online © ML Comm

ORIGINAL ARTICLE Print ISSN 1738-3684 / On-line ISSN 1976-3026

DOI 10.4306/pi.2010.7.3.177 OPEN ACCESS

Impairment of Concept Formation Ability in Children with ADHD:

Comparisons between Lower Grades and Higher Grades
Hye Jeong Hong, Jong Bum Lee, Jin Sung Kim, Wan Seok Seo,
Bon Hoon Koo, Dai Seg Bai and Jin Young Jeong
Department of Psychiatry, College of Medicine, Yeungnam University, Daegu, Korea

ObjectiveaaWe investigated executive functions (EFs), as evaluated by the Wisconsin Card Sorting Test (WCST), and other EF between
lower grades (LG) and higher grades (HG) in elementary-school-age attention deficit hyperactivity disorder (ADHD) children.
MethodsaaWe classified a sample of 112 ADHD children into 4 groups (composed of 28 each) based on age (LG vs. HG) and WCST
performance [lower vs. higher performance on WCST, defined by the number of completed categories (CC)] Participants in each group
were matched according to age, gender, ADHD subtype, and intelligence. We used the Wechsler intelligence Scale for Children 3rd edi-
tion to test intelligence and the Computerized Neurocognitive Function Test-IV, which included the WCST, to test EF.
ResultsaaComparisons of EFs scores in LG ADHD children showed statistically significant differences in performing digit spans back-
ward, some verbal learning scores, including all memory scores, and Stroop test scores. However, comparisons of EF scores in HG ADHD
children did not show any statistically significant differences. Correlation analyses of the CC and EF variables and stepwise multiple regres-
sion analysis in LG ADHD children showed a combination of the backward form of the Digit span test and Visual span test in lower-
performance ADHD participants significantly predicted the number of CC (R2=0.273, p<0.001).
ConclusionaaThis study suggests that the design of any battery of neuropsychological tests for measuring EF in ADHD children should
first consider age before interpreting developmental variations and neuropsychological test results. Researchers should consider the dynam-
ics of relationships within EF, as measured by neuropsychological tests. Psychiatry Investig 2010;7:177-188

Key WordsaaAttention deficit hyperactivity disorder, Wisconsin Card Sorting Test, Concept formation ability, Age, Developmental variation,
Working memory.

INTRODUCTION propriate problem-solving set to attain a later goal. However,

moderate effect size (ES) and lack of universal EF deficits among
The search for accurate and reliable measures of attention-de- ADHD individuals suggest that EF weaknesses are neither
ficit hyperactivity disorder (ADHD) symptoms has not yield- necessary nor sufficient to cause ADHD,2 and furthermore, EF
ed a litmus test, and the heterogeneity of the disorder itself pre- deficits are typical of developmental disorders in general.3 EF
clude any one test from claiming pinpoint accuracy, and AD- develops throughout childhood and adolescence and plays an
HD-related evaluation incorporates a wide variety of psycho- important role in a child’s cognitive functioning, behavior, emo-
logical and neuropsychological tests.1 One of the most pro- tional control, and social interaction. Attentional control appears
minent neuropsychological theories of ADHD suggests that to emerge in infancy and to develop rapidly in early childhood.
its symptoms arise from a primary deficit in executive functions In contrast, cognitive flexibility, goal setting, and information
(EF), defined as neurocognitive processes that maintain an ap- processing experience a critical period of development, between
7 and 9 years of age, and are relatively mature by 12 years of
Received: February 11, 2010 Revised: April 10, 2010
Accepted: May 30, 2010 Available online: July 9, 2010 age.4 Furthermore, EF, as a concept, has had serious problems,5
 Correspondence: Jong Bum Lee, MD, PhD and it is not single skill. As a result, an individual’s ability to
Department of Psychiatry, College of Medicine, Yeungnam University, 317-1 perform the non-executive requirement of a task can contami-
Daemyeong-dong, Nam-gu, Daegu 705-717, Korea
Tel: +82-53-620-3340, Fax: +82-53-629-0256, E-mail: [email protected] nate that individual’s performance on EF tasks.6 In addition
cc This is an Open Access article distributed under the terms of the Creative Commons
to showing the overall reliability and validity of EF tests’ re-
Attribution Non-Commercial License (http://creativecommons.org/licenses/by-
nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduc- sults, it is important to show that not only do specific execu-
tion in any medium, provided the original work is properly cited. tive-function tasks differentiate clinical from non-clinical pop-

Copyright © 2010 Korean Neuropsychiatric Association 177

Impairment of Concept Formation Ability in Children with ADHD

ulations, but they effectively discriminate among various clini- Mental Disorders, 4th edition, Text Revision (DSM-IV-TR)
cal populations.7 criteria.14 Our inclusion criteria were diagnosis of ADHD and
The facts that not every person with ADHD is impaired age 7-12 years. Institutional Review Board reviewed and ap-
on every test and some individuals with ADHD perform with- proved the protocol. With regard to ADHD diagnosis, at least
in the normal range on all or most measures demonstrate AD- two clinicians interviewed the patients to ensure that they met
HD heterogeneity. They also show potential moderators of AD- DSM-IV-TR diagnostic criteria for ADHD, as evidenced by
HD neuropsychological heterogeneity are family history, co- the presence of at least 6 of 9 inattention symptoms (with or
morbid disorders, ADHD subtypes, and developmental differ- without 6 of 9 hyperactive/impulsive symptoms) and as as-
ences.8 Heterogeneity in ADHD9-11 is evident in the expression certained in a semi-structured interview, using a DSM symp-
of the two ADHD symptom domains: neuropsychological im- tom checklist. The study excluded participants who had an in-
pairments and comorbid behavior problems.12 Wåhlstedt et telligence quotient (IQ) below 70, by the Wechsler Intelligence
al.12 pointed to the importance of viewing ADHD as a hetero- Scale for Children-3rd edition (WISC-III);15,16 a prior history
geneous condition with regard to both neuropsychological of more than 1 month of ADHD medication treatment; a pres-
functioning’s and comorbidity’s differential impacts on differ- ent or past history of substance abuse or addiction (except nic-
ent ADHD symptom groups and the two ADHD symptom otine abuse/addiction); a present or past history of psychiatric
domains. disease (axis I or II diagnosis other than ADHD); a neurologi-
With regard to developmental differences, although there are cal disease; a medical condition that could alter cerebral func-
more than a hundred studies examining neuropsychological tioning (i.e., cardiovascular, endocrinological, oncological, and
functioning in childhood ADHD, there are relatively few stud- autoimmune diseases); and/or head trauma with loss of con-
ies examined such functioning in ADHD preschoolers, ado- sciousness of more than 30 min.
lescents, and/or adults.13 Although many researchers have ex-
tensively studied elementary school-age ADHD children’s neu- Group classification and verification of sampling
ropsychological functioning, such studies have not considered biases
these children’s neurodevelopment characteristics. We classified the participants into four groups, based on
Given that executive processes are dependent on the fron- their number of completed categories (CC) on the WCST and
tal lobe systems’ integrity, it is likely that EF skills demonstrate their age (LG 7-9 years vs. HG 10-12 years). WCST perform-
functional improvements, which research can align with neu- ance classification criteria were age-referenced to WCST norm
ropsychological developments within the prefrontal cortex.4 In values,17,18 and ADHD participants performing lower in one
this study, EFs showed ongoing developmental functioning, or more categories than the age reference norm were the lower
not yet fully developed. Furthermore, some EFs were not yet, or performance group (L-ADHD). ADHD participants perform-
were slowly, developing in lower grades (LG) elementary chil- ing the same as the age-referenced norm were the normal per-
dren as opposed to higher grades (HG) elementary children. formance group (N-ADHD). Finally, we divided participants
Furthermore, we found it suggestive that EF measures had into the LG and HG. We matched numbers of ADHD partici-
more inter- or intra-differences in LG elementary children than pants, mean age, and gender ratio, and ADHD subtype and
HG elementary children. Therefore, we compared EF, espe- verified these by statistical analyses.
cially concept formation ability, evaluated by the Wisconsin There were 28 L-ADHD and 28 N-ADHD participants in
Card Sorting Test (WCST), and established at late childhood, the LG group, and 4 (14.3%) of the participants in each of these
according to neurodevelopment stages of the children. Also, L-ADHD and N-ADHD sections were female. The remainder,
before comparison, we controlled for possible effects of age, 24, were male (85.7%). In the LG group, mean age (8.07±
gender, comorbid disorders, ADHD subtype, and intelligence 0.74) and intelligence (96.71±9.80) did not vary significantly
between the two groups. Finally, we explored the relationships between the L-ADHD and N-ADHD sections.
between WCST performance and other EFs in ADHD children. With regard to the overall ratio of ADHD subtypes, there
were 10 predominantly inattentive-type participants (17.9%);
METHODS the remainder (46) were combined type (82.2%). The subtype
ratio did not vary significantly between L-ADHD and N-AD-
Participants sampling HD participants.
We sampled participants from outpatient units in Y Universi- There were 56 matched participants in the HG group, sel-
ty Hospitals in Korea over a period of 3 years, from March 1, ected for the same ratio as in the LG group (28 L-ADHD; 28 N-
2006 through June 30, 2008. Research psychiatrists diagnosed ADHD), and 2 (7.1%) of the participants in each of these L-
the participants per the Diagnostic and Statistical Manual of ADHD and N-ADHD sections were female. The remainder,

178 Psychiatry Investig 2010;7:177-188

 HJ Hong et al.

26, were male (92.9%). In the HG group, mean age (10.98± In the cognitive flexibility domain, we evaluated conceptual
0.77) and intelligence (98.77±13.28) did not vary significant- transfer ability by the Trail Making Test (TMT) A & B. In the
ly between the L-ADHD and N-ADHD sections. goal setting domain, we evaluated concept formation ability
With regard to the overall ratio of ADHD subtypes, there by the WCST. In the information processing domain, we evalu-
were 22 predominantly inattentive-type participants (60.7%); ated speed of processing by reaction times for each neuropsy-
the remainder (34) were combined type (60.7%). The subtype chological test. Finally, we evaluated verbal and visual work-
ratio did not vary significantly between L-ADHD and N-AD- ing memory (WM) and memory process by Digit span test
HD participants. In comparing the LG to the HG group, the (DST) and Visual span test (VST): forward and backward, the
gender ratio, intelligences, and ratio of ADHD subtypes did Verbal learning test (VLT), and the Visual recognition test (VRT).
not vary significantly; of course, age did (Table 1). The WCST26-29 was originally developed to assess abstrac-
tion ability in normal persons.26,27 However, it is now gaining
Procedures increasing popularity as a clinical neuropsychological instru-
All participants completed the behavioral rating scales, Ko- ment28 and is expanding into childhood uses.29 The test used 8
rean Personality Inventory for Children (K-PIC), K-WISC- main scores: numbers of CC, total trials, numbers to complete
III, and EF tests before their final diagnoses and medications. 1st category, total errors, perseverative responses, persevera-
Three licensed clinical psychologists administered the tests, af- tive errors, non-perseverative errors, maintaining set failure.
ter we randomly, blindly, and singly assigned the participants. Other EFs included the Visual and auditory CPT (VCPT &
ACPT), a psychological test that measures a person’s sustained
Materials and selective attention and impulsivity. There are 4 main sc-
ores. Correct response indicates the number of times the sub-
K-WISC-III15,16 ject responded to the target stimulus (pressed the button at
The WISC was the first published intelligence test by We- number 3). Reaction time measures the amount of time be-
chsler.19 The third edition was published 1991, and the Kore- tween the presentation of the stimulus and the participants’ re-
an version of the WISC-III was published by Kwak et al. 16 This sponse. Omission errors indicate the number of times the tar-
test comprises ten core subtests and five supplemental ones. get was presented, but the subject did not click the response
These subtests then generate Full Scale IQ, Verbal IQ, and key or button. Commission errors indicate the number of times
Performance IQ, as well as four composite scores known as the participant responded but no target was presented (pressed
indices: Verbal comprehension, Perceptual organization, Free- buttons at numbers other than 3). The Stroop test is based on
dom from distractibility, Processing speed. the observation that individuals can read words much faster
than they can identify and name colors. The cognitive dimen-
K-PIC20,21 sion tapped by the Stroop test is associated with cognitive flex-
The PIC is an objective multidimensional test of child and ibility, resistance to interference from outside stimuli, creativ-
adolescent emotional and cognitive status, first published by ity, and psychopathology. The TMT requires a subject to “con-
Wirt et al.20 The Korean version of the PIC was published by nect-the-dots” of 25 consecutive targets on a computer screen.
Kim et al.21 The administrative booklet consists of 255 items Two versions are available: 1) in which the targets are all num-
to be completed by the child’s parents, or some other rater who bers (1, 2, 3, etc.), and 2) in which the participant must alter-
knows the child well. The 16 KPI-C’s subscales consist of 4 nate between numbers and letters (1, A, 2, C, etc.). The goal is
validity scales, the ego resilience scale, and 11 clinical scales to finish the test as quickly as possible, and the primary perfor-
(Verbal, Performance, Anxiety, Depression, Somatization, De- mance metric is the time taken to complete the test. The Digit
linquency, Hyperactivity, Family, Social interaction, Psychoti- span test (DST) is an adaptation from the Wechsler battery (the
cism, and Autism scales). intelligence scale and memory scales), in which the participant
must recall a series of digits of increasing length by touching
Evaluation of EF the digits on a screen, in order. For the reverse digit span, the
We evaluated EF via computerized neuropsychological tests subject is asked to touch the digits on the screen in reverse or-
in the Computerized Neurocognitive Function Test-IV (CNT).22 der. The VST is an adaptation of the Corsi block-tapping test;30
CNT’s reliability and validity for adults,23 children18,24 and peo- after white circles blink in a prearranged sequence on a touch
ple with ADHD25 are well-established. In the attentional con- monitor, the participant must attempt to copy this blinking
trol domain, we evaluated attention ability by a visual and au- pattern. The VLT is a modified Rey Auditory VLT and assess-
ditory continuous performance test (CPT) and inhibition abi- es immediate and delayed verbal memory; the format of the
lity by the Stroop color-word interference test (Stroop test). test follows the Rey Auditory VLT, but the word lists were in

www.psychiatryinvestigation.org 179
Impairment of Concept Formation Ability in Children with ADHD

Korean. The test consists of a learning phase, with 5 presenta- Statistical analysis
tions of a 15-word list (list A), an interference phase, with an- We used the Chi-square test (ADHD, subtypes), Fisher ex-
other 15-word list (list B), a short-term delayed recall (A6) of the act test (gender), and t-test (age, education, IQ, and WCST sc-
first list, and a long-term delayed test, after 20 minutes, which ores) for verification of sampling biases. We used the t-test for
consists of both recall and recognition. Outcome measures in- our analysis of EF scores between the groups and correlation
clude trial A1 (number of words recalled in the first trial), trial analysis and stepwise multiple regression analysis for the re-
A5 (number of words recalled in the fifth trial), distracter recall lationship between WCST and other EFs. We analyzed the
(B6), 6th trial (short-term delayed recall; A6), delayed recall data were analyzed using the Statistical Package for the Social
(A7; recall after 20 minutes), recognition (number of words Sciences (SPSS) for Windows, version 17.0 (SPSS Inc., Chicago,
correctly identified on the cued recall trial), total recall (total IL, USA) with a significance level of 0.05. The ES were calcu-
number of words correct, from 1st to 5th trials), learning index lates according to the method of Cohen. The ES index for the
(A5-A1; number for trial A5 minus number for trial A1), pro- t-test of the difference between independent means is d, the
active interference [(A1-B)/A1] and retroactive interference [(A5- difference repressed in units of (i.e., divided by) the within-
A6)/A5]. The VRT follows the format of the Rey Auditory VLT. population standard deviation. An ES of 0.2 was considered
Fifteen different target figures, composed of circles, squares, of a small effect and an ES of 0.5 was considered a medium ef-
triangles, lines, and dots, display separately for 1 second each, fect, 0.8 was considered a large effect.31
and a participant must select these target figures from a pool
of the target figures and 15 interference figures. The test con- RESULTS
sists of a learning phase, with 5 consecutive presentations of
the 15 figures, and a delayed recognition test after 20 minutes. Comparisons of WCST and K-PIC scores by group
Outcome measures include trial A1 (number of figures recog- classification
nized correctly in the first trial), trial A5 (number of figures re- The LG group’s WCST performance showed the L-AD-
cognized correctly in the fifth trial), delayed recognition (A6; HD’s CC were significantly lower than those of the N-ADHD
recognition after 20 minutes), total recognition (total number (L-ADHD, 3.11±0.79; N-ADHD, 5.64±0.68, t=-12.92; p<
of figures correctly recognized, from 1st to 5th trials), and learn- 0.01; ES=3.43). The L-ADHD section showed significantly
ing index (A5-A1; number for trial A5 minus number for higher performances on total trials (L-ADHD, 128.00±0.00;
trial A1). N-ADHD, 111.61±18.51; t=4.69; p<0.01; ES=8.89), trials to

Table 1. Demographic data

Variables L-ADHD N-ADHD Total
LG Gender
Male (%) 24 (85.7) 24 (85.7) 48 (85.7)
Female (%) 04 (14.3) 04 (14.3) 08 (14.3)
Age (mean±SD) 8.07±0.66 08.07±0.81 08.07±0.74
Intelligence (mean±SD) 95.50±10.99 97.93±8.46 96.71±9.80
Subtype
ADHD, predominantly inattentive type 04 (14.3) 06 (21.4) 10 (17.9)
ADHD, combined type 24 (85.7) 22 (78.6) 46 (82.1)
HG Gender
Male (%) 26 (92.9) 26 (92.9) 52 (92.9)
Female (%) 2 (7.1) 2 (7.1) 4 (7.1)
Age (mean±SD) 10.93±0.770 11.04±0.800 10.98±0.770
Intelligence (mean±SD) 99.25±12.67 99.29±14.08 98.77±13.28
Subtype
ADHD, predominantly inattentive type 11 (39.3) 11 (39.3) 22 (39.3)
ADHD, combined type 17 (60.7) 17 (60.7) 34 (60.7)
SD: standard deviation, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed categories
of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, ADHD: attention
deficit hyperactivity disorder, WCST: Wisconsin Card Sorting Test

180 Psychiatry Investig 2010;7:177-188

 HJ Hong et al.

1st category (L-ADHD, 20.93±23.76; N-ADHD, 7.93±9.81; errors, commission errors, and reaction times for the ACPT
t=2.68; p<0.05; ES=0.72), total errors (L-ADHD, 34.11±11.51; and VCPT. These sections also did not differ significantly in
N-ADHD, 23.00±8.34; t=4.13; p<0.01; ES=1.11) and non- the HG group (Table 4).
perseverative errors (L-ADHD, 9.93±7.63; N-ADHD, 6.00± On the DST, the LG group’s L-ADHD section did not differ
3.46; t=2.48; p<0.05; ES=0.66), but no significant difference significantly from the N-ADHD section on the forward sub-
on the other variables. Meanwhile, the HG group’s WCST per- test, but the L-ADHD section scored significantly lower than
formance showed the L-ADHD’s completed category results the N-ADHD on the backward subtest (L-ADHD, 3.18±0.94;
were significantly lower than those of the N-ADHD (L-ADHD, N-ADHD, 3.82±0.86; t=-2.66; p<0.05; ES=0.71). On the VLT,
2.64±0.95; N-ADHD, 6.00±0.00; t=-18.68; p<0.01; ES=5.00). the groups did not differ significantly on either subtest. The LG
The L-ADHD section also showed significantly higher perfor- group’s L-ADHD section scored significantly lower on the dis-
mances on total trials (L-ADHD, 128.00±0.00; N-ADHD, 108.61 tracter recall (L-ADHD, 3.79±2.17; N-ADHD, 5.00±2.11; t=
±15.67; t=6.55; p<0.01; ES=1.75), trials to 1st category (L- -2.13; p<0.05; ES=0.57), delayed recall (L-ADHD, 7.18±2.31;
ADHD, 21.61±22.21; N-ADHD, 7.07±10.12; t=3.15; p<0.01; N-ADHD, 8.82±2.87; t=-2.36; p<0.05; ES=0.63), delayed
ES=0.84), and failure to maintain set (L-ADHD, 2.54±2.00; recognition (L-ADHD, 11.21±3.10; N-ADHD, 12.82±1.72;
N-ADHD, 1.39±0.96; t=2.74; p<0.01; ES=0.73). However, the t=-2.40; p<0.05; ES=0.64), and total recall, but the sections
groups did not differ significantly in total errors or the other did not differ significantly on the other index. Also, the sec-
variables (Table 2). The LG group’s KPI-C subscale scores tions did not differ significantly with regard to the VRT index.
showed that the groups did not differ significantly on the va- Furthermore, in the HG group, the sections did not differ sig-
lidity and clinical scales. Also, the HG group’s L-ADHD sec- nificantly on forward and backward subtests of the DST, VST,
tion KPI-C scores showed no significant difference from the VRT, and all indices of the VLT (Table 5).
N-ADHD section on the validity and clinical scales (Table 3). On TMT type A and B, the sections did not differ signifi-
cantly in the LG group. Also, there was no significant differ-
Comparisons of EFs scores between L-ADHD and ence in the response time for Word reading, a subtest of the
N-ADHD in LG ADHD and HG ADHD Stroop test, but the L-ADHD section performed significantly
On CPT performance, the LG group’s L-ADHD section did slower than the N-ADHD did on Color reading (L-ADHD,
not differ significantly from the N-ADHD section on omission 32.55±11.55; N-ADHD, 26.49±7.53; t=2.32; p<0.05; ES=

Table 2. Comparisons of WCST performance between L-ADHD and N-ADHD in each age group
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
LG (N=56) Completed categories 003.11±0.79 005.64±0.68 12.923 0.001 3.43
Total trials 128.00±0.00 111.61±18.51 4.686 0.001 8.89
Trials to 1st category 020.93±23.76 007.93±9.81 2.676 0.011 0.72
Total errors 034.11±11.51 023.00±8.34 4.134 0.001 1.11
Perseverative errors 016.04±7.33 014.29±5.38 1.018 0.313 0.27
Non-perseverative errors 009.93±7.63 006.00±3.46 2.479 0.018 0.66
Perseverative responses 027.71±10.52 023.00±9.55 1.757 0.085 0.47
Failure to maintain set 001.82±0.72 001.71±0.94 0.479 0.634 0.13
HG (N=56) Completed categories 002.64±0.95 0006.0±0.00 -18.676 0.001 5.00
Total trials 0128.0±0.00 108.61±15.67 6.548 0.001 1.75
Trials to 1st category 021.61±22.21 007.07±10.12 3.151 0.003 0.84
Total errors 024.25±13.02 023.21±8.60 0.351 0.727 0.09
Perseverative errors 012.11±9.58 014.14±5.91 -0.957 0.344 0.26
Non-perseverative errors 009.25±8.55 007.18±5.02 1.105 0.275 0.30
Perseverative responses 020.96±15.18 021.93±8.51 -0.293 0.771 0.08
Failure to maintain set 002.54±2.00 001.39±0.96 2.739 0.009 0.73
SD: standard deviation, ES: effect size, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed
categories of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, ADHD: at-
tention deficit hyperactivity disorder, WCST: Wisconsin Card Sorting Test

www.psychiatryinvestigation.org 181
Impairment of Concept Formation Ability in Children with ADHD

Table 3. Comparisons of Korean personality inventory scores between children impaired and non-impaired by ADHD in each age group
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
LG (N=56) Test-retest 47.79±10.97 47.00±10.53 0.273 0.786 0.07
Lie 41.32±7.830 40.32±11.77 0.374 0.710 0.10
Frequency 49.36±14.90 57.36±18.41 -1.787 0.080 0.48
Ego resilience 41.25±11.70 39.75±13.93 0.436 0.664 0.12
Verbal development 59.79±13.47 53.68±13.71 1.682 0.098 0.45
Performance development 57.04±12.94 53.50±14.16 0.975 0.334 0.26
Anxiety 47.39±10.16 51.32±11.02 -1.387 0.171 0.37
Depression 53.39±11.61 55.11±10.96 -0.568 0.572 0.15
Somatization 45.68±9.470 47.93±8.810 -0.921 0.361 0.25
Delinquency 52.25±12.65 57.89±11.44 -1.750 0.086 0.47
Hyperactivity 60.11±11.79 60.36±12.72 -0.076 0.939 0.02
Familial disharmony 51.71±9.870 53.43±17.79 -0.446 0.658 0.12
Socialization 53.57±0.860 55.00±8.170 -0.708 0.482 0.25
Psychosis 49.43±10.20 50.86±13.74 -0.442 0.661 0.11
Autism 58.21±13.95 56.64±13.88 0.423 0.674 0.11
HG (N=56) Test-retest 49.07±11.56 47.96±11.42 0.361 0.720 0.10
Lie 41.96±10.27 39.86±9.620 0.793 0.431 0.21
Frequency 50.46±14.58 52.07±11.48 -0.458 0.649 0.12
Ego resilience 41.00±11.74 39.25±11.49 0.564 0.575 0.15
Verbal development 53.14±12.58 55.14±11.05 -0.632 0.530 0.17
Performance development 52.43±11.84 52.82±12.18 -0.122 0.903 0.03
Anxiety 52.75±12.10 56.82±13.37 -1.195 0.237 0.31
Depression 57.36±12.31 58.64±11.00 -0.412 0.682 0.11
Somatization 46.96±8.970 46.75±9.760 0.086 0.932 0.34
Delinquency 55.04±9.010 56.96±12.12 -0.676 0.502 0.18
Hyperactivity 58.46±11.57 62.11±14.34 -1.046 0.300 0.28
Familial disharmony 50.61±12.15 50.11±11.59 0.158 0.875 0.04
Socialization 56.67±8.490 57.79±7.830 -0.508 0.613 0.14
Psychosis 47.37±17.45 43.43±18.70 0.807 0.423 0.22
Autism 56.41±15.75 58.29±14.49 -0.461 0.647 0.12
SD: standard deviation, ES: effect size, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed
categories of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, ADHD: at-
tention deficit hyperactivity disorder, WCST: Wisconsin Card Sorting Test

0.62), Word reading of Color word (L-ADHD, 19.93±6.04; and EFs variables in the L-group showed that the number of
N-ADHD, 16.72±3.43; t=2.44; p<0.05; ES=0.65), and Color CC was positively related to response time on the DST back-
naming of Color word (L-ADHD, 25.79±9.83; N-ADHD, ward subtest (r=0.460, p<0.001) and on the VST (r=0.340, p<
21.38±5.97; t=2.03; p<0.05; ES=0.54). The HG group’s sec- 0.05). In addition, the number of CC was positively correlat-
tions, on the other hand, did not differ significantly in response ed with 5th trial (r=0.297, p<0.05), distracter recall (r=0.340,
times on TMT type A and B and the subtests of the Stroop test p<0.05), 6th trial (r=0.356, p<0.05), delayed recall (r=0.356,
(Table 6). p<0.01), delayed recognition (r=0.334, r<0.05), total recall
(r=0.349, p<0.01), which is an index of the VLT, and 5th trial
Relationships between CC and EFs scores in LG (r=0.286, p<0.05), which is an index of VRT. The number of
ADHD and HG ADHD CC was negatively correlated with errors on the TMT type A
The results of the correlation analysis of the number of CC (r=-0.322, p<0.05) and B (r=-0.273, p<0.05), color reading (r=

182 Psychiatry Investig 2010;7:177-188

 HJ Hong et al.

Table 4. Comparisons of CPT performance between L-ADHD and N-ADHD sections in each age group
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
LG (N=56) Omission errors
ACPT 045.07±21.85 41.00±18.49 0.753 0.455 0.20
VCPT 030.71±26.78 24.93±21.57 0.890 0.377 0.24
Commission errors
ACPT 041.29±16.65 48.50±48.63 -0.743 0.461 0.20
VCPT 033.07±27.85 29.36±25.87 0.517 0.607 0.14
Reaction times (msec)
ACPT 658.11±79.45 627.21±114.01 1.176 0.245 0.32
VCPT 607.32±81.26 603.29±78.170 0.189 0.851 0.05
HG (N=56) Omission errors
ACPT 025.18±15.41 19.75±13.23 1.414 0.163 0.38
VCPT 012.96±15.65 14.75±19.85 -0.374 0.710 0.10
Commission errors
ACPT 021.57±11.68 23.00±21.67 -0.307 0.760 0.08
VCPT 016.04±15.48 14.79±17.07 0.287 0.775 0.08
Reaction times (msec)
ACPT 629.61±66.43 620.07±78.610 0.490 0.626 0.13
VCPT 499.71±68.06 513.89±90.350 -0.663 0.510 0.18
SD: standard deviation, ES: effect size, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed
categories of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, CPT: con-
tinuous performance test, ACPT: auditory CPT, VCPT: visual CPT, msec: 1/100 second, ADHD: attention deficit hyperactivity disorder, WCST:
Wisconsin Card Sorting Test

-0.354, p<0.01), word reading of color word (r=-0.368, p<0.01), ning and organization, complex problem-solving, and response
and color naming of color word (r=-0.361, p<0.01) in the St- inhibition).13 If a cutoff of 1.5 standard deviations (based on
roop test. On the other hand, there was no correlation be- the control samples) on performance of some EF measures de-
tween the number of CC and EFs variables in the H-group fines EF impairment, then, on any individual measure, be-
(Table 7). tween 3.7% and 56.7% of Korean children with ADHD are EF
We conducted stepwise multiple regression analyses to com- impaired,32 between 16% and 51% of ADHD children are EF
pare the relative contribution of the EF variables to the num- impaired (on other measures). However, a comparison of mul-
ber of CC. The results showed that, in L-ADHD participants, tiple deficits revealed only 10% of ADHD children showed
the number of CC was significantly predicted by a combina- deficits across all five domains (stop signal reaction time, re-
tion of the backward test on the DST and VST (R2=0.273, p< action time variability, Stroop color-word interference effect,
0.001). Tests of the individual predictors revealed that the Continuous Performance Task commission errors, and TMT
backward tests of the DST (t=3.81, p<0.01) and VST (t=2.12, B time). By contrast, 21% of children with ADHD (and 53%
p<0.05) contributed significantly to the regression equation of controls) were unimpaired on all five measures.33
(Table 8). One of the challenges for understanding EF in children is
that EF skills develop rapidly through childhood, with the sug-
DISCUSSION gestion that progression is not necessarily linear, but may occur
in spurts. Further, it appears that components of EF might
Researchers have studied the neuropsychological func- demonstrate different developmental trajectories, adding to
tioning of elementary-school-age ADHD children since the the complexity of the domains.4 Within EF, EF skills such as
early 1970s, and, while not all studies showed positive results, information processing, cognitive flexibility, and goal-setting
in their entirety these studies indicated that, as a group, chil- ability develop rapidly when children are between 7 and 10
dren with ADHD exhibit sub-average or relatively weak per- years of age.34-36 Such EF can be evaluated by various neuro-
formance on various tasks of vigilance, verbal learning (par- psychological tests. In particular, the WCST is a commonly-
ticularly encoding), WM, and EF (such as set-shifting, plan- used measure with both clinical and research utility.37 The most

www.psychiatryinvestigation.org 183
Impairment of Concept Formation Ability in Children with ADHD

Table 5. Comparisons of digit span, visual span, verbal learning test, and visual recognition test performance between L-ADHD and N-
ADHD sections for each age group
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
LG (N=56) Digit span
Forward 5.11±1.310 5.21±0.920 -0.354 0.725 0.08
Backward 3.18±0.940 3.82±0.860 -2.658 0.010 0.71
Visual span
Forward 4.61±1.230 5.00±1.120 -1.250 0.217 0.33
Backward 3.50±1.450 3.93±1.210 -1.197 0.236 0.32
Verbal learning test
1st trial 3.96±2.200 4.61±1.810 -1.193 0.238 0.32
5th trial 9.43±2.690 10.43±2.890 -1.342 0.185 0.36
Distracter recall 3.79±2.170 5.00±2.110 -2.126 0.038 0.57
6th trial 7.43±3.490 9.07±3.240 -1.825 0.074 0.49
Delayed recall 7.18±2.310 8.82±2.870 -2.360 0.022 0.63
Delayed recognition 11.21±3.100 12.82±1.720 -2.401 0.020 0.64
Total recall 35.75±10.64 42.11±9.840 -2.321 0.024 0.62
Learning index 5.46±3.240 5.82±2.960 -0.431 0.668 0.12
Proactive interference -0.24±0.930 -0.31±0.880 0.312 0.756 0.08
Retroactive interference 0.13±0.670 0.12±0.210 0.042 0.967 0.02
Visual recognition test
1st trail 9.50±3.040 10.36±2.200 -1.210 0.232 0.32
5th trial 12.57±1.530 13.29±1.490 -1.774 0.082 0.48
Delayed recognition 12.54±1.570 12.46±3.110 0.108 0.914 0.03
Total recognition 58.21±7.120 60.64±9.700 -1.068 0.290 0.29
Learning index 3.07±2.370 2.93±1.720 0.258 0.797 0.07
HG (N=56) Digit span
Forward 6.07±0.980 5.96±1.570 0.306 0.761 0.08
Backward 4.11±1.200 4.39±1.170 -0.905 0.370 0.24
Visual span
Forward 5.57±1.400 5.43±1.290 0.397 0.693 0.10
Backward 4.64±1.520 5.00±1.440 -0.902 0.371 0.24
Verbal learning test
1st trial 4.61±2.100 5.75±2.780 -1.735 0.088 0.33
5th trial 11.29±2.680 12.36±2.180 -1.641 0.107 0.44
Distracter recall 5.36±2.380 6.18±1.870 -1.438 0.156 0.38
6th trial 9.82±2.420 10.54±3.170 -0.948 0.347 0.26
Delayed recall 8.89±2.390 9.29±3.840 -0.460 0.648 0.13
Delayed recognition 13.14±2.140 12.43±3.880 0.853 0.399 0.23
Total recall 43.32±0.310 48.21±10.70 -1.743 0.087 0.65
Learning index 6.68±2.330 06.61±2.780 0.104 0.917 0.03
Proactive interference -0.37±0.870 -0.31±0.920 -0.264 0.793 0.07
Retroactive interference 0.08±0.340 0.15±0.200 -1.001 0.321 0.25
Visual recognition test
1st trail 9.89±2.960 10.93±1.740 -1.596 0.116 0.43
5th trial 12.79±2.970 13.50±1.840 -1.082 0.284 0.29

184 Psychiatry Investig 2010;7:177-188

 HJ Hong et al.

Table 5. Continued
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
HG (N=56) Visual recognition test
Delayed recognition 12.22±3.730 12.46±3.890 -0.235 0.815 0.06
Total recognition 58.93±11.49 63.04±8.000 -1.553 0.126 0.42
Learning index 2.89±1.950 02.57±1.570 0.679 0.500 0.18
SD: standard deviation, ES: effect size, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed
categories of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, ADHD: at-
tention deficit hyperactivity disorder, WCST: Wisconsin Card Sorting Test

Table 6. Comparisons of TMT type A & B and Stroop test performances between L-ADHD and N-ADHD sections in each age group
L-ADHD (N=28) N-ADHD (N=28)
Variables t p ES
(mean±SD) (mean±SD)
LG (N=56) TMT type A
Times (sec) 47.39±14.88 58.93±58.84 -1.006 0.322 0.27
Errors 03.00±2.93 02.14±2.21 1.236 0.222 0.33
TMT type B
Times (sec) 89.54±26.21 91.21±33.44 -0.209 0.835 0.05
Errors 08.11±6.70 05.54±5.20 1.605 0.114 0.43
Stroop test
Word reading (sec) 18.48±4.99 18.04±7.17 0.267 0.790 0.07
Color reading (sec) 32.55±11.55 26.49±7.53 2.324 0.025 0.62
Word reading of Color word (sec) 19.93±6.04 16.72±3.43 2.444 0.019 0.65
Color naming of Color Word (sec) 25.79±9.83 21.38±5.97 2.027 0.049 0.54
HG (N=56) TMT type A
Times (sec) 34.09±13.48 31.97±12.99 0.600 0.551 0.16
Errors 01.93±4.14 02.29±5.02 -0.290 0.773 0.08
TMT type B
Times (sec) 61.48±17.05 60.91±25.45 0.098 0.922 0.03
Errors 03.46±3.58 03.29±6.50 0.127 0.899 0.03
Stroop test
Word reading (sec) 15.23±4.31 13.80±2.87 1.460 0.150 0.39
Color reading (sec) 22.82±5.19 20.35±6.26 1.610 0.113 0.43
Word reading of Color word (sec) 15.09±3.27 14.57±4.09 0.518 0.606 0.14
Color naming of Color Word (sec) 16.92± 4.28 15.48±6.26 1.009 0.318 0.27
SD: standard deviation, ES: effect size, LG: lower grades, HG: higher grades, L-ADHD: ADHD participants with lower performance in completed
categories of WCST than the age norm reference, N-ADHD: ADHD participants performing the same as the age-referenced norm, TMT: trail
making test, sec: second, ADHD: attention deficit hyperactivity disorder, WCST: Wisconsin Card Sorting Test

commonly-reported WCST performance indices are the num- cent correct, number of CC, total errors, and perseverative er-
ber of CC and the number of perseverative errors. In a meta- rors, but there were variations in ESs both across studies as
analytic review of age-related differences in old age,38 the per- well as within WCST variables, and poor performance is not
severative error measure was marginally more sensitive to age sufficient for a diagnosis of ADHD. In a study of factor anal-
differences, compared with categories achieved. Thus, it may ysis and developmental WCST performance trends in Korean
be a better metric of EF, if a single WCST score is to be used. elementary school children, Ko et al.17 and Lee et al.18 showed
Meta-analytic study results of WCST with children37 suggest WCST scoring variables had a important difference in devel-
that, across all of the studies, children with ADHD fairly con- opmental trends, and three factors (conceptual formation, per-
sistently exhibited poorer performance as compared to chil- severation, and ability to sustain attention) effectively explained
dren without clinical diagnoses of ADHD, as measured by per- the nature of these differences. However, the studies showed

www.psychiatryinvestigation.org 185
Impairment of Concept Formation Ability in Children with ADHD

Table 7. Correlation matrix between numbers of completed catego- Table 7. Continued

ries and EF variables EFs variables
Lower grades Higher grades
Lower grades Higher grades EFs variables children children
EFs variables children children r p r p
r p r p
Stroop test
Continues Performance Test
Color reading (sec) -0.354 0.007 -0.202 0.135
Omission errors Word reading of Color -0.368 0.005 -0.071 0.605
ACPT -0.149 0.274 -0.213 0.115 word (sec)
VCPT -0.159 0.242 0.017 0.903 Color naming of Color -0.361 0.006 -0.107 0.431
Commission errors word (sec)
ACPT 0.027 0.845 0.040 0.767 ACPT: auditory continues performance test, VCPT: Visual CPT,
TMT: trail making test, msec: 1/100 second, sec: second
VCPT -0.199 0.141 -0.131 0.335
Reaction times (msec) some differences. Ko et al.17 and Lee et al.18 used the comput-
erized version of the WCST, but the WCST performance in-
ACPT -0.113 0.408 -0.097 0.476
structions and feedback differed in language and modalities.
VCPT -0.068 0.617 0.070 0.608
There were no age effects in failure to maintain set17 or the
Digit span
number of perseverative errors.18 However, one factor (concept
Forward 0.117 0.392 -0.079 0.561 formation), including the number of CC, accounted for more
Backward 0.460 0.000 0.191 0.159 of the common variance than other factors did.17,24 Therefore,
Visual span we adopted number of CC for the WCST performance index.
Forward 0.263 0.051 0.075 0.584 Meta-analysis for demonstrating sensitivity and specificity
Backward 0.340 0.010 0.218 0.107 for identifying those EF deficits associated with ADHD sug-
Verbal learning test gested that across all of the studies, individuals with ADHD
1st trial 0.244 0.070 0.230 0.088 fairly consistently exhibit poorer performance on the WCST,
5th trial 0.297 0.026 0.186 0.170
as compared to individuals without clinical diagnosis, as mea-
sured by percent correct, number of CC, total errors, and per-
Distracter recall 0.340 0.010 0.237 0.079
severative errors. Notably, various clinical groups have per-
6th trial 0.340 0.010 0.099 0.470
formed more poorly than ADHD groups in a number of stu-
Delayed recall 0.356 0.007 0.023 0.865 dies. Thus, while impaired performance on the WCST may
Delayed recognition 0.334 0.012 -0.068 0.621 be indicative of an underlying neurological disorder, most
Total recall 0.349 0.008 0.205 0.130 likely related to frontal lobe function, poor performance is not
Learning index 0.110 0.418 -0.046 0.739 sufficient for a diagnosis of ADHD.37 Developmental norms
Proactive interference -0.007 0.961 0.071 0.605 by age for an unselected sample of normal children in grades 1
Retroactive interference -0.005 0.972 0.115 0.397 through 6 indicated that children make rapid gains in the num-
Visual recognition test
ber of CC and significantly reduce the number of persevera-
tive errors with advancing age, such that, by 10 years of age,
1st trail 0.151 0.267 0.222 0.101
their performance on the WCST is indistinguishable from
5th trial 0.286 0.033 0.151 0.266
that of adults.39 Therefore, tests such as the WCST may be in-
Delayed recognition 0.119 0.383 0.024 0.862 appropriate for young children since the prefrontal regions,
Total recognition 0.199 0.142 0.231 0.087 and especially the dorsolateral areas, do not become fully
Learning index 0.018 0.895 -0.097 0.475 functional until later in development.40 Furthermore, the cog-
TMT type A nitive activities mediated by the prefrontal regions do not be-
Times (msec) 0.068 0.616 -0.076 0.578 gin to develop until adolescence, and a disorder of the frontal
Errors -0.322 0.015 0.055 0.688 area can be symptomless until age 12-15 or even older.41 Le-
TMT type B
sions in the prefrontal areas, therefore, may appear “silent’ be-
fore the age of 7, since perseveration seems to be the norm.
Times (msec) -0.097 0.475 -0.021 0.880
However, by age 7, children surpass the performance of adults
Errors -0.273 0.042 -0.023 0.869
with focal frontal lesions but not those with focal nonfrontal
Stroop test lesions, suggesting that at this age the frontal regions are be-
Word reading (sec) -0.056 0.684 -0.207 0.126 ginning to become operational, although they are not yet func-

186 Psychiatry Investig 2010;7:177-188

 HJ Hong et al.

Table 8. The stepwise multiple regression analysis results for the number of completed categories by EF variables
Unstandardized coefficient Standard coefficients
Index/Groups Model in difference t p
B Standard error Beta
Lower grades children (constant) 1.150 0.743 1.549 0.127
(R2=0.273, p<0.001) Digit span test; backward 0.627 0.185 1.549 3.807 0.001
Visual span test; backward 0.278 0.131 0.254 2.116 0.039
EF: executive functions

tionally mature.39 nor sufficient to draw such a distinction because of other con-
WM is critical to conscious thought, because it permits in- founding or unknown contaminating factors. The adminis-
ternal representation of information, to guide decision-making tration and interpretation of age-inappropriate neuropsycho-
and overt behavior during an activity so that behavior is not do- logical tests have caused controversy in research and clinical
minated by the immediate sensory cues in the environment.42 practice. The WSCT is a commonly-used measure in both re-
WM is likely crucial for performance on the WCST, given that search and clinical practice, 37 and 75.5% of neuropsycholo-
the participant must keep in mind information about previ- gists have reported using the WCST as part of their battery.51
ous sorts while simultaneously processing information to de- However, ADHD individuals fairly consistently exhibit poor-
termine the next sort.43,44 Therefore, age differences on the er performance on the WCST, compared to individuals with-
WCST may be a function of reduced WM, a capacity with out any such clinical diagnosis as measured by some indi-
well-established age-related deficits.45 The fact that LG ADHD ces.37 Moreover, the causes of poor WCST performance are
perform more poorly on the WCST than LG normal children different, especially in elementary-school-age children. The
might be due or related to WM or verbal memory ability, but cause of poorer performance than normal in lower-grades ch-
the lower WCST performance of HG ADHD children compar- ildren with ADHD was not shown by WCST alone; other EF,
ed to HG normal children was not explicable by any other EF. such as verbal and visual WM ability, explained 27.3% of the
This means WCST performance in LG ADHD was related to variance. This suggests that a neuropsychological test battery
developmental variations of EF and did not show an impaired design for measuring EF in ADHD children should consider
EF only. However, HG ADHD children showed impaired EF age first. For developmental variations, and in interpretation
of requiring from WCST performance discrete to other EFs. of neuropsychological tests results, researchers should consid-
In line with the views of Alexander and Stuss,46 EF matures at er the dynamic relationships of EF measured by neuropsycho-
different rates, especially cognitive flexibility (the ability to cope logical tests, but a separate domain of EF-oriented interpreta-
with multi-dimensional switching tasks), and planning and tion would not be preferable.
organizing skills develop rapidly between 7 and 9 or 10 years of
age 34-36 compared to other EFs.4 These means that we did REFERENCES
not assess fully established EF in LG ADHD and could not dif- 1. Gordon M, Barkley RA, Lovett BJ. Tests and observational measures
ferentiate these EF impairments from developmental cogni- (3rd ed). In: Bakrley RA, editor. Attention deficit hyperactivity disor-
der: a handbook for diagnosis and treatment. New York: The Guilford
tive delay or other variations. Press, 2006, p.369-388.
The Stroop test measure of cognitive control assesses the ease 2. Willcutt EG, Doyle AE, Nigg JT, Faraone SV, Pennigton BF. Validity of
with which a person can maintain a goal in mind and suppress the executive theory of attention deficit/hyperactivity disorder: a meta-
analytic review. Biol Psychiatry 2005;57:1336-1346.
a habitual response in favor of a less familiar one. In children,
3. Pennington BF, Ozonoff S. Executive functions and developmental
the Stroop interference effect is minimal when they are just be- psychopathology. J Child Psychol Psychiatry 1996;37:51-87.
ginning to learn to read and increases (in the first few grades) 4. Anderson P. Assessment and development of executive function (EF)
as children gain in reading fluency. Thereafter, it declines grad- during childhood. Child Neuropsychol 2002;8:71-82.
5. Morgan AB, Lilenfeld SO. A meta-analytic review of the relation anti-
ually as children gain control over the automatic response of social behavior and neuropsychological measures of executive func-
reading.47 As with WCST performance, WM contributes to tion. Clin Psychol Rev 2000;20:113-136.
Stroop interference effects, and individual differences in WM 6. Collette F, Vad der Linden M. Brain imaging of the central executive
component of working memory. Neurosci Biobehav Rev 2002;26:105-
capacity predict performance on the Stroop task, indicating 125.
the importance of goal maintenance in the face of competi- 7. Romine CB, Reynolds CR. A model of the development of frontal lobe
tion from habit.48 After all, the Stroop interference effect is functioning: findings from a meta-analysis. Appl Neuropsychol 2005;
12:190-201.
affected by WM ability, but the Stroop test evaluates EF dif-
8. Doyle AE. Executive functions in attention-deficit/hyperactivity disor-
ferently than the WCST or TMT B.49,50 EF weakness in ADHD der. J Clin Psychiatry 2006;67 Suppl 8:21-26.
children compared to normal children is neither necessary 9. Castellanos FX, Sounga-Barke EJ, Milham MP, Tannock R. Character-

www.psychiatryinvestigation.org 187
Impairment of Concept Formation Ability in Children with ADHD

izing cognition in ADHD: beyond executive dysfunction. Trends Cogn symptoms in children with ADHD. Unpublished doctorial thesis, Dae-
Sci 2006;10:117-123. gu: Yungnam University Undergraduate School; 2009.
10. Nigg JT. What causes ADHD? Understanding what goes wrong and why. 33. Nigg JT, Wilcutt EG, Doyle AE, Sonuga-Barke EJ. Causal heterogeneity
New York: Guildford Press; 2006. in attention-difecit/hyperactivity disorder: do we need neuropsycho-
11. Willcutt EG, Brodsky K, Chhabildas N, Shanahan M, Yerys B, Scott A, logically impaired subtypes? Biol Psychiatry 2005;57:1224-1230.
et al. The neuropsychology of ADHD: validity of the executive func- 34. Anderson P, Anderson V, Lajoie G. The Tower of London Test: validation
tion hypothesis. In attention deficit hyperactivity disorder: from genes and standardization for pediatric population. The Clin Psychologist 1996;
to patents. Totowa, NJ: Human, 2005b, p.185-213. 10:54-65.
12. Wåhlstedt C, Thorell LB, Bohlin G. Heterogeneity in ADHD: neuro- 35. Anderson P, Anderson V, Northam E, Taylor H. Standardization of the
psychological pathway, comorbidity and symptom domains. J Abnorm Contingency Naming Test for school-aged children: a new measure of
Child Psychol 2009;37:551-564. reactive flexibility. Clin Neuropsychol Assess 2000;1:247-273.
13. Seidman LJ. Neuropsychological functioning in people with ADHD 36. Krikorian R, Bartok J. Developmental data for the Porteus Maze Test.
across the lifespan. Clin Psychol Rev 2006;26:466-485. The Clin Neuropsychologist 1998;12:305-310.
14. American Psychiatric Association. Diagnostic and statistical manual 37. Romine CB, Lee D, Wolfe ME, Homack S, George C, Riccio CA. Wis-
of mental disorder, 4th ed-revison. Washington, DC: American Psychi- consin Card Sorting Test with children: a meta-analytic study of sensi-
atric Press; 2000. tivity and specificity. Arch Clin Neuropsychol 2004;19:1027-1041.
15. Wechsler D. Wechsler intelligence scale for children (3rd ed). San An- 38. Rhodes MG. Age-realted differences in performance on the Wisconsin
tonio, TC: Psychological Corporation; 1991. card sorting test: a meta-analytic review. Psychol Aging 2004;19:482-494.
16. Kwak GJ, Park HY, Kim CT. Korean-Wechsler intelligence scale for 39. Chelune GJ, Baer RA. Developmental norms for the Wisconsin card
children (3rd ed). Seoul: Special Education Publishing Co.; 2001. Sorting test. J Clin Exp Neuropsychol 1986;8:219-228.
17. Ko RW, Kang YW, Park MC, Oh KJ. Developmental trends of perfor- 40. Goldman PS. An alternative to developmental plasticity: Heterology of
mance on the Wisconsin Card Sorting Test in Korean children. Korean CNS structures in infants and adults. In: Stein D, Rosen J, Butters N,
J Clin Psychol 2002;21:461-478. editors. Plasticity and recovery of function in the central nervous system.
18. Lee JB, Kim JS, Seo WS, Shin HJ, Bai DS, Lee JH. The neuropsycho- New York: Academic Press, 1974, p.149-174.
logical characteristics of the elementary school aged child by comput- 41. Golden CJ. The Lurai-Nebraska children’s battery: theory and formula-
erized neurocognitive function test. Korean J Psychosomatic Medicine tion. In: Hynd GW, Obrzut JE, editors. Neurospsychologcial assessment
2003b;11:118-136. and the school-age child: issues and procedures. New York: Grune & St-
19. Wechsler D. Wechsler intelligence scale for children. New York: Psy- ratton, 1981, p.277-302.
chological Corporation; 1949. 42. Martinussen R, Hayden J, Hogg-Johnson S, Tannock R. A meta-analysis
20. Wirt RD, Lachar D, Klinedisnt JK, Seat PD. Multidimensional descrip- of working memory impairments in children with attention-deficit/hy-
tion of child personality: a manual for the personality inventory for peractivity disorder. J Am Acad Child Adolesc Psychiatry 2005;44:377-
children. Los Angeles: Western Psychological Services; 1984. 384.
21. Kim ST, Kim JH, Song SH, Lee HK, Ju YH, Hong ChH, Hwang ST. 43. Dehane S, Changeux JP. The Wisconsin Card Sorting Test: theoretical
Korean personality inventory for children. Seoul: Korea Guidance. analysis and modeling in a neuronal network. Cereb Cotex 1991;1:62-79.
22. Maxmedica. Computerized Neurocognitive Function Test (4th ed). 44. Hartman M, Bolton E, Fehnei SE. Accounting for age differences on the
Seoul: Madmedical Co.; 2004. Wisconsin Card Sorting Test: decreased working memory, not inflexi-
23. Ha KS, Kwon JS, Lyoo IK, Kong SW, Lee DW, Youn T. Development bility. Psychol Aging 2001;16:385-399.
and standardization process, and factor analysis of the Computerized 45. Verhaeghen P, Marcoen A, Gossens L. Facts and fiction about memory ag-
Cognitive Function Test system for Korea adults. J Korean Neuropsy- ing: a qualitative integration of research findings. J Gerontol 1993;48:
chiatric Assoc 2002;41:551-562. P157-P171.
24. Lee JB, Kim JS, Seo WS, Shin HJ, Bai DS, Lee HL. The validity and reli- 46. Alexander MP, Stuss DT. Disorders of frontal lobe functioning. Semin
ability of computerized neurocognitive function test in the elementary Neurol 2000;20:427-437.
school child. Korean J Psychosomatic Medicine 2003a;11:97-117. 47. Strauss E, Sherman EMS, Spreen O. A compendium of neuropsycho-
25. Koo BH, Bai DS. Diagnostic availability of a Computerized Neurocog- logical tests: administration, norms, and commentary (3rd edition). New
nitive Function Test for Children with ADHD. J Korean Soc Biol Psy- York: Oxford University Press, 2006, p.477-499.
chiatry 2006;12:39-49. 48. Kane MJ, Engle RW. Working-memory capacity and the control of at-
26. Berg EA. A simple objective treatment for measuring flexibility in think- tention: the contributions of goal neglect, response competition, and
ing. J Gen Psychol 1948;39:5-22. task set to Stroop interference. J Exp Psychol Gen 2003;132:47-70.
27. Milner B. Effects of different brain lesions on card sorting. Arc Neurol 49. Bondi MW, Serody AB, Chan AS, Eberson-Shumate SC, Delis DC,
1963;9:90-100. Hansen LA, et al. Cognitive and neuropathologic correlates of Stroop
28. Heaton RK. Wisconsin Card Sorting Test (WCST). Odessa, FL: Psy- Color-Word Test performance in Alzheimer’s disease. Neuropsycholo-
chological Assessment Resource; 1981. gy 2002;16:335-343.
29. Heaton RK, Chelune GJ, Talley JL, Kay GG, Curtiss G. Wisconsin 50. Boone KB, Pontón MO, Gorsuch RL, González JJ, Miller BL. Factor
Card Sorting Test Manual: revised and expanded. Odessa, FL: Psycho- analysis of four measures of prefrontal lobe functioning. Arch Clin Neu-
logical Assessment Resource; 1993. ropsychol 1998;13:585-595.
30. Milner B. Interhemispheric differences in the localization of psycho- 51. Rabin LA, Burton LA, Barr WB. Utilization rates of ecologically ori-
logical processes in man. Br Med Bull 1971;27:272-277. ented instruments among clinical neuropsychologists. Clin Neuropsy-
31. Cohen J. A power primer. Psychol Bull 1992;112:155-159. chol 2007;21:727-743.
32. Lee HL. Psychopathology and neurocognitive function on behavioral

188 Psychiatry Investig 2010;7:177-188