You are on page 1of 10

Archives of Clinical Neuropsychology 22 (2007) 969–978

The relationship of intelligence to executive function and


non-executive function measures in a sample of average,
above average, and gifted youth
Sharon Arffa ∗
The Watson Institute, 301 Camp Meeting Road, Sewickley, PA 15143, United States
Accepted 12 August 2007

Abstract
This study explores the relationship of intelligence to the Wisconsin Card Sorting Test, Stroop Color-Word Test, Oral Word Fluency
Test, Design Fluency Test, Trail Making Test, contrasted with Rey Complex Figure Test, Rey Auditory Verbal Learning Test, Wide
Range Achievement Test, and Underlining Test in average, above average and gifted children. Full-Scale IQ was significantly related
to Wisconsin Card Sort Perseverative and Non-Perseverative Errors, Stroop Color-Word Test, Color-Word condition, Controlled Oral
Word Fluency, Design Fluency, Rey Complex Figure, and Underlining conditions but not Trails or Rey Auditory Verbal Learning
Test. MANCOVA’s show gifted children outperformed other children on the executive but not the non-executive tests. Finally, the
nature of the neuropsychological/IQ relationship was explored by further analyses.
© 2007 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved.

Keywords: Intelligence; Executive function; Gifted; Youth

Executive function is a term for a number of complex cognitive processes that are interdependent and critical to
purposeful, goal directed behavior (Lezak, Howieson, Loring, & Hannay, 2004). Executive functions refer to a variety
of correlated abilities ranging from simple voluntary initiation and inhibition of behavior to those involving complex
planning, problem solving, and insight. Planning problem solving, and insight certainly correspond to psychological
and even lay concepts of “intelligent behavior”. However, evidence for a relationship of intelligence tests to executive
function measures is not strong. It is well known that frank frontal lesions do not impair IQ (Damasio & Anderson,
1993; Milner, 1982). The few studies that have examined psychometric intelligence and executive functions have been
inconsistent. In several studies with adults, executive function measures were not substantially related to IQ (Donders
& Kirsch, 1991; Johnstone, Holland & Larimore, 2000).
Some types of executive functions may be more associated with IQ than others. Seidenberg, Giordani, Berent, and
Boll (1983) found that neuropsychological measures requiring conceptual problem-solving ability, mental efficiency,
and language-related skills were more strongly related to IQ, while simple motor, constructional, and perceptual tasks,
were not. Research in cognitive psychology as well as clinical neuropsychology has emphasized the diversity in
traditional and laboratory tasks of executive function. A conference committee from the National Institute of Child
Health and Human Development (Eslinger, 1996) concluded that 40% of participants agreed on the presence of six

∗ Tel.: +1 412 749 2889.


E-mail address: sharona@thewatsoninstitute.org.

0887-6177/$ – see front matter © 2007 National Academy of Neuropsychology. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.acn.2007.08.001
970 S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978

latent variables: planning/preparatory set, inhibitory control, sequencing of behavior, flexibility and shifting mental set,
organization of behavior and self monitoring. Working memory has also been linked to executive function (Friedman &
Miyake, 2004). Specifically relating psychometric IQ to laboratory tasks of executive function, Friedman et al. (2006),
found updating tasks (those requiring adding and deleting information in working memory) to be highly correlated with
intelligence as measured by WAIS Full-Scale IQ, select subtests, and Ravens Progressive Matrices. Inhibiting (tasks
requiring suppressing automatic responses) or shifting (tasks requiring shifting between subtasks) was not strongly
related to IQ. The authors suggested that updating and working memory requires an individual to sustain attention in
order to process relevant and ignore irrelevant information, and in this manner, corresponds to Binet’s definition of
intelligence, which requires an individual to first perceive his world, log perception in memory, and to rework them.
In other studies the lack of executive function/IQ relationship has been attributed to ceilings on certain tests (e.g.,
Trail Making Test; Waldmann, Dickson, Monahan, & Kazelskis, 1992) or restricted IQ ranges (Wiens & Matazzaro,
1988). For example, adults with superior IQ did better on the Category Test, Oral Word Fluency Test, and Stroop
Color-Word Association Test than did a group with average IQ (Parsons, 1984). Arffa, Lovell, Podell, and Goldberg
(1998) found intelligence to be significant qualifier of age trends on the Wisconsin Card Sort in a pediatric sample
ranging from above average to highly gifted children.
Most research on the relationship of IQ to performance on executive function tests has been conducted with adults.
Like the adult literature, research in children is also inconsistent. Welsh, Pennington, and Grossier (1991) found no
correlation with IQ and several measures of executive function in 6–12-year-old children. No strong IQ/executive
function correlations were identified on tasks of motor planning (Golden, 1978), the Wisconsin Card Sorting Test
(Boone, Ghafferian, Lesser, Hill-Gutierrez, and Berman (1993), and the Trail Making Test (Ardila, Pineda, & Rosselli,
2000; Waldmann et al., 1992). Ardila et al. (2000) found correlations between absolute magnitudes of .30 and .37
on verbal fluency measures and Wisconsin Card Sort Perseverative Errors and Verbal and Full-Scale IQ, but not with
Trail Making Tests or other Wisconsin Card Sort conditions. The authors interpreted this as showing a reasonable
dissociation between IQ and executive function. Riccio, Hall, Morgan, and Hynd (1994), studying children with
attention deficit hyperactivity disorder, ages 6–16 years, found that Wechsler Intelligence Scale for Children-Revised
(WISC-R) performance IQ correlated with Wisconsin Card Sorting Test (WCST) variables for those between the ages
of 9 years and 11 years, 11 months, but not with the older children. Other investigators have highlighted the moderating
effect of IQ on executive functions, particularly at higher IQ levels (Arffa et al., 1998; Baron, 2003; Mahone et al., 2002).
Research clearly shows that executive functions improve with development throughout infancy, childhood, and
adolescence. The moderating influence of intelligence on executive function performance in children remains unclear.
The purpose of the present study is to determine the role of level of intellectual performance on several executive
and non-executive function tests, in an average, above average, and superior sample over several age ranges. It was
hypothesized, that significant relationships would occur in all or most of the executive function measures in this
pediatric sample spanning a large IQ range. There was no a priori hypothesis regarding non-executive functions or
whether gender relationships qualified the IQ relationships. Another purpose of the study was to provide practical
information, in the form of mean scores, to aid clinicians who see brain injured children who are premorbidly gifted.
Finally, a third purpose was to shed light on the underlying constructs common to intelligence tests and executive
function measures typically used in clinical practice.

1. Method

1.1. Participants

A sample of 45 normal children with Wechsler Intelligence Scale for Children-III (WISC-III) Full-Scale IQ above
130 (mean = 138.8; 20 male), 55 normal children with Full-Scale IQs between 115 and 129 (mean = 122.7; 24 male)
and 48 normal children with Full-Scale IQs between 90 and 114 (mean = 107; 27 male) were administered the tests.
Participants were normal and gifted school children from four different school districts in a north-central metropolitan
city, ranging in age from 6 to 15. There were 13 six-year olds, 10 seven-year olds, 18 eight-year olds, 28 nine-year
olds, 22 ten-year olds, 19 eleven-year olds, 21 twelve-year olds, 13 thirteen-year olds, 11 fourteen-year olds, and 3
fifteen-year olds. Eighty-eight percent were White, one was African American, and five were Asian Americans; there
were three non-White in the average IQ group, and one in each of the other groups. Six and 14-year olds did not
complete the Design Fluency measure.
S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978 971

1.2. Materials

1.2.1. Executive functioning tests


The Wisconsin Card Sorting Test, the Trail Making Test (Trails), the Stroop Color and Word Test (Stroop), the
controlled Oral Word Association Test (COWAT), and the Design Fluency Test (DF) were used as neuropsychological
tests of executive function. These tests are commonly used clinical measures typically classified as “executive” in
nature.

1.2.1.1. The Wisconsin Card Sorting Test ( Heaton, Chelune, Talley, Kay, & Curtis, 1993). This standard clinical
measure requires participants to match cards by color, shape, and number to four key cards. Examinees are not told
how to sort the cards, but must determine the correct category from the feedback from the examiner. The category
changes periodically. The six sort condition as well as the examiner administered condition were used. The WCST
was scored for total number of categories, failures to maintain set, Perseverative Errors (PE), Non-Perseverative Errors
(NPE), total errors, and trials to complete first category according to guidelines in the manual (Heaton et al., 1993).

1.2.1.2. The Trail Making Test ( Reitan & Davidson, 1974; Rourke & Finlayson, 1978). This test is a downward
extension of a commonly used adult measure. It consists of two parts. Part A requires the examinee to draw pencil lines
connecting numbers 1–15 randomly positioned on an 8.5 × 11 sheet of paper. Part B requires participants to alternate
between numbers and letters in order. Total time to completion is recorded, which includes time taken for the examiner
to point out mistakes.

1.2.1.3. The Stroop Color and Word Test ( Golden, 1978). This test has three trials. First, the participant rapidly reads
color words. Secondly, the participant rapidly names the color of “Xs” printed in colored ink. Thirdly, the participant
must say the color of ink words are printed in. For example, examinees must say “green” if the ink is green but the
printed word is “blue”. All trials have a 45 s time limit, errors are corrected, and total score is recorded.

1.2.1.4. The controlled Oral Word Fluency Test ( Ruff, Light, Parker, & Levin, 1996). This test requires participants
to produce as many words beginning with the letters, F, A, S respectively in 60 s.

1.2.1.5. The Design Fluency Test ( Jones-Gotman & Milner, 1977). This test consists of two conditions, only the fixed
condition being used. In the fixed condition, subjects were asked to generate as many different figures as possible, each
contained in two vertical parallel lines, for 4 min.

1.2.2. Non-executive measures


The non-executive measures selected (Stroop Color and Word conditions), Underlining conditions, Rey Auditory
Verbal Learning Test (RAVLT), Rey Complex Figure (RCF), and Trails A are believed to tap more basic attention
functions or memory. Since these functions may be correlated with executive ability, any differences between the two
categories of tests would be especially robust. Trails and Stroop conditions are described above.

1.2.2.1. Rey Auditory Verbal Learning Test ( Wiens, McMinn, & Crossen, 1988). This is verbal list learning task where
a word list is repeated five times and the participant is required to recall as many of the words as possible, in any order.
There are also interference and delayed recall conditions. The composite score for all five trials was used.

1.2.2.2. Rey Complex Figure ( Knight & Kaplan, 2003). The RCF has accumulated a considerable literature as a
test of visual-spatial perception/construction and memory. The relationship to executive function has been considered
although this is more for the organization condition rather than the copy condition. A complex geometric form is
presented to the participants who are instructed to draw the design on their paper as accurately as possible.

1.2.2.3. Underlining Test, condition 2, 4 and 9 (U2, U4, U9, Rourke & Orr, 1977). These tests require a subject
to rapidly underline a target presented at the top of the page, picking it out among distractors. Condition 2 requires
the participant to rapidly underline a Greek Cross, ignoring seven other geometric forms randomly spaced across an
8.5 × 11 in paper within 30 s. Condition 4 requires the participant to rapidly underline a complex form consisting of a
972 S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978

square within a square, ignoring other geometric forms, differing in spatial presentation, randomly spaced across an
8.5 × 11 in paper within 60 s. Condition 9 requires the participant to rapidly underline the letter combination “fsbm”,
ignoring distractors of these same letters in different order randomly spaced across an 8.5 × 11 in paper within 60 s.
The total score is correct targets minus incorrect targets.

1.2.3. Intelligence and achievement tests


The Wechsler Intelligence Scale for Children-III (Wechsler, 1991), Wide Range Achievement Test Reading and
Math (WRAT-3, Stone, Jastak, & Wilkinson, 1995) were given according to standard instructions. The achievement
measures were included as they classically are highly related to intelligence and serve as useful contrast with the other
measures (Guilford & Fruchter, 1978).

1.3. Procedure

The children and their parents were presented with IRB approved informed consents and had agreed to participate
in this study. Children were carefully screened for neurological dysfunction, learning disability, emotional distur-
bance, and attention deficits through a telephone interview prior to testing. The above measures were administered in
counterbalanced order along with a graduate student according to standard instructions.

1.4. Data analysis

Raw scores were used for all data analyses. In preliminary analysis, the partial correlations controlling for age were
similar for Full-Scale, Verbal, and Performance IQ’s, and the Full-Scale IQ was adopted as the IQ index in multiple
regression analyses. In a review of raw data, there was not sufficient variance in the WCST number of categories
(most children achieved six categories), Trials to Criterion, and Failure to Maintain Set, and these were removed from
further analyses. Multiple regression was used to assess the contribution of general intelligence to both executive
and non-executive function measures, as well as gender relationships. Multiple regression is an excellent method to
determine the overall co-relationship of intelligence and neuropsychological measures. However, in clinical practice,
it is sometimes important to know whether gifted children perform like non-gifted children on neuropsychological
exam. One such scenario is the neuropsychologist examining for residual of mild concussion in a premorbidly gifted
child. MANCOVA was therefore utilized to determine whether the significant IQ relationships were revealed when
groups were clustered in traditional IQ levels (average, above average/superior, and superior/gifted). Mean scores will
be reported for those measures in which the gifted children are significantly superior.
A final goal of the study is to further delineate the nature of significant relationships in intelligence and neuropsy-
chological measures. This was largely an exploratory endeavor, with no a priori hypotheses. Since intelligence is a
construct consisting of a number of different abilities and factors, the relationships of the subtests of the WISC-III with
each of the executive and non-executive function measures were explored in a preliminary analysis. The pattern of
relationships suggested that a composite based on Arithmetic, Coding, and Digit span (which was termed a “working
memory” composite), a composite based on Information, Similarities, Vocabulary, and Comprehension (which was
termed “verbal comprehension”), and a composite based on Picture completion, Picture arrangement, Block Design,
and Object Assembly (which was termed “perceptual organization”) the most efficient to use in further analyses. These
composite group titles are the same derived from factor analytic studies (Kaufman, 1994), but the working memory
composite is actually more akin to the WISC-R Freedom from Distractibility factor rather than the WISC-III working
memory factor. The Symbol Search subtest was not administered routinely.

2. Results

2.1. Stepwise multiple regression

To further assess the contribution of general intelligence to both executive and non-executive function measures, 16
stepwise multiple regression were calculated. All age regressions were significant, but are not reported as gender and
IQ relationships were the subject of interest. Because computation of separate multiple analyses raises the possibility
of achieving significance by chance occurrence, a Bonferroni alpha adjustment was used as a more stringent test of F
S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978 973

Table 1
Multiple regression results for executive function, non-executive function and achievement tests
Gender Age IQ

p R2 p R2 R2 change p R2 R2 change

Executive function tests


Wisconsin Card Sort
PEa ns .05 .04 .001 .14 .10
NPEb ns .02 .043 <.001 .15 .12
Stroop Color-Word Test
Stroop CWc .038 .034 <.001 .45 .43 <.001 .55 .09
Fluency Tests
COWATd ns <.001 .42 <.001 .52 .10
DFe ns <.001 .49 .05 .34 .12
Trail Making Test
B ns .00 .16 ns
Non-executive function tests
Stroop Color-Word Test
Word ns <.001 .58 <.001 .62 .04
Color ns <.001 .44 <.001 .49 .05
Memory tests
RAVLTf .006 .09 .004 .13 .04 ns
Trail Making Test
A ns .004 .08 ns
Rey Complex Figure Test
Copy ns <.001 .28 <.001 .46 .18
Underlining Test
Condition 2 ns <.001 .44 <.001 .50 .07
Condition 4 ns <.001 .17 .003 .25 .09
Condition 9 ns <.001 .41 <.001 .49 .09
Wide Range Achievement Test 3
Reading ns <.001 .13 <.001 .41 .28
Math ns <.001 .23 <.001 .37 .14
a Perseverative Errors.
b Non-Perseverative Errors.
c Color Word.
d Controlled Oral Word Fluency Test.
e Design Fluency.
f Rey Auditory Verbal Learning Test.

level (see Table 1). Six multiple regressions were calculated to determine the relationship between gender, age, and
intelligence to executive tests WCST Perseverative and Non-Perseverative Errors, Stroop Color-Word condition Total,
Oral Word Fluency Total, Design Fluency 4 min fixed condition total, and Trail B Total Time. Raw scores of each of
the six variables were used, first using gender, age, then the WISC-III Full-Scale raw scores, in the stepwise formula.
Age and WISC-III raw scores were entered as continuous variables. The stepwise regression for WCST Perseverative
Errors revealed a significant R2 change, F(1, 96) = 11.32, p < .001. Intelligence accounted for 10% of the variance. The
stepwise regression for WCST Non-Perseverative Errors revealed a significant R2 change, F(1, 95) = 14.18, p < .001.
Intelligence accounted for 12% of the variance. The stepwise regression for Stroop Color-Word condition revealed a
significant R2 change for intelligence and gender, F(1, 123) = 25.87, p < .001, and F(1, 125) = 4.45, p < .01, respectively.
Intelligence accounted for 9% of the variance and gender for 4% of the variance in total scores. The stepwise regression
for Oral Word Fluency total score revealed a significant R2 change, F(1, 132) = 26.96, p < .001. Intelligence accounted
for 10% of the variance. The stepwise regression for Design Fluency total score revealed a significant R2 change, F(1,
974 S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978

86) = 15.15, p < .001. Intelligence accounted for 12% of the variance. The stepwise regression for Trails B total time
revealed was non-significant.
Eight multiple regressions were calculated to determine the relationship between gender, age, and intelligence to
non-executive tests, Stroop Word and Color conditions, Rey Auditory Verbal Learning Total score, Trails A total
score, Rey Complex Figure total score, and the three Underlining Test condition totals. The stepwise regression for the
Stroop Word and Color conditions revealed significant R2 changes, F(1, 123) = 12.28, p < .001, and F(1, 123) = 12.53,
p < .001, respectively. Intelligence accounted for 4% of the variance the word, and 5% of the variance in the color
condition. The stepwise regression for Rey Auditory Verbal Learning Total score revealed a significant R2 change
for gender, but not intelligence, F(1, 78) = 7.86, p < .005. Gender accounted for 9% of the variance. The stepwise
regression for underlining conditions, 2, 4, and 9 significant R2 changes, F(1, 108) = 36.33, p < .001, F(1, 99) = 12.64,
p < .001, F(1, 95) = 15.6, p < .001, respectively. Intelligence accounted for 7% in Underlining #2 scores, and for 9%
of the variance in each of the two other conditions. The stepwise regression for Trails A Total Time revealed was
non-significant. Finally two multiple regressions were calculated to determine the relationship between gender, age,
and IQ to achievement functions using the WRAT-3 Reading and Math subtests. The stepwise regression for WRAT-3
Reading Total score revealed a significant R2 change, F(1, 96) = 11.32, p < .001. Intelligence accounted for 28% of
the variance. The stepwise regression for WRAT-3 Math Total revealed a significant R2 change, F(1, 115) = 55.74,
p < .001. Intelligence accounted for 14% of the variance.

2.2. MANCOVA

Three separate MANCOVA’s were computed to maximize loss from missing data and ensure larger sample sizes.
Age was used as the covariate (see Table 2). Only the variables found to have a significant intelligence relationship
on multiple regression were assessed. One MANCOVA yielded a significant main effect for intelligence for WRAT-3
Reading (F = 18.8 (2, 112), p < .001), WRAT-3 Math (F = 10.0 (2, 112), p < .001), Stroop Color-Word Score (F = 7.89 (2,
112), p < .001), Stroop Color Score (F = 5.35 (2, 112), p < .01), Oral Word Fluency Score (F = 14.78 (2, 112), p < .001),
and Rey Complex Figure Score (F = 12.42 (2, 112), p < .001). Stroop Word condition was non-significant (F = 2.6 (2,
122), p > .05). A second MANCOVA included the WCST variables in one analysis, yielding a significant main effect
for intelligence for Perseverative Errors (F = 5.14 (2, 104), p < .01), and Non-Perseverative Errors (F = 5.52 (2, 104),
p < .005). A third MANCOVA included all other measures, resulting in a significant main effect for intelligence for
Design Fluency (F = 10 (2, 83), p < .001), but not for any of the Underlining conditions 2, 4, and 8 (F = 1.73; 1.11, 1.5
(2, 88), p > .05, respectively). In Table 3 the means of the gifted children for each of the neuropsychological measures
in which they were significantly superior is listed.

2.3. Exploratory multiple regression with WISC-III composites

Multiple regressions were recomputed using a stepwise approach, adding age, then a composite based on Arith-
metic, Coding, and Digit span (“working memory”), a composite based on Information, Similarities, Vocabulary, and
Comprehension (“verbal comprehension”), and a composite based on Picture completion, Picture arrangement, Block
Design, and Object Assembly (“perceptual organization”).
The first composite, working memory, accounted for a significant amount of the variance in the WCST PE and
NPE conditions, the Stroop Color-Word condition, and the DF condition, but the verbal comprehension or percep-
tual organization factors did not add information. Both the working memory and verbal comprehension composites
accounted for a significant amount of the variance in the OWF test. For the non-executive measures, when entered first,
the working memory composite accounted for the only significant amount of intelligence test variance for the Stroop
Word and Color condition, the RAVLT, and U8. For the other two underlining conditions, both the working memory
and the perceptual organization composites contributed significantly, and for the RCF, only the perceptual organization
composite was significantly related.

3. Discussion

Of the six executive function indices subjected to multiple regression, five were significantly related to intelligence
in a regression analysis. The amount of the variance contributed by intelligence to executive function is modest but
S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978 975

Table 2
MANCOVA results (age as covariate) for executive and non-executive function tests
IQ

p F LSD post hoc analyses

Executive function tests


Wisconsin Card Sort
PEa .007 5.1 Gifted group superior to above average group
NPEb .01 4.9 Gifted group superior to above average group
Stroop Color-Word Test
CW totalc .03 3.6 Gifted group superior to both groups
Fluency Tests
COWATd .05 4.3 Gifted group superior to both groups and above average superior to average
DFe .05 5.8 Gifted group superior to both groups and above average superior to average
Trail Making Test
B ns
Non-executive function tests
Stroop Color-Word Test
Word ns
Color ns
Memory tests
RAVLTf .002 6.8 Above average group superior to average
Trail Making Test
A ns
Rey Complex Figure Test
Copy .002 7.1 Above average group superior to average
Underlining Test
Condition 2 ns
Condition 4 ns
Condition 9 ns
Achievement Tests
WRAT-3g Reading ns
WRAT-3 Math .001 5.6 Gifted group superior to both groups and above average superior to average
a Perseverative Errors.
b Non-Perseverative Errors.
c Color Word.
d Controlled Oral Word Fluency Test.
e Design Fluency.
f Rey Auditory Verbal Learning Test.
g Wide Range Achievement Test—3rd Edition.

statistically significant. Full-Scale intelligence accounted for 9–12% of the variance in WCST errors, Stroop Color-
Word inhibition, and verbal and non-verbal fluency measures. Trails B were unrelated to IQ. These results support the
previous research of Ardila et al. (2000), Arffa et al. (1998), Baron (2003) and Mahone et al. (2002) who found that
executive functions are significantly related to intelligence, especially when the sample contains high IQ individuals,
and when there is adequate variance in the raw data.
The WRAT-3 Reading and Arithmetic achievement measures, consistent with previous research, show a strong
correlation with Full-Scale IQ, accounting for 14–28% of the variance in IQ scores. Of the eight non-executive
neuropsychological test conditions, six were significantly related to intelligence in a regression analysis. The amount
of variance contributed by intelligence was more variable than in the executive function analyses, also modest and
ranging from 4 to 18%. Aside from RCF which accounted for the 18% of variance, the relationships were generally
more modest than that on the executive function tests.
976 S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978

Table 3
Mean scores for gifted children by age on neuropsychological measures with significant effects of intelligence
Age Test

N WCSTa Perseverative WCST Non-Perseverative Stroop Color-Word Controlled Oral Word Design Fluency fixed
Errorsb Errorsc Numberd Fluency Numbere condition numberf

6 3 9.5 9.5 27 13.3


7 3 12.5 14 21 21.3 6.5
8 5 8.5 6.7 30.2 27.8 11.0
9 12 5 6 32.1 27.8 14.3
10 8 7.6 6.4 38.6 35.6 17
11 5 5 3.2 35.6 36.6 17.7
12 3 6.6 6.3 50.3 36.3 18.7
13 4 8.6 10.7 50 45.3 18
14 2 4 6 52 63
a Wisconsin Card Sorting Test.
b A MANCOVA with age as covariate yielded a significant IQ condition and 130–150 group is significantly superior to 115–129 group and 90–114
group at .05 level based upon LSD post hoc tests.
c A MANCOVA with age as covariate yielded a significant IQ condition and 130–150 group is significantly superior to 115–129 at .05 level based

upon LSD post hoc test


d A MANCOVA with age as covariate yielded a significant IQ condition F = 3.5 (2, 112), p < .05. The 130–150 group was significantly superior

to 115–129 group and 90–114 group at .05 level based upon LSD post hoc tests.
e A MANCOVA with age as covariate yielded a significant IQ condition F = 8.5 (2, 112), p < .001. The 130–150 group was significantly superior

to both lower IQ groups and 115–129 group was superior to 90–114 group at .05 level based upon LSD post hoc tests.
f A MANCOVA with age as covariate yielded a significant IQ condition F = 10.75 (2, 84), p < .001. The 130–150 group was significantly superior

to both lower IQ groups and 115–129 group is superior to 90–114 group at .05 level based upon LSD post hoc tests.

Clearly, achievement is more strongly related to IQ than are either executive or non-executive function tests. However,
executive function measures were slightly more robust in their relationship to IQ than the non-executive measures.
IQ should be a major consideration when performing clinical evaluations with premorbidly gifted children. And
gifted children performed significantly better than the average and in some cases both the average and above average
groups on MANCOVA’s. In contrast to the executive tests, gifted children did not perform better than above average and
average children on non-executive tests in a MANCOVA, although the above average group outperformed the average
group. When using MANCOVA, the Stroop Word condition and the Underlining conditions were non-significant.
The gifted children did not consistently show superior performance on non-executive measures compared to executive
function measures. On the RCF, and RAVLT, above average children outperformed average children but gifted children
did not outperform above average children. This may have been due to a test ceiling effect. On the RCF, for example,
both gifted and above average adolescents performed close to the top score. While it is important to consult manuals
for norms appropriate to gifted children, caution should be taken when using certain extrapolated or smoothed norms.
For example, children 8 years and older with 145 IQs should have four or fewer total errors on the WCST and a 14
year old would have only two errors, according to Heaton et al. (1993) manual. However, some errors are necessary
(usually 5–6) in order to shift sets, and the data from the current sample (Table 2) confirm that actual error rates are
much higher.
Since intelligence is a multidimensional construct, and IQ subtests tap a variety of mental abilities, further analyses
were performed to further delineate the relationship of IQ to executive and attention and memory tasks. Most executive
function measures and a number of attention measures were more related to subtests that comprise the Freedom
from Distractibility and Processing Speed factors of the WISC-III. The Arithmetic, Digit Span, and Coding subtests
clustered together well in preliminary analysis and were considered in one composite index termed working memory.
This working memory index accounted for most of the variance (attributable to IQ) on all the executive measures except
OWF, which also was related to the verbal comprehension composite. Working memory also accounted for most of
the variance on the non-executive attention and memory tasks, although there was more variability with this group of
measures, and the perceptual organization composite figured highly as well. As a comparison, the achievement subtests
appear to be related more generally to overall IQ, showing significant correlations with many WISC-III subtests. Clearly,
most of the significant intelligence test variance for many executive function measures lies with subtests Arithmetic,
S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978 977

Digit Span, and Coding, at least in a child population. This supports the work of Friedman et al. (2006) who found
IQ in adults was related to executive function tests which require an individual to process relevant information, hold
it in mind, manipulate, and reproduce in an orderly, efficient manner. Curiously, those functions considered “higher
level” executive functions like verbal and non-verbal reasoning, problem solving, and abstraction, may be less related
to psychometric intelligence than previously thought (Seidenberg et al., 1983), at least in children.
Since it is theoretically possible that gifted children approach solving neuropsychological tasks in a different
manner than other children, correlation and multiple regressions were explored for this subgroup only. This procedure
significantly reduces the variance in the database, making direct comparisons difficult. However, aside from the WCST
variables, the strongest relationship between neuropsychological measures and IQ remained the working memory
composite. The WCST test/IQ test relationship was not significant, likely because the variance in the gifted group was
seriously reduced, as even young children performed at optimal levels. It is likely that gifted children employ the same
cognitive strategies in solving executive function tasks than do average to superior children.
These results are relevant to current research in executive function in ADHD populations. Executive function tests
have not reliably discriminated ADHD populations often because significant relationships are usually removed when IQ
is controlled for (Scheres et al., 2004; Sergeant, Geurts, & Oosterlaan, 2003; Riccio, Homack, Pizzitola-Jaratt, & Wolfe,
2006; Roodenrys, 2006). This has led researchers to posit that executive dysfunction may not be impaired in ADHD
children. However, since it is reported that subtests of Digit Span, Coding, and Arithmetic may be depressed in some
ADHD populations, ostensibly because these subtests tap working memory (Anastopoulos, Spisto, & Maher, 1994),
and working memory may the primary source of the covariance between IQ and executive function, then removing IQ
from discrimination analysis in ADHD children removes the variance that would most reliably discriminate ADHD
children from other populations (on executive function variables). It is recommended that any statistical equations for
IQ in studies comparing subgroups on executive function measures, should probably be based on a composite of the
verbal and performance subtests, and not those that comprise the Freedom from Distractibility or Perceptual Speed
factors.
An attempt was made to statistically control for the numerous multiple regressions, and to perform analyses sequen-
tially based upon hypotheses. However, the multiple analyses which raise the risk of significance by chance remains a
limitations of this study. In addition, clinical measures are not as specific as laboratory measures, and future research
could be improved by using both in future research.

Acknowledgement

This research was supported by a grant from the Allegheny Singer Research Foundation.

References

Anastopoulos, A. D., Spisto, M. A., & Maher, M. C. (1994). The WISC-III freedom from distractibility factor: Its utility in identifying children with
attention deficit hyperactivity disorder. Psychological Assessment, 6, 368–371.
Ardila, A., Pineda, D., & Rosselli, M. (2000). Correlation between intelligence test scores and executive function measures. Archives of Clinical
Neuropsychology, 15(1), 31–36.
Arffa, S., Lovell, M., Podell, K., & Goldberg, E. (1998). Wisconson Card Sorting Test performance in above average and superior school children:
Relationship to intelligence and age. Archives of Clinical Neuropsychology, 13, 713–720.
Baron, I. (2003). Neuropsychological evaluation of the child. New York, NY: Oxford University Press.
Boone, K. B., Ghaffarian, S., Lesser, I. M., & Hill-Gutierrez, E. (1993). Wisconsin Card Sorting Test performance in healthy, older adults: Relationship
to age, sex, education, and IQ. Journal of Clinical Psychology, 49(1), 54–60.
Damasio, A. R., & Anderson, S. W. (1993). The frontal lobes. In K. M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology (3rd ed., pp.
409–460). New York: Oxford University Press.
Donders, J., & Kirsch, N. (1991). Nature and implications of selective impairment on the Booklet Category Test and Wisconsin Card Sorting Test.
Clinical Neuropsychologist, 5(1), 78–82.
Eslinger, P. J. (1996). Conceptualizing, describing, and measuring components of executive function: A summary. In G. Reid Lyon, A. Norman, &
Krasnegor (Eds.), Attention, memory, and executive function (pp. 263–278). Baltimore, MD: Paul H. Brookes.
Friedman, N. P., & Miyake, A. (2004). The relations among inhibition and interference control functions: A latent-variable analysis. Journal of
Experimental Psychology: General, 133, 101–135.
Friedman, N. P., Miyake, A., Corley, R. P., Young, S. E., DeFries, J. C., & Hewitt, J. K. (2006). Not all executive functions are related to intelligence.
Psychological Science, 17(2), 172–179.
Golden, C. J. (1978). Stroop color and word test. Wood Dale, IL: Stoelting Co.
978 S. Arffa / Archives of Clinical Neuropsychology 22 (2007) 969–978

Guilford, J. P., & Fruchter, B. (1978). Fundamental statistics in psychology and education (6th ed.). New York: Mc-Graw-Hill.
Heaton, R. K., Chelune, G. J., Talley, J. L., Kay, G. G., & Curtiss, G. (1993). Wisconsin Card Sorting Test manual: Revised and expanded Psychological
Assessment Resources. Odessa, FL: Psychological Assessment Resources.
Johnstone, B., Holland, D., & Larimore, C. (2000). Language and academic abilities. In G. Groth-Marnat (Ed.), Neuropsychological assessment in
clinical practice: A guide to test interpretation and integration (pp. 335–354). New York, NY: John Wiley & Sons, Inc.
Jones-Gotman, M., & Milner, B. (1977). Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia, 15,
653–674.
Kaufman, A. S. (1994). Intelligent testing with WISC-III. New York: Wiley.
Knight, J. A., & Kaplan, E. (2003). The Handbook of Rey-Osterrieth complex figure usage: Clinical and research applications. New York, NY: John
Wiley & Sons, Inc.
Lezak, M. D., Howieson, D. B., Loring, D. W., & Hannay, H. J. (2004). Neuropsychological assessment (4th ed.). Oxford University Press.
Mahone, E. M., Hagelthorn, K. M., Cutting, L. E., Schuerholz, L. J., Pelletier, S. F., Rawlins, C., et al. (2002). Effects of IQ on executive function
measures in children with ADHD. Psychology Press, 8(1), 52–65.
Milner, B. (1982). Some cognitive effects of frontal-lobe lesions in man. Philosophical Transactions of the Royal Society of London, 298, 211–226.
Parsons, E. J. C. (1984). An investigation of appropriate neuropsychological assessment procedures with adults in the higher intelligence ranges.
Dissertation Abstracts International, 45(5), 1592B.
Reitan, R. M., & Davidson, L. A. (Eds.). (1974). Clinical neuropsychology: Current status and applications. (pp. 53–89). New York: John Wiley &
Sons.
Riccio, C. A., Hall, J., Morgan, A., & Hynd, G. W. (1994). Executive function and the Wisconsin Card Sorting Test: Relationship with behavioral
ratings and cognitive ability. Developmental Neuropsychology, 10(3), 215–229.
Riccio, C. A., Homack, S., Pizzitola-Jarratt, K., & Wolfe, M. E. (2006). Differences in academic and executive function domains among children
with ADHD predominantly inattentive and combined. Archives of Clinical Neuropsychology, 21(7), 657–667.
Roodenrys, S. (2006). Working memory function in attention Deficit Hyperactivity disorder. In T. P. Alloway & S. E. Gathercole (Eds.), Working
memory and neurodevelopment disorders. New York: Psychology Press.
Rourke, B. P., & Finlayson, M. A. J. (1978). Neuropsychological significance of variations in patterns of academic performance: Verbal and
visual-spatial abilities. Journal of Abnormal Child Psychology, 6, 1.
Rourke, B. P., & Orr, R. R. (1977). Prediction of the reading and spelling performances of normal and retarded readers: A four-year follow-up.
Journal of Abnormal Child Psychology, 18, 1–21.
Ruff, R. M., Light, R. H., Parker, S. B., & Levin, H. S. (1996). Benton Controlled Oral Word Association Test: Reliability and updated norms.
Archives of Clinical Neuropsychology, 11(4), 329–338.
Scheres, A., Oosterlaan, J., Geurts, H., Morein-Zamir, S., Meiran, N., Schut, H., et al. (2004). Executive functioning in boys with ADHD: Primarily
an inhibition deficit? Archives of Clinical Neuropsychology, 19, 569–594.
Seidenberg, M., Giordani, B., Berent, S., & Boll, T. (1983). IQ level and performance of the Halstead–Reitan Neuropsychological Test battery for
older children. Journal of Consulting and Clinical Psychology, 51(3), 406–413.
Sergeant, J. A., Geurts, H., & Oosterlaan, J. (2003). How specific is a deficit of executive functioning for Attention-Deficit/Hyperactivity Disorder?
Behavioural Brain Research, 130, 2–3.
Stone, M. H., Jastak, S., & Wilkinson, G. (1995). Wide Range Achievement Test WRAT. Wilmington: Delaware, Jastak Associates.
Waldmann, B. W., Dickson, A. L., Monahan, M. C., & Kazelskis, R. (1992). The relationship between intellectual ability and adult performance on
the Trail Making Test and the Symbol Digit Modalities. Journal of Clinical Psychology, 48(3), 360–363.
Wechsler, D. (1991). Wechsler Intelligence Scale for Children-III. San Antonio, TX: Psychological Corp.
Welsh, M. C., Pennington, B. F., & Groisser, D. B. (1991). A normative-developmental study of executive function: A window on prefrontal function
in children. Developmental Neuropsychology, 7(2), 131–149.
Wiens, A. N., & Matazzaro, J. D. (1988). WAIS and MMPI correlates of the Halstead–Reitan neuropsychological battery in normal male subjects.
Journal of Nervous and Mental Diseases, 164, 112–123.
Wiens, A. N, McMinn, M. R., & Crossen, J. R. (1988). Rey Auditory Verbal Learning Test: Development of norms for healthy young. The Clinical
Neuropsychologist.

You might also like