; Hausdorf & Risavy, 2015; La Grange & Roodt,
2007). Secondly, numerous examples of SMA that proved to be a valid job performance predictor (more than GMA) in
certain context can be found as well; these are: mechanical comprehension and mechanical reasoning for manufacturing
employees (Muchinsky, 1993), performance correctness (Thomas, Barrett, & Alexander, 1996) and performance speed
for clerks (Whetzel et al., 2011) or perceptual speed for warehouse workers (Mount, Oh, & Burns, 2008). Schmidt
(2002) cited examples of nine validation studies where the comparison between GMA and SMA could be made in nine
occupations. In six cases, SMA had a superior predictive magnitude over GMA. Marcus, Johnston & Rothstein (2007)
investigated managers in the forestry sector and found that composites based on several SMA had a superior predictive
validity over GMA. Finally, as Krumm et al. (2014) mentioned, more recent evidence from meta-analytic studies shows
that SMA could be not only an additional, but even a more important predictor of performance than GMA.
The above study results contradict the mainstream viewpoints of the principal role of GMA in job performance
prediction. This may, at first glance, seem a “cherry picking” practice, as primary study results are always more prone to
error than meta-analytic reviews. It is however far from the truth, as thorough analysis lists a series of factors that may
moderate the relation between both GMA and SMA with job performance. Firstly, when SMA results are computed in
alignment with a given criterion and estimated specifically for the occupation group examined, it could account for a
great deal of variance in job performance (Ree & Earles, 1992; Reeve, 2004; Schneider & Newman, 2015). Next, GMA
appears to lose validity towards job performance while other characteristics, e.g. social skills, are being considered
simultaneously (Cote & Miners, 2006; Schneider & Newman, 2015). Furthermore, while more robust measures of job
performance than supervisory ratings are used, GMA appeared to be a weak or an insignificant predictor (La Grange &
Business and Management Studies Vol. 4, No. 3; 2018
3
Roodt, 2007; Vinchur, Schippmann, Switzer, & Roth, 1998). Finally, the occupational context itself seems to be a major
moderator of the validity magnitude for the predictors discussed.
1.2.2 Concerns about the Universal Validity of GMA in Performance Prediction in a Variety of Occupations
As stated above, the validity of both GMA and SMA seems to vary greatly between different occupations. Ree and
colleagues (Olea & Ree, 1994; Ree, Earles, & Teachout, 1994) found predictive validity rates of GMA from 0.26 for air
traffic operators and up to 0.71 for laboratory specialists or mechanics. In the study of Bertua et al. (2005), the
coefficients range was also broad, from 0.14 and 0.16 for clerks and drivers, up to 0.33 for engineers. Similarly, in
Salgado et al. (2003), they were from 0.12 - 0.20 for police and drivers, up to 0.31 - 0.34 for clerks and sales staff. It
appears that the magnitude of the criterion-related validity of GMA for different occupational groups varied from
moderately weak to exceptionally strong. Unfortunately, this differentiation is often omitted in meta-analytic reports, as
authors often emphasize few general coefficients estimation and neglect to reports the contextual information or
detailed and systematical subgroups comparison. For that reason, sometimes these actual differences in predictive
validity of GMA become indistinguishable.
1.2.3 Concerns about the Lack of Incremental Validity from SMA
Incremental validity of SMA in job performance prediction, based on Ree and colleagues' work (Carretta & Ree, 2000;
Olea & Ree, 1994; Ree & Earles, 1992; Ree et al., 1994), is considered to be almost insignificant. Once again, this
might be due to the context considered; as Schneider & Newman (2015) noticed, this is the case only when all
occupations that occurred in a sample (e.g. over 80) are considered together.
There may be however methodological reasons to question the evidence regarding the lack of value of SMA (Reeve,
2004). There is an issue with the incremental validity analysis method itself, which is a major method used in primary
validation studies (Stanhope & Surface, 2014). By using hierarchical regression, one can compare two models one
with only an initial predictor (GMA in this case) and a second one with an initial and an additional predictor (e.g. SMA).
If there is an increase in the coefficient of determination between the models, one could state that the additional
predictor has incremental validity. The issue is that, when predictors are included in the regression, in this approach the
shared variance between initial and additional predictors and dependent variables is accounted for in the initial predictor
only (Lang et al., 2010; Stanhope & Surface, 2014).
According to Lang et al. (2010), “as shared variance between GMA (…) and the narrower cognitive ability measures
belongs to either GMA or narrower cognitive ability constructs (…), incremental validity analysis does not correspond
to the model’s assumptions” (p. 604). This statement is based on nested factor models (NFM) of intelligence, which
claim that variance in ability tests is explained by both GMA and SMA, contrary to theories originating in the
Spearmanian tradition, with its assumption of the causal effect from GMA to SMA. As NFT seems to be more accepted
and supported by data (see Lang et al., 2010; Richardson & Norgate, 2015; Verbeke et al., 2008), there is an issue with
the acceptance of results based on incremental validity analysis with GMA as an initial predictor; caution should be
recommended at least.
The fact is that the predictive validity of SMA may be considered at the same time to be insignificant, little or highly
significant (based on the same dataset), while the only moderator will be the statistical method used (Lang et al., 2010;
Reeve, 2004). Researchers consequently need a statistical technique capable of determining the explained variance
belonging to each ability construct even when the measures of the constructs are considerably correlated. A technique of
this kind is relative importance analysis (RIA). As Johnson and LeBreton (2004) described, it allows one to estimate
“the proportionate contribution each predictor makes to R2, considering both its direct effect (e.g. its correlation with the
criterion) and its effect, when combined with the other variables in the regression equation” (p. 240). This method
seems to be much more useful in determining the predictive validity magnitude of SMA and GMA in validation studies
and it seems reasonable to at least employ both hierarchical regression models and RIA to compare their results in
determining the relative validity of GMA and SMA, as substantial doubts surrounding this matter could be identified.
1.3 Hypothesis Development
The empirical reasons for the consideration of the relative validity of GMA and SMA are sufficient. However, there are
also strong theoretical grounds to undermine the claims of the superior role of GMA in job performance prediction. A
well-established theory of individual differences in job performance by Motowidlo, Borman and Schmidt (1997) will
serve as a general framework. Based on this theory, one could expect both a relatively better validity of SMA than GMA
and a significant differentiation of abilities validity in different organizational contexts.
The key assumption in this theory is that the variance in job performance is caused by a variability in characteristic
adaptations, which are the specific skills and patterns of employees’ behaviors. Characteristic adaptations are, in turn,
the results of interactions between the basic tendencies (individual differences in personality and abilities) and learning
Business and Management Studies Vol. 4, No. 3; 2018
4
experience derived from the environment. Basically, characteristic adaptations are implementation of behaviors required
in one’s job in a faultless and easy manner. Execution of some specific behavior or task is, in fact, enhanced by a certain,
narrow ability (Sternberg, 2001). As tasks and behaviors required in a given occupation tend to be similar (within a
certain group), they should be enhanced by a sole SMA (or few at best). Concurrently, as GMA overly enhance general
problem solving and functioning, they should be responsible for a variety of behaviors, including those insignificant
within a given occupational context. Therefore, their validity would be on average important, however less substantial
than the one of SMA. Subsequently, actions required in a given occupation differed greatly (between groups) and
therefore a characteristic adaptation needed to perform well should vary too. This led to the logical assumption that
basic tendencies required in a given occupation must be different. As a result, the predictive validity of the test of these
abilities would vary for different occupational groups. This theoretical explanation, together with empirical evidence
discussed above, led to the following two hypothesis:
H1. A validity of SMA toward job performance within a given occupational group is relatively better than GMA.
H2. An occupational group moderates the validity of GMA toward job performance.
2. Method
2.1 Participants’ Characteristics
The participants of this study were Polish candidates (N = 4033) for positions from fifteen different occupational groups
(board of directors, buyers, clerks, constructors, consultants, customer service, financiers, HR, IT specialists,
manufactory workers, marketing specialists, QA’s, researchers, salespeople, and transition specialists). The study
included 2,237 (55.47%) women and 1,796 (44.53%) men. As many as 1,446 (35.85%) were in managerial positions.
Participants’ average age was 31 (SD = 6.57) and they were mostly moderately experienced (M = 7.55; SD = 6.47). The
high dispersion suggests that a broad scope of experience levels was represented in the sample. Participants were
assigned to occupational groups based on their most recent job position.
2.2 Measurements
During the study, the job performance was measured by two measures (both rating and quantitative), while cognitive
abilities were measured by a series of on-line cognitive tests. Besides, control variables such as age, overall professional
experience and occupational experience (within a given occupational group) were measured.
2.2.1 Job Performance (Quantitative)
The first variable that operationalized a job performance was the financial outcomes of participants. The mean salary
from the preceding three months was reported by the participants. The assumption here is that better-performing
employees are paid better and they are employed by the best-paying companies or at higher positions. The salary is
moreover often strongly dependent on the work outcome, as the pay for performance system is a common practice, in
particular in certain occupations. To avoid the impact of such confounding factors as wage disparities between
occupations, the data were scaled within the analyzed groups before the computations. As the salary is considered to be
sensitive information, it was optional for participants to provide this information (to avoid measurement unreliability
resulting from false data), yet only cases with no missing data were included into study.
2.2.2 Job Performance (Rating)
The second criterion of job performance was self-reported ratings of the average level of job goals completed (e.g. KPI
or sales targets). The participants were asked to rate to what degree they realize their professional goals, either formal
(e.g. included in management-by-objective system) or informal (e.g. set by a supervisor and communicated in an
informal manner). The rating was performed with the aid of a 6-point scale, where 1 means achieving less than 60% of
the goals and 6 means the accomplishment of over 100%.
2.2.3 Ability Tests
To measure cognitive abilities, an online test dedicated for personnel selection was used. The test was developed in
accordance with strict psychometrical procedures (and validated on a group of 5,572 participants). As a result, the tool
met the required standards for psychological tests: it was reliable, as the minimum test-retest coefficients for a 1-month
period was 0.89 for each subtest and valid, e.g. selected scales correlated highly with expected scales from WAIS-R and
CFT-3 and the factor analysis confirms an expected, hierarchical structure of the results. The tests were developed on
the basis of the CHC theory (see Schneider & Newman, 2015) and therefore each test was designed to measure a chosen
narrow ability. Each test consisted of 20 tasks and was time-limited. There were seven tests:
Quantitative reasoning. The ability to perform tasks based on calculation and mathematical operations.
Visualization. The ability to perform tasks that require spatial imagination as well as being able to find and predict
Business and Management Studies Vol. 4, No. 3; 2018
5
analogies in given systems of spatial elements.
Reading comprehension. The ability to interpret information and drawing proper conclusions based on various read
information.
Deductive reasoning. The ability to apply patterns of logical reasoning as well as drawing proper conclusions based on
incomplete information and predefined rules for reasoning.
Lexical knowledge. Verbal ability consisting of a broad vocabulary and an ability to connect elements into basic
relations on the basis of knowledge of their meaning.
Inductive reasoning. The ability to recognize relationships between given constructs and predict fitting elements based
on rules that need to be identified.
Word fluency. The ability to quickly generate words based on a given criterion.
On the basis of the results of these tests, two further variables were computed.
2.2.4 GMA
With the help of the Ordinary Least Squares (OLS) method of factor analysis (to find the minimum residual solution), a
single general factor was calculated and extracted. This variable corresponds to the general factor from stratum III in
CHC theory and operationalized GMA as well. As different methods of estimating g tend to show a high stability (see
Ree & Earles, 1991), no further general factors were calculated.
2.2.5 SMA
This composite was calculated according to the SVT, which means that each narrow ability test result was weighted
(based on the validity coefficient for job performance criterion) specifically for a chosen occupational group (Reeve,
2004; Stanhope & Surface, 2014). These composite scores reflect the specifically weighted choice of SMA, which are
valid only for a certain occupation and which have been designed for a given group.
2.3 Research Procedure
The data for this study were gathered during actual recruitment processes conducted by a staffing agency. The
candidates were able to identify job positions in online job boarding services and then fill out applications on a
dedicated platform. Each of the candidates who replied to job ads could create an account with their demographic data,
where the criterion measures and abilities tests were conducted. The procedure was administered online and the
candidates completed tests in their own time and place; however, they were informed that their scores will be verified
by the agency during further stages of the recruitment procedure (this regards both criterion measures and ability tests
results). The participants were provided with detailed information about the procedures and tests, so they could make
their own decisions whether they would participate in the study and pass the data to the researcher.
3. Results
The results of the measures along with a correlation table are shown in Table 1. The validity of every measured
predictor towards both job performance criteria is presented in Table 2. Based on these results, three further analyses
were performed: hierarchical regression models, moderation analysis and relative importance analysis.
3.1 Observed Validity
First of all, all predictors appeared to have a comparable correlation in the case of both job performance criteria and
similar patterns of relationship presented. The observed validity coefficient for the whole sample shows that GMA has
moderate validity. Several SMA tests, e.g. quantitative reasoning, reading comprehension and verbal potential, proved
to be comparable or slightly weaker predictors, while the rest had in general an unsatisfactory correlation with both job
performance criteria. However, the results differ substantially when occupational validity coefficients were analyzed. In
one group of consultants both SMA and GMA had no validity at all. In the case of clerks, financiers and QA’s, GMA
had no validity toward both job performance criteria. There were seven groups in which GMA had the strongest
correlation with both job performance criteria measured among all single predictors considered.
However, when occupational groups were analyzed, it appeared that there was no single group where GMA would be a
significant and a more valid predictor that the SMA composite. Furthermore, there were seven groups in which a single
narrow ability test could be identified as the most valid predictor. These data are in favor of both hypotheses, as SMA
proved to correlate stronger with job performance measures than GMA did, and the GMA validity differed between
groups.
Business and Management Studies Vol. 4, No. 3; 2018
6
Table 1. Summary of study results
mean
sd
1
2
3
6
7
9
10
12
13
1. age
30.93
6.57
2. experience (overall)
7.55
6.47
0.91
3. experience (occupational)
5.18
5.29
0.72
0.76
4. job performance (financial
outcome)
4 236
2
671
0.38
0.39
0.35
5. job performance (self-rating)
3.78
0.88
0.13
0.14
0.12
6. QR
16.80
3.79
0.02
0.04
0.04
7. V
16.86
3.00
-0.0
3
-0.0
3
-0.0
4
0.3
7
8. RC
15.47
3.08
-0.0
3
-0.0
3
-0.0
2
0.5
2
0.3
6
9. DR
14.58
3.74
-0.0
3
-0.0
6
-0.0
2
0.2
9
0.2
2
10. LK
17.25
2.69
0.09
0.09
0.06
0.2
4
0.1
8
0.2
0
11. IR
19.79
2.26
-0.0
9
-0.0
9
-0.0
9
0.3
3
0.2
6
0.2
5
0.2
7
12. WF
23.43
5.20
-0.1
2
-0.1
3
-0.1
1
0.2
4
0.2
2
0.1
9
0.1
9
13. SMA
-0.03
0.89
0.05
0.08
0.06
0.5
6
0.4
2
0.4
2
0.4
2
0.2
4
14. GMA
0.00
1.15
-0.0
1
-0.0
3
-0.0
3
0.7
9
0.6
0
0.5
1
0.4
6
0.4
5
0.7
0
Note. QR = quantitative reasoning; V = visualization; RC = reading comprehension; DR = deductive reasoning; LK =
lexical knowledge; IR = inductive reasoning; WF = word fluency.
Table 2. Observed validity coefficients
department
n
QR
V
RC
DR
LK
ID
WF
GMA
SMA
board
102
0.47*** /
0.47***
0.45*** /
0.44***
0.29* /
0.34**
0.34** /
0.34**
0.25* /
0.19
0.30* /
0.40**
0.01 /
0.22*
0.51*** /
0.57***
0.58*** /
0.59***
buyers
45
0.37* /
0.43***
-0.13 /
0.07
0.46** /
0.52***
0.39** /
0.28
0.2 /
0.39**
0.02 /
0.18
0.12 /
-0.08
0.4** /
0.5***
0.52*** /
0.6***
clerks
589
0.05 /
0.03
0.06 /
-0.02
0.02 /
0.02
0.01 /
0.01
0.10* /
0.06
-0.06 /
-0.02
-0.01 /
0.08
0.04 /
0.03
0.10* /
0.01
constructors
200
0.22* /
0.31**
0.26** /
0.25**
0.21* /
0.32**
0.05 /
0.22*
0.09* /
0.25*
0.24* /
0.32**
0.24* /
0.36**
0.29** /
-0.44***
0.34** /
0.46***
consultants
78
0.20 /
0.13
0.06 /
-0.01
0.11 /
0.06
-0.09 /
-0.07
-0.07 /
-0.02
-0.2 /
-0.14
-0.01 /
0.01
0.07 /
0.03
0.02 /
0.01
customer
service
425
0.38** /
0.42***
0.20** /
0.23**
0.26** /
0.33**
0.16* /
0.22*
0.11* /
0.18*
0.17* /
0.31**
0.05 /
0.20**
0.36** /
0.47***
0.38** /
0.48***
financiers
415
0.13* /
0.03
0.05 /
0.04
0.04 /
-0.01
0.16* /
0.21**
0.10* /
0.11*
-0.02 /
0.01
0.11* /
0.03
0.09 /
0.06
0.24* /
0.23*
HR
170
0.01 /
-0.06
-0.14 /
-0.07
-0.10 /
-0.12
-0.12 /
-0.08
-0.02 /
-0.14
0.21* /
0.22*
0.16* /
0.20*
0.15* /
0.19*
0.23* /
0.27*
IT
253
0.46*** /
0.53***
0.22* /
0.29**
0.41*** /
0.46***
0.37** /
0.47***
0.33** /
0.41**
0.18* /
0.31**
0.13* /
0.18*
0.51*** /
0.63***
0.54*** /
0.65***
manufactory
workers
227
0.26* /
0.41***
0.15* /
0.27*
0.24* /
0.38**
0.22* /
0.29*
0.12 /
0.25*
0.02 /
0.20*
0.13* /
0.11*
0.24* /
0.46***
0.3** /
0.47***
marketing
specialists
215
0.34** /
0.39**
0.20* /
0.29**
0.42*** /
0.41***
0.22* /
0.31**
0.25* /
0.27**
0.20* /
0.24*
0.16* /
0.23*
0.46*** /
0.53***
0.46*** /
0.53***
QA’s
34
-0.11 /
0.19
0.40** /
0.23
0.29 /
0.09
0.08 /
0.27
-0.03 /
0.18
0.01 /
0.23
0.16 /
0.40**
-0.11 /
0.33
0.40** /
0.40**
researchers
39
0.64*** /
0.69***
0.50*** /
0.48**
0.18 /
0.25**
0.57*** /
0.51***
0.31** /
0.34*
0.46** /
0.38*
0.05 /
0.10
0.68*** /
0.70***
0.83*** /
0.79***
salespeople
966
0.41*** /
0.46***
0.25* /
0.36**
0.34** /
0.45***
0.21* /
0.30**
0.30** /
0.35**
0.15* /
0.33**
0.09** /
0.24**
0.43*** /
0.59***
0.46*** /
0.60***
transition
specialists
275
0.20* /
0.29**
0.07 /
0.14*
0.25* /
0.29**
0.22* /
0.26**
0.28** /
0.28**
0.25* /
0.28**
0.10 /
0.16*
0.33** /
0.42***
0.39** /
0.43***
all
4033
0.31** /
0.31**
0.17* /
0.21*
0.25** /
0.27**
0.18* /
0.23*
0.20* /
0.22*
0.10* /
0.19*
0.01 /
0.12*
0.31** /
0.37***
0.40*** /
0.51***
Business and Management Studies Vol. 4, No. 3; 2018
7
Note. The first number is the coefficient for quantitative measure, the second for rating; QR = quantitative reasoning; V
= visualization; RC = reading comprehension; DR = deductive reasoning; LK = lexical knowledge; IR = inductive
reasoning; WF = word fluency.
* p < 0.05; ** p < 0.01; ***p < 0.001.
3.2 Hierarchical Regression Analysis
Next, hierarchical regression models were built and compared (Table 3) and incremental validity analysis was
performed. Overall, only in two cases, e.g. clerks and consulting specialists, abilities failed to significantly explained
the variance in both job performance criteria. Furthermore, after building an initial model with only GMA as a predictor,
and the second one with GMA and SMA composite as predictors, only in three cases (customer service, marketing and
QA’s) SMA did not show incremental validity (however, in the case of customer service and QA’s, only when the rating
of performance was being considered). In every other group, SMA proved to have significant and substantial
incremental validity, which was confirmed by outcomes of ANOVA tests. On average, in the case of job performance
ratings, SMA increased R2 of the final model by 19% and by 37% in the case of job performance quantitative measure
(in comparison with the initial model, with only GMA).
Table 3. Results of incremental validity analysis
Dependant variable: financial outcome
Dependant variable: rating
R2 for models:
R2 for
models:
department
GMAa
SMAa
Botha
R2
Fb
GMAa
SMAa
Botha
R2
Fb
board
0.27
0.34
0.37
0.11 (29%)
16.85***
0.33
0.35
0.35
0.02
(6%)
2.91*
buyers
0.16
0.27
0.30
0.15 (48%)
8.87***
0.25
0.36
0.40
0.15 (38%)
10.75***
clerks
0.00
0.01
0.01
0.01 (86%)
5.04*
0.00
0.00
0.00
0.00
(0%)
0.14
constructors
0.08
0.11
0.12
0.04 (32%)
8.77***
0.19
0.21
0.21
0.02 (10%)
5.36**
consultants
0.01
0.01
0.01
0.00
(0%)
0.12
0.01
0.01
0.01
0.00
(0%)
0.24
customer service
0.13
0.14
0.16
0.03 (20%)
15.5***
0.22
0.23
0.23
0.01
(1%)
1.29
financiers
0.01
0.06
0.07
0.06 (89%)
26.57***
0.00
0.05
0.06
0.06 (95%)
26.18***
HR
0.02
0.06
0.06
0.03 (59%)
5.72**
0.04
0.07
0.07
0.04 (50%)
6.41**
IT
0.26
0.29
0.30
0.04 (12%)
12.6***
0.39
0.42
0.43
0.04
(9%)
17.23***
manufactory workers
0.06
0.09
0.12
0.07 (54%)
17***
0.21
0.22
0.22
0.01
(7%)
4.12*
marketing specialists
0.21
0.22
0.22
0.01
(2%)
0.92
0.28
0.28
0.29
0.01
(1%)
0.78
QA’s
0.01
0.16
0.25
0.23 (95%)
9.64***
0.11
0.16
0.19
0.08 (41%)
2.89**
researchers
0.46
0.70
0.70
0.24 (35%)
28.56***
0.50
0.62
0.63
0.13 (21%)
12.42***
salespeople
0.19
0.21
0.22
0.03 (15%)
40.37***
0.35
0.36
0.36
0.01
(1%)
6.33**
transition specialists
0.11
0.15
0.15
0.04 (27%)
13.41***
0.17
0.19
0.19
0.02
(8%)
5.32**
Note. a predictors included in models; b ANOVA test results for incremental validity significance.
* p < 0.05; ** p < 0.01; ***p < 0.001.
Importantly, there were groups in which SMA increased the percentage of explained variance in job performance by
half or more. These results oppose the aforementioned claims of a lack of incremental validity of SMA and therefore
lend further support to Hypothesis 1.
3.3 Moderation Analysis
Along with the hierarchical regression analysis, the moderation analysis was employed. For this part, the regression
models were built with only GMA as a predictor of job performance measures variance. Also, an occupational group (as
a categorical variable) was included as a moderator of this relationship, along with the interaction effects of predictors.
If the GMA were a universally valid job performance predictor, no interaction effect (no moderation) should be
observed. An initial model (with only GMA) for the whole sample accounted for only 9% of job performance
quantitative variance (F(2, 4031) = 422.55; p < 0.001) and 14% of job performance rating
(F(2, 4031) = 632.73; p < 0.001). When the moderator was included, the model accounted for 25% of the quantitative
measure (F(30, 4003) = 46.50; p < 0.001) and 26% of rating (F(30, 4003) = 48.01; p < 0.001). Importantly, the
Business and Management Studies Vol. 4, No. 3; 2018
8
interaction effects of GMA and occupational group were significant in both regression models and ANOVA tests
confirmed that the models with and without interaction effects differed significantly in terms of quantitative (F = 30.03;
df = 28; p < 0.001) and rating (F = 23.58; df = 28; p < 0.001) measures of job performance. All in all, these data support
Hypothesis 2.
3.4 Relative Importance Analysis
In the final step, a relative importance analysis with the relative weights method was performed (Table 4). This allows a
determination of the contribution that every predictor included made to the total variance of job performance, both by
itself and in combination with the other predictors. More precisely, the relative weight analysis was employed
(Tonidandel & LeBreton, 2011) and the RWA-WEB was used to calculate the results (Tonidandel & LeBreton, 2014).
This method helps to determine the exact proportion of contribution that is expressed as a percentage for each predictor,
which adds up to 100%. The larger the percentage for a predictor, the more relatively important it is, even when other
variables are taken into account (Tonidandel & LeBreton, 2014).
Table 4. Results of relative weights analysis.
department
QR
V
RC
DR
LK
IR
WF
GMA
board
19.41 /
19.99
21.05 /
18.85
7.49 /
8.87
15.85 /
15.67
10.68 /
4.17
5.26 /
13.98
6.01 /
3.27
14.25 /
15.19
buyers
12.05 /
14.9
18.79 /
2.24
23.15 /
27.05
14.17 /
5.92
5.06 /
23.78
6.94 /
1.94
1.43 /
8.68
18.42 /
15.49
constructors
10.1 /
10.49
22.82 /
6.72
10.62 /
12.92
3.64 /
4.57
1.38 /
9.53
19.57 /
17.15
19.2 /
25.64
12.66 /
12.98
customer
service
40.85 /
31.17
8.8 /
6.61
12.88 /
13.48
5.05 /
6.01
2.36 /
3.68
5.82 /
15.49
3.57 /
4.89
20.68 /
18.67
financiers
17.07 /
1.47
1.9 /
1.84
1.71 /
2.61
26.17 /
65.47
14.41 /
18.59
3.4 /
1.06
26.84 /
2.52
8.51 /
6.45
HR
17.31 /
4.82
12.51 /
1.94
5.33 /
6.79
7.89 /
2.34
1.12 /
12
31.38 /
34.82
11.38 /
26.67
13.09 /
10.62
IT
24.91 /
24.91
3.56 /
3.73
17.86 /
13.2
16.41 /
19.41
15.5 /
17.21
2.75 /
5.35
1.49 /
1.43
17.52 /
14.74
manufactory
workers
20.69 /
28.33
5.52 /
10.48
13.16 /
17.2
17.79 /
14.25
3.01 /
7.54
3.6 /
3.89
23.44 /
1.27
12.8 /
17.04
marketing
specialists
17.33 /
19.17
5.6 /
9.95
29.9 /
20.7
5.82 /
12.65
10.51 /
9.13
5.46 /
5.08
3.24 /
6.03
22.16 /
17.29
QA’s
2.77 /
7.66
49.67 /
8.83
18.98 /
5.79
1.10 /
11.91
1.26 /
4.21
0.94 /
10.02
15.56 /
39.12
9.72 /
12.47
researchers
18.44 /
27.76
20.89 /
18.11
3.31 /
5.16
25.24 /
19.67
3.21 /
5.11
15.49 /
7.99
1.18 /
0.69
12.24 /
15.5
salespeople
28.37 /
18.97
8.02 /
12.28
16.32 /
17.83
5.93 /
9.17
17.68 /
12.93
3.67 /
7.53
1.95 /
4.00
18.07 /
17.29
transition
specialists
8.97 /
15.8
3.13 /
2.49
17.06 /
16.38
13.22 /
13.72
27.29 /
19.82
14.35 /
13.76
1.64 /
2.93
14.34 /
15.09
Note. All results are presented in percentages; the first number is the relative importance for quantitative measure, the
second for rating; QR = quantitative reasoning; V = visualization; RC = reading comprehension; DR = deductive
reasoning; LK = lexical knowledge; IR = inductive reasoning; WF = word fluency.
The groups where GMA and SMA significantly explain variance in job performance were taken into account at this
stage. Importantly, within these 13 groups, GMA did not even once contribute to the overall variance mostly. In almost
every case there were single narrow ability test results that had a superior contribution to job performance variance than
GMA. On average, a single most relatively important predictor (a SMA result) contributed 108% more in the case of the
quantitative job performance measure and 140% more in the case of the rating measure than GMA did. In some cases
(e.g. customer service group, financiers and HR specialists), a single SMA contributed over twice as much as GMA to
the job performance measure variance. The pattern of contribution was, however, strongly diversified. There were no
SMA that could be considered universally to be the most important predictor regardless of occupation. Each SMA was
the predictor that contributed most to the variability of job performance in at least one group. The outcome of this
analysis supports Hypothesis 1, as SMA proved to be relatively more important than GMA in predicting the variance of
job performance measures.
4. Discussion
The SVT perspective, emphasizing the role of SMA in personnel selection, fits better the data, as results showed
patterns exactly in accordance with the assumptions of this theory, i.e. the job-unique weighting composites based on
Business and Management Studies Vol. 4, No. 3; 2018
9
several SMA were far superior to GMA alone in predicting job performance variance. The data presented above support
both hypotheses. Drawn from Motowidlo et al. (1997) theory, one could assume SMA to be a basic tendency that leads
to job performance in certain occupation, which is also relatively more important than GMA.
Schmidt (2002) noticed that GMA “has higher validity than any single aptitude” (p. 189), which is a credo of the
unitarian perspective. As shown in this study, SMA had a superior validity and importance than GMA in job
performance variance prediction. Furthermore, researchers who support the unitarian perspective (Carretta & Ree, 2000;
Ones, Viswesvaran, & Dilchert, 2006; Schmidt, 2002) claimed that differences between jobs do not affect the
generalizability of GMA validity. Differences in validity magnitude and importance across occupations seem to support
the opposing statement. Next, when job differentiation was considered, SMA proved to have at least decent incremental
validity, and importantly, adding specific weighted component to GMA in the regression model substantially increased
predictive validity. Consequently, the results were contrary to the claims that SMA added little or nothing more than
GMA to job performance prediction.
Overall, the validity of SMA in job performance prediction was beyond what GMA could offer. Viswesvaran & One
(2002) pointed out that SVT could not be confirmed, as GMA is responsible for most of the predicted variance in
variables. The results presented support an opposite conclusion. Proponents of the unitarian perspective also claimed
that SVT should be disconfirmed, as validity of SMA comes from, and is dependent on, GMA. Thanks to the
methodology employed, the study results support a contrasting statement, as the relative weight analysis allows one to
determine the contribution of a predictor, irrespective of its relation with other predictors (Lang et al., 2010). As the
contribution of SMA was considerably larger than that of GMA, the results fit SVT better.
The results presented come from a study that does not suffer from the most pressing methodological issue that Reeve
(2004) pointed out and which caused “at least the questioning of some of the evidence regarding the lack of value of
narrow abilities” (p. 624). However, the results needed to be interpreted particularly carefully, as decades of studies and
many meta-analytic results cannot be of course disproved by a single study. Nevertheless, in the author’s opinion the
data presented are robust enough to serve as proof of the concept and provoke a data-driven discussion on the actual and
recent relative validity and importance of both GMA and SMA. As Krumm et al. (2014) noted, moderators and
boundary conditions of the major role of GMA in performance prediction are important observations from the point of
view of a validity study. This study provides data to suggest that an occupation could be such a moderator.
It is important to consider the reasons for the discrepancy between the study presented and previous findings. The
dispersion of results between groups should be considered crucial, because it shows that analysis within occupations
emphasizes the general differences between the predictive validity of both SMA and GMA and is essential in providing
evidence for unitarian or SVT perspectives. This part of results and therefore the contribution of the study to the
literature seems to be of utmost importance, as previous studies often failed to take occupational context into
consideration in a systematic and comprehensive manner. Interestingly, Schneider & Newman (2015) described the
compatibility principle that could explain some of SMA predictive validity. They pointed out that GMA has better
validity in the prediction of general job performance and SMA are better predictors when it comes to specific job
performance. The results of this study also seem to confirm this observation, as the criterion used here seems to be
strongly connected with specific job performance the salary is often based on a series of indicators dedicated to a
certain profession, and the rating of the percentage of the achieved goals relied solely on job-specific tasks.
Many meta-analytic studies did not take occupational context into consideration to a satisfactory degree and this might
be the reason for the discrepancy between the results presented and the mainstream view on GMA and SMA validity.
However, one could list a series of further, substantial reasons and explanations. Firstly, the meta-analyses summarized
studies from decades ago (Schmidt & Hunter, 1998). Since this study is based on recent data, it may reflect the more
current state of affairs in job performance determination. Furthermore, few studies focused on the composites of SMA
results specifically weighted for a given context, emphasizing instead the analysis of single scores. There is evidence
that this might moderate the actual incremental validity of SMA observed in a study (Reeve, 2004; Schneider &
Newman, 2015). Finally, many validity studies were conducted on already employed participants, and therefore the
range restriction of their abilities and performance was high. Thanks to the methodology employed, the results
presented were free from this auto-selection issue, and might reflect outcomes closer to actual ones.
4.1 Theoretical Implication
Most importantly, a plausible theoretical explanation for the observed results could be identified. The validity of GMA
and the superior role of SMA could be interpreted on the grounds of the triarchic theory of intelligence (Sternberg,
1997). According to the componential parts of Sternberg’s theory, one could distinguish different components of
information processing. The metacomponents are responsible for the management of one’s mind and play executive
function. They are responsible for a correct identification of a given problem and for making decision regarding the best
Business and Management Studies Vol. 4, No. 3; 2018
10
way to resolve that problem (Diamond, 2013; Sternberg, 1985). The performance components are in turn the processes
that are directly used to solve a given problem; they carry out the action that metacomponents command (Sternberg,
1997). On this basis, one can assume that GMA play the role of metacomponents, used in identifying a job problem and
deciding on how to perform, while SMA serve as performance components and are actually used to handle job tasks
easily and faultlessly. That is why in the majority of cases a moderate level of GMA is required to perform at a
minimum level. This process explained also the high importance and incremental validity of SMA, as performance
components have a complementary role over metacomponents.
4.2 Practical Implication
Results from the study presented can be used by HR practitioners in competency modeling processes and could support
decisions concerning the selection tools chosen by companies. Information about the validity of SMA and GMA could
be used both as grounds for new competency models based on abilities and for the development of current models used
in organizations. Organizations should be interested in basing decisions about HR processes on empirical evidence,
because, as the study has demonstrated, they may be able to carry out successful personnel selection using a limited
range of tools (a few tests, instead of a whole battery) with comparable results (but saving time and resources).
4.3 Limitation and Further Guidelines
There are of course some limitations to this research that should be noted. Measurements of performance in this study
gave rise to concerns. Such measures as salary and self-reports are vulnerable to many cofounders. Both of the
measures used have problems with content validity. On the other hand, the use of quantitative criteria is one of the most
contributing parts of this study. As Rojon et al. (2015) noted, many authors avoid them as they often result in lower
coefficients. Furthermore, the two measures of construct in interest were employed, according to the criterion-driven
approach, which also should contribute to reliable results (Kaplan, Cortina, & Ruark, 2010). Nevertheless, the measures
are debatable, at best. This, of course, limits the possibility to generalize conclusions drawn from the results to certain
criteria and reduces to some degree the comparability of this study. It is, however, worth noting that the intention of the
author was to provide proof of the concept and to deliver data to identify a relative validity of GMA and SMA. It seems
reasonable that the employed measures could be regarded sufficient for those purposes.
One could list a series of guidelines that need to be followed in validation studies if they are to provide useful data. As
analyzing predictive validity with or without occupational grouping has a major impact on the results, the coefficients
should always be presented for certain groups (the more specific, the better). Future validation research should focus on
a systematic comparison of the predictive validity of GMA and SMA across a series of occupations, with the use of a
series of job performance criteria. At this point, studies investigating single predictors for one occupational group seem
to have very limited utility. Overall, this will allow researchers to conduct more detailed systematic review studies and
compare the relative importance of predictors.
5. Conclusion
There is a reasonable basis to have a further debate on SVT in personnel selection and it is no wonder that this
perspective has been gaining importance in recent years (Lang et al., 2010; Reeve, 2004; Schneider & Newman, 2015;
Stanhope & Surface, 2014). It is worth referring to Schmidt’s significant statement (2002) that “there cannot be a
serious debate(p. 188) whether GMA is the best performance predictor. No one questions the valid role of GMA as a
job performance predictor, but there certainly should be a debate.
References
Barros, E., Kausel, E. E., Cuadra, F., & az, D. A. (2014). Using General Mental Ability and Personality Traits to
Predict Job Performance in Three Chilean Organizations. International Journal of Selection and Assessment, 22(4),
432438.
Bertua, C., Anderson, N., & Salgado, J. F. (2005). The predictive validity of cognitive ability tests: A UK meta-analysis.
Journal of Occupational and Organizational Psychology, 78(3), 387409.
https://doi.org/10.1348/096317905X26994
Carretta, T. R., & Ree, M. (2000). General and specific cognitive and psychomotor abilities in personnel selection: The
prediction of training and job performance. International Journal of Selection and Assessment, 8(4), 227236.
https://doi.org/Doi 10.1111/1468-2389.00152
Carretta, T. R., & Ree, M. J. (1997). Expanding the Nexus of Cognitive and Psychomotor Abilities. International
Journal of Selection and Assessment, 5(3), 149158. https://doi.org/10.1111/1468-2389.00055
Cote, S., & Miners, C. T. H. (2006). Emotional Intelligence , Cognitive Intelligence , and Job Performance.
Administrative Science Quarterly, 51(1), 128.
Business and Management Studies Vol. 4, No. 3; 2018
11
Diamond, A. (2013). Executive Functions. Annual Review of Psychology, 64(1), 135168.
https://doi.org/10.1146/annurev-psych-113011-143750
Downey, L. A., Lee, B., & Stough, C. (2011). Recruitment Consultant Revenue: Relationships with IQ, personality, and
emotional intelligence. International Journal of Selection and Assessment, 19(3), 280286.
https://doi.org/10.1111/j.1468-2389.2011.00557.x
Hausdorf, P. A., & Risavy, S. D. (2015). Predicting training and job performance for transit operators. International
Journal of Selection and Assessment, 23(2), 15. https://doi.org/10.1111/ijsa.12107
Hirsh, H. R., Northrop, L. C., & Schmidt, F. L. (1986). Validity generalization results for law enforcement occupations.
Personnel Psychology, 39(2), 399420. https://doi.org/10.1111/j.1744-6570.1986.tb00589.x
Hogan, J., Hogan, R., & Gregory, S. (1992). Validation of a sales representative selection inventory. Journal of Business
and Psychology, 7(2), 161171. https://doi.org/10.1007/BF01013926
Hunter, J. E., & Hunter, R. F. (1984). Validity and utility of alternative predictors of job performance. Psychological
Bulletin, 96(1), 7298. https://doi.org/10.1037/0033-2909.96.1.72
Johnson, J. W., & Lebreton, J. M. (2004). History and Use of Relative Importance Indices in Organizational Research.
Organizational Research Methods, 7(3), 238257. https://doi.org/10.1177/1094428104266510
Kaplan, S., Cortina, J., & Ruark, G. A. (2010). Oops…. We Did It Again: Industrial–Organizational’s Focus on
Emotional Intelligence Instead of on Its Relationships to Work Outcomes. Industrial and Organizational
Psychology, 3(02), 171177. https://doi.org/10.1111/j.1754-9434.2010.01220.x
Krumm, S., Schmidt, A. L., & Lipnevich, A. A. (2014). Specific Cognitive Abilities at Work. Journal of Personnel
Psychology, 13(3), 117122. https://doi.org/10.1027/1866-5888/a000117
La Grange, L., & Roodt, G. (2007). Personality and cognitive ability as predictors of the job performance of insurance
sales people. SA Journal of Industrial Psychology, 27(3), 3543.
Lakens, D., Hilgard, J., & Staaks, J. (2016). On the reproducibility of meta-analyses: six practical recommendations.
BMC Psychology, 4(1), 24. https://doi.org/10.1186/s40359-016-0126-3
Lang, J. W. B., Kersting, M., lsheger, U. R., & Lang, J. (2010). General mental ability, narrower cognitive abilities,
and job performance: The perspective of the nested-factors model of cognitive abilities. Personnel Psychology,
63(3), 595640. https://doi.org/10.1111/j.1744-6570.2010.01182.x
LeBreton, J. M., Scherer, K. T., & James, L. R. (2014). Corrections for Criterion Reliability in Validity Generalization:
A False Prophet in a Land of Suspended Judgment. Industrial and Organizational Psychology, 7(4), 478500.
https://doi.org/10.1111/iops.12184
Marcus, B., Goffin, R. D., Johnston, N. G., & Rothstein, M. G. (2007). Personality and Cognitive Ability as Predictors
of Typical and Maximum Managerial Performance. Human Performance, 20(3), 275285.
https://doi.org/10.1080/08959280701333362
Motowidlo, S. J., Borman, W. C., & Schmit, M. J. (1997). A Theory of Individual Differences in Task and Contextual
Performance. Human Performance, 10(2), 7183. https://doi.org/10.1207/s15327043hup1002_1
Mount, M. K., Oh, I. S., & Burns, M. (2008). Incremental validity of perceptual speed and accuracy over general mental
ability. Personnel Psychology, 61, 113139. https://doi.org/10.1111/j.1744-6570.2008.00107.x
Muchinsky, P. M. (1993). Validation of Intelligence and Mechanical Aptitude Tests in Selecting Employees for
Manufacturing Jobs. Journal of Business and Psychology, 7(4), 373382.
Olea, M., & Ree, M. J. (1994). Predicting pilot and navigator criteria: Not much more than g. Journal of Applied
Psychology, 79(6), 845851. https://doi.org/10.1037/0021-9010.79.6.845
Ones, D. S., Viswesvaran, C., & Dilchert, S. (2006). Cognitive ability in selection decisions. In D. Wilheml & R. Engle
(Eds.), Understanding and measuring intelligence. London: Sage Publications.
Ree, M. J., & Earles, J. A. (1992). Intelligence Is the Best Predictor of Job Performance. Current Directions in
Psychological Science, 1(3), 8689. https://doi.org/10.1111/1467-8721.ep10768746
Ree, M. J., Earles, J. A., & Teachout, M. S. (1994). Predicting Job Performance: Not Much More Than g. Journal of
Applied Psychology, 79(4), 518524.
Reeve, C. L. (2004). Differential ability antecedents of general and specific dimensions of declarative knowledge: More
than g. Intelligence, 32(6), 621652. https://doi.org/10.1016/j.intell.2004.07.006
Business and Management Studies Vol. 4, No. 3; 2018
12
Richardson, K., & Norgate, S. H. (2015). Does IQ Really Predict Job Performance? Applied Developmental Science,
8691(February), 117. https://doi.org/10.1080/10888691.2014.983635
Rojon, C., McDowall, A., & Saunders, M. N. K. (2015). The Relationships Between Traditional Selection Assessments
and Workplace Performance Criteria Specificity: A Comparative Meta-Analysis. Human Performance, 28(1), 125.
https://doi.org/10.1080/08959285.2014.974757
Salgado, J. F., Anderson, N., Moscoso, S., Bertua, C., de Fruyt, F., & Rolland, J. P. (2003). A meta-analytic study of
general mental ability validity for different occupations in the European community. The Journal of Applied
Psychology, 88(6), 10681081. https://doi.org/10.1037/0021-9010.88.6.1068
Schmidt, F. L. (2002). The role of general cognitive ability and job performance: Why there cannot be a debate. Human
Performance, 15(1/2), 187210.
Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: Practical
and theoretical implications of 85 years of research findings. Psychological Bulletin, 124(2), 262274.
https://doi.org/10.1037/0033-2909.124.2.262
Schneider, W. J., & Newman, D. A. (2015). Intelligence is multidimensional: Theoretical review and implications of
specific cognitive abilities. Human Resource Management Review, 25(1), 1227.
https://doi.org/10.1016/j.hrmr.2014.09.004
Stanhope, D. S., & Surface, E. A. (2014). Examining the incremental validity and relative importance of specific
cognitive abilities in a training context. Journal of Personnel Psychology, 13(3), 146156.
https://doi.org/10.1027/1866-5888/a000116
Sternberg, R. J. (1985). Beyond IQ: A Triarchic Theory of Intelligence. Cambridge: Cambridge University Press.
Sternberg, R. J. (1997). A Triarchic View of Giftedness: Theory and Practice. In N. Coleangelo & G. A. Davis (Eds.),
Handbook of Gifted Education (pp. 4353). Boston, MA: Allyn and Bacon.
Sternberg, R. J. (2001). The relationship between academic and practical intelligence: a case study in Kenya.
Intelligence, 29(5), 401418. https://doi.org/10.1016/S0160-2896(01)00065-4
Thomas, J. C., Barrett, G. V., & Alexander, R. A. (1996). The relationship of specific mental ability measures compared
to a general mental ability measure to quality and quantity performance on a clerical job sample. Journal of
Business and Psychology, 11(1), 3541. https://doi.org/10.1007/BF02278253
Tonidandel, S., & LeBreton, J. M. (2011). Relative Importance Analysis: A Useful Supplement to Regression Analysis.
Journal of Business and Psychology, 26(1), 19. https://doi.org/10.1007/s10869-010-9204-3
Tonidandel, S., & LeBreton, J. M. (2014). RWA Web: A Free, Comprehensive, Web-Based, and User-Friendly Tool for
Relative Weight Analyses. Journal of Business and Psychology, 110. https://doi.org/10.1007/s10869-014-9351-z
Verbeke, W. J., Belschak, F. D., Bakker, A. B., & Dietz, B. (2008). When Intelligence Is (Dys)Functional for Achieving
Sales Performance. Journal of Marketing, 72(4), 4457. https://doi.org/10.1509/jmkg.72.4.44
Vinchur, A. J., Schippmann, J. S., Switzer, F. S., & Roth, P. L. (1998). A Meta-Analytic Review of Predictors of Job
Performance for Salespeople. Journal of Applied Psychology, 83(4), 586597.
https://doi.org/10.1037/0021-9010.83.4.586
Viswesvaran, C., & Ones, D. S. (2002). Agreements and Disagreements on the Role of General Mental Ability (GMA)
in Industrial, Work, and Organizational Psychology. Human Performance, 15(1), 211231.
https://doi.org/10.1207/S15327043HUP1501&02_13
Whetzel, D. L., McCloy, R. A., Hooper, A., Russell, T. L., Waters, S. D., Campbell, W. J., & Ramos, R. A. (2011).
Meta-Analysis of Clerical Performance Predictors: Still stable after all these years. International Journal of
Selection and Assessment, 19(1), 4150. https://doi.org/10.1111/j.1468-2389.2010.00533.x
Copyrights
Copyright for this article is retained by the author(s), with first publication rights granted to the journal.
This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution license
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
... According to researchers and HR professionals, cognitive ability is one of the most important factors in predicting job performance (Grobelny 2018). However, there is an ongoing debate in the literature about whether general or specific cognitive abilities predict job performance better. ...
... These suggest that the predictive validity of specific abilities also varies widely across occupations. Grobelny's (2018) own research backed up his ideas. Specific cognitive abilities had higher predictive power than general cognitive ability, and the predictive power of different abilities depended on the job in question. ...
... Based on the research provided by Grobelny (2018) and labor market experience, the system was designed to measure the following abilities: ...
Chapter
Full-text available
Gamified flipped learning has become admired in language teaching and learning as a creative and effective way to improve student learning and encourage them to comprehend and process new knowledge. The purpose of this study was to investigate how a gamified flipped EFL classroom affected second-grade students’ vocabulary learning in a private Turkish school. The students were assigned into two groups: experimental and control. There were 20 participants in the experimental group, and 20 in the control group. Teacher reflective journals and structured student interviews were used in this study. Following the treatment of the experimental group, both the control and experimental groups were given a posttest. The results of this study’s experimental group posttest attempted to show that using gamified flipped classrooms improved the experimental group’s mean score, and also had a positive impact on students’ engagement, and increased their motivation toward learning in in-class gamified activities. The key findings of structured interviews were students wanted more interaction inside the videos. They also claimed that they liked the use of hands-on activities in online synchronous classes instead of having only instruction.KeywordsFlipped learningGamificationEFLGamified flipped learningPrimary studentsVocabulary learning
... Furthermore, GMA and the Big Five were measured on a higher, more abstract level (i.e., broad concepts), because we wanted to study GMA and the Big Five on the same level of abstraction as reported in well-known meta-analyses (Schmidt and Hunter, 1998;e.g., Salgado and Anderson, 2003). Some studies highlight the role of narrow personality traits (e.g., facets of conscientiousness) and specific aptitudes (e.g., psychomotor abilities) as predictors of job performance (e.g., Schmidt, 2002;Dudley et al., 2006;Grobelny, 2018) as well. Additionally, character strengths and facets of the Big Five overlap (e.g., perseverance as a character strength with achievement thriving and self-discipline as facets of conscientiousness, self-regulation as a character strength with impulsiveness as a facet of neuroticism), although they are not redundant (Noftle et al., 2011;McGrath et al., 2020). ...
Article
Full-text available
Over the last decades, various predictors have proven relevant for job performance [e.g., general mental ability (GMA), broad personality traits, such as the Big Five]. However, prediction of job performance is far from perfect, and further potentially relevant predictors need to be investigated. Narrower personality traits, such as individuals' character strengths, have emerged as meaningfully related to different aspects of job performance. However, it is still unclear whether character strengths can explain additional variance in job performance over and above already known powerful predictors. Consequently, the present study aimed at (1) examining the incremental validity of character strengths as predictors of job performance beyond GMA and/or the Big Five traits and (2) identifying the most important predictors of job performance out of the 24 character strengths, GMA, and the Big Five. Job performance was operationalized with multidimensional measures of both productive and counterproductive work behavior. A sample of 169 employees from different occupations completed web-based self-assessments on character strengths, GMA, and the Big Five. Additionally, the employees' supervisors provided web-based ratings of their job performance. Results showed that character strengths incrementally predicted job performance beyond GMA, the Big Five, or GMA plus the Big Five; explained variance increased up to 54.8, 43.1, and 38.4%, respectively, depending on the dimension of job performance. Exploratory relative weight analyses revealed that for each of the dimensions of job performance, at least one character strength explained a numerically higher amount of variance than GMA and the Big Five, except for individual task proactivity, where GMA exhibited the numerically highest amount of explained variance. The present study shows that character strengths are relevant predictors of job performance in addition to GMA and other conceptualizations of personality (i.e., the Big Five). This also highlights the role of socio-emotional skills, such as character strengths, for the understanding of performance outcomes above and beyond cognitive ability.
... However, Personality Trait could not predict the relationship of Employee Job Performance but was significant. This is confirmed by Grobelny (2018) that he also found out that the general and specific mental abilities which are also professional skills predicted for Job performance. ...
Article
Full-text available
Introduction: In today’s global environment there is a competition on performances of organizations. Business organizations are seeking highly skilled professionals to be employed to achieve high performance and productivity over other competitors. The purpose of this research is to determine the effects of personality traits, and professional skills on employee performance in an Automotive Corporation in the Philippines. Methodology: The study is a quantitative research, and a descriptive-correlation. A self-constructed survey questionnaire was distributed to 97 employees of Automotive Corporation using convenience sampling. The statistical tools used for analyzing the results from the SPSS 22, was Pearson correlation to analyze the relationship and standard deviation and the mean for the descriptive study, and t-test and ANOVA were used to analyze the difference, and in terms of predictions, linear regression was used. Result: the Study revealed that the relationship between Professional Skills has a positive significant relationship with employee performance but there is no significant relationship between personality Traits and employee performance, though openness was dominant with the highest mean. The study also showed that professional skills, personality traits, and employee performance are high. Considering the sex, males had a higher employee performance than females. Further results revealed that only professional skills predicted employee performance. Discussion: The result of the study implies that business owners need employees that are able to get the job done because employee performance is critical to the overall success of the company. The study recommends that industries should pay attention to Professional Skills, and Personality Traits of workers because it is important in business organizations.
Chapter
The most important role of education is to promote successful employment by equipping students with the abilities, skills and knowledge they need. In our study, we present the results of a programme to assess disadvantaged workers using digital measurement tools and to develop a training programme based on the results. The results of the programme were compared with a reference group of currently unemployed people who were already working. Based on this, we conducted a training and then assessed the participants again after one year. The skills developed by participants were significantly closer to those of people already in work, which increases the success rate of job search.KeywordsAssessmentCognitive abilityDisadvantagedTraining
Article
Full-text available
Background Meta-analyses play an important role in cumulative science by combining information across multiple studies and attempting to provide effect size estimates corrected for publication bias. Research on the reproducibility of meta-analyses reveals that errors are common, and the percentage of effect size calculations that cannot be reproduced is much higher than is desirable. Furthermore, the flexibility in inclusion criteria when performing a meta-analysis, combined with the many conflicting conclusions drawn by meta-analyses of the same set of studies performed by different researchers, has led some people to doubt whether meta-analyses can provide objective conclusions. DiscussionThe present article highlights the need to improve the reproducibility of meta-analyses to facilitate the identification of errors, allow researchers to examine the impact of subjective choices such as inclusion criteria, and update the meta-analysis after several years. Reproducibility can be improved by applying standardized reporting guidelines and sharing all meta-analytic data underlying the meta-analysis, including quotes from articles to specify how effect sizes were calculated. Pre-registration of the research protocol (which can be peer-reviewed using novel ‘registered report’ formats) can be used to distinguish a-priori analysis plans from data-driven choices, and reduce the amount of criticism after the results are known. SummaryThe recommendations put forward in this article aim to improve the reproducibility of meta-analyses. In addition, they have the benefit of “future-proofing” meta-analyses by allowing the shared data to be re-analyzed as new theoretical viewpoints emerge or as novel statistical techniques are developed. Adoption of these practices will lead to increased credibility of meta-analytic conclusions, and facilitate cumulative scientific knowledge.
Article
Full-text available
A meta-analysis on the validity of tests of general mental ability (GMA) and specific cognitive abilities for predicting job performance and training success in the UK was conducted. An extensive literature search resulted in a database of 283 independent samples with job performance as the criterion (N=13,262), and 223 with training success as the criterion (N=75,311). Primary studies were also coded by occupational group, resulting in seven main groups (clerical, engineer, professional, driver, operator, manager, and sales), and by type of specific ability test (verbal, numerical, perceptual, and spatial). Results indicate that GMA and specific ability tests are valid predictors of both job performance and training success, with operational validities in the magnitude of .5–.6. Minor differences between these UK findings and previous US meta-analyses are reported. As expected, operational validities were moderated by occupational group, with occupational families possessing greater job complexity demonstrating higher operational validities between cognitive tests and job performance and training success. Implications for the practical use of tests of GMA and specific cognitive abilities in the context of UK selection practices are discussed in conclusion.
Chapter
Cognitive ability in selection decisions Intelligence testing has had a checkered history in psychological sciences. Several terms such as cognitive ability, general mental ability , and g factor have been used in the literature (Viswesvaran & Ones, 2002). In this chapter, we use the term cognitive ability to refer to these concepts generically, and we reserve the term general mental ability (GMA) to refer to the general factor that spans these measures. GMA is a general informationprocessing capacity and is extracted as a general factor (the first unrotated factor) from a battery of specific ability tests. 1 It is closely associated with reasoning and judgment abilities (see Chapter 21 , this volume). A century of scientific research has shown that GMA is predictive of socioeconomic status, academic achievement, health-related behaviors, social outcomes, occupational status, and even death (Brand, 1987). Ree and Carretta (2002) note that GMA predicts a wide range ...
Article
This meta-analysis evaluated predictors of both objective and subjective sales performance. Biodata measures and sales ability inventories were good predictors of the ratings criterion, with corrected rs of .52 and .45, respectively. Potency (a subdimension of the Big 5 personality dimension Extraversion) predicted supervisor ratings of performance (r =.28) and objective measures of sales (r =.26). Achievement (a component of the Conscientiousness dimension) predicted ratings (r =.25) and objective sales (r=.41). General cognitive ability showed a correlation of .40 with ratings but only .04 with objective sales. Similarly, age predicted ratings (r =.26) but not objective sales (r = -.06). On the basis of a small number of studies, interest appears to be a promising predictor of sales success.
Article
Although considerable meta-analytic research has validated the use of cognitive ability tests, structured interviews, and personality tests with training and job performance criteria, few studies have investigated the validity of these measures with transit operators. There are the only two single studies of concurrent validation research specifically with transit operators. This article presents the results of a predictive validation study conducted with transit operator applicants for a large urban transit authority in Canada. Key knowledge, skills, and abilities were determined for the role and used as a basis for the design and choice of predictors and criteria. Four predictors were used in the study: education, cognitive ability, personality assessment, and structured interview. Criteria included training performance (formative and summative), probationary performance, preventable accidents, and lost time injuries. Validation results supported cognitive ability, structured interview, and several personality factors as predictors of training performance, but less so for job performance. The use of formative training ratings greatly augmented the evidence supporting the predictors beyond typical organizational criteria.
Article
IQ has played a prominent part in developmental and adult psychology for decades. In the absence of a clear theoretical model of internal cognitive functions, however, construct validity for IQ tests has always been difficult to establish. Test validity, therefore, has always been indirect, by correlating individual differences in test scores with what are assumed to be other criteria of intelligence. Job performance has, for several reasons, been one such criterion. Correlations of around 0.5 have been regularly cited as evidence of test validity, and as justification for the use of the tests in developmental studies, in educational and occupational selection and in research programs on sources of individual differences. Here, those correlations are examined together with the quality of the original data and the many corrections needed to arrive at them. It is concluded that considerable caution needs to be exercised in citing such correlations for test validation purposes.
Article
We examine the extent to which g and specific abilities predict learning in a training context by juxtaposing contrasting theories of cognitive ability (i.e., Spearmanian theory and nested-factors theory) and using an analytical strategy (i.e., multidimensional importance analysis) that aligns with underlying theoretical assumptions of both theories. When conducting incremental validity analysis (Spearmanian approach), specific abilities did not add much to the prediction of learning. However, when conducting relative weights analysis (nested-factors approach), specific abilities were the dominant predictors of learning. Results suggest different theoretical and analytical approaches can lead to different results. Results also suggest specific abilities may provide predictive utility beyond g in the training context, but recognizing this utility depends on one’s theoretical assumptions and methodological approach.