Mckelvie 1989

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

Psychological Reports, 1989, 65, 161-162.

@ Psychological Reports 1989

T H E WONDERLIC PERSONNEL TEST: RELIABILITY AND


VALIDITY IN AN ACADEMIC SETTING '

STUART J. McKELVIE
Bishop's Universib

Summary.-Based on total of 290 undergraduates, the split-half reliability of the


Wonderlic Personnel Test was .87 and the Pearson correlation between test score and
mean grade was .21. Implications are presented for the use of this test in an academic
setting.

The Wonderlic Personnel Test (WPT) is a 12-min., 50-item spiral-


omnibus test of "problem-solving ability" (Wonderlic, 1983) that is widely
used as a screening device in business and industry (Murphy, 1984).
However, it may also be viewed as a test of general intelligence (Davou &
McKelvie, 1984) since its items are based on the original Otis Test of
Mental Ability (Wonderlic, 1983) and since scores correlate fairly well ( 5 6
to .80) with aptitude G (General Learning Ability) of the General Aptitude
Test Battery (Wonderlic, 1983) and very highly (.93) with the WAIS Full
Scale IQ (Dodrill, 1981). Given that general intelligence tests are often used
in academic settings, the test might be useful, particularly since it is avail-
able in 16 alternative forms. This note provides data concerning the internal
consistency and concurrent validity of the test with undergraduate students.
As part of a larger project, samples of 290 full-time undergraduates in
three consecutive years (1987-1989; n = 79, 99, 112) were administered the
Wonderlic under standard conditions. The samples were chosen to be repre-
sentative of the local university population, which was stratified by division
(Natural Science, Social Science, Humanities, Business), year of study
(three), and sex. Testing was conducted by students in a course on psycho-
logical testing. Each student administered the test to about 10 subjects, who
were recruited for this part of the project if they permitted the author (not
the testers) to view their transcripts.
Odd-even split-half reliability coefficients were computed and adjusted
with the Spearman-Brown formula, to give values of .89, .89, and .83 for
the three calendar years, the number for the combined samples being .87.
This estimate agrees with the values (.88, .94) provided in the manual
(Wonderlic, 1983), showing that the internal consistency of the test is
acceptable in both academic and nonacademic settings. To calculate grades

'I thank d those who participated and assisted in this study, particularly Mary Latuli pe,
Barbara McLeUan, Susie Shields, Sue Stuart, and Lynn White who helped to collate the fata,
and Sandra Gallichon and Kim Passey, who retrieved the transcripts. Reprint requests should
be sent to Smart J. McKelvie, Department of Psychology, Bishop's Univenity, Lennoxville,
Quebec, J I M 127 Canada.
162 S. J. McKELVIE

on the same basis, only courses taken in the most recent semester were con-
sidered, numbers for individual students ranging from three to six, the mode
being five. For the three consecutive years, Pearson correlations between
test scores and mean grades (concurrent validity) were .23 ( p < .05), .17
( p < .lo), and .23 ( p < .02), respectively; the combined value was .21.
Notably, this estimate is identical to the one given in the manual for the
correlation of test score with GPA. In addition, the Ms and SDs in the three
calendar years were 26.1 and 6.6 (1987), 27.0 and 6.4 (1988), and 26.8 and
5.8 (1989). These means were not significantly different (F,,,,, = .67, p > .20),
and the over-all M and SD were 26.7 and 6.2. The mean falls between the
test scores in the manual for high school (20.8) and college (29.6) graduates,
suggesting that the present over-all mean score is consistent with previous
findings.
These data show that the Wonderlic is internally consistent for under-
graduates and that their mean score falls as expected on the basis of
Wonderlic's norms. However, the low validity coefficient reported in the
manual and confirmed here suggests that the test has little practical value as
a predictor of individual grades, accounting as it does for only 4.4% of the
variance. Of course, this value might be slightly depressed by a restriction in
range, since only enrolled students were tested. Indeed, the current SDs
were slightly lower than the most relevant ones in the manual, which were
about 7.0. O n the other hand, the test might be useful as a general screen-
ing device or even as an addition to other admission data, particularly if the
selection ratio at the institution were low (Anastasi, 1988, p. 174). Probably
the most useful function for the Wonderlic in the academic setting, how-
ever, would be as a research instrument when a quick estimate of general
intelligence is needed. So I advise my students.
REFERENCES
ANASTASI,A. (1988) Psychological testing. (6th ed.) New York: Macmillan.
D~vou,D., & MCKELVIE,S. J. (1984) Relationship between study habits and performance on
an intelligence test with limited and unlimited time. Psychological Reports, 54, 367-371.
MURPHY,K. R. (1984) The Wonderlic Personnel Test. In D. J. Keyser & R. C. Sweetland
(Eds.), Test critiques. Vol. 1. Pp. 769-775.
WONDERLIC, E. F. (1983) Wonderlic Personnel Test manual. NorthField, IL: E. F. Wonderlic &
Assoc.

Accepted July 28, 1989.

You might also like