Clinical Utility of Cancellation on the WISC-IV

Abstract

This study examined empirical evidence for clinical utility of the Wechsler Intelligence Scale for Children, fourth edition (WISC-IV) cancellation subtest by comparing data from 597 clinical and 597 matched control children. The results of dependent t and sequential logistic regression analyses demonstrated that (a) children with intellectual disabilities, motor impairments, head injuries, Autistic/Aszperger’s disorder, ADHD and learning disabilities, and mathematics disorder showed significant deficits on the cancellation subtest; (b) children with intellectual disabilities and Asperger’s disorder benefited when stimuli were randomly aligned, but children with ADHD benefited from structured conditions; (c) beyond the full-scaled IQ (FSIQ) and General Ability Index (GAI)–Cognitive Proficiency Index (CPI) discrepancy scores, the cancellation subtest added unique diagnostic power to identify children with reading disorders, mild intellectual disabilities, closed head injuries, and motor impairments. These results suggest the utility of the cancellation subtest in clinical assessment.

Keywords

cancellation clinical diagnosis WISC-IV

Cancellation is a new supplemental processing speed subtest in the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV). The test asks the child to scan both a random and a structured arrangement of pictures, and marks target pictures within a specified time limit (Wechsler, 2003a). In neuropsychological settings, cancellation tasks are traditionally used as measures of visual inattention. These tasks primarily assess the capacity for sustained attention, accuracy and speed of visual scanning and discrimination, and activation and inhibition of responses (Lezak, Howieson, Bigler, & Tranel, 2012; Miller & Hale, 2008). Performance deficits in these tasks were reported to be associated with spatial inattention problems (Halligan, Marshall, & Wade, 1989), particularly for stroke patients with unilateral spatial inattention (Vanier et al., 1990; Ferber & Karnath, 2001). No other study has examined the incremental utility of this test when it was considered for an intellectual assessment battery, such as WISC-IV.

In WISC-IV hierarchical confirmatory factor analyses, cancellation is the subtest with the lowest g loading and the highest uniqueness (Chen, Keith, Weiss, Zhu, & Li, 2010; Keith, Fine, Taub, Reynolds, & Kranzler, 2006; Weiss, Keith, Zhu, & Chen, in press). Thus, larger percentages of subtest variance could be explained by factors other than those specified in the WISC-IV structure: full-scaled IQ (FSIQ), verbal comprehension (VCI), perceptual reasoning (PRI), working memory (WMI), and processing speed (PSI). This subtest was also evaluated by Ortiz who identified it as one of the few WISC-IV subtests with low cultural loading and linguistic demands. Therefore, this test may yield results more equitable for diverse individuals (Flanagan & Kaufman, 2004).

Wechsler tests are frequently used as part of diagnostic assessments (Prifitera, Saklofske, & Weiss, 2008). The consensus on the primary utilities of these batteries is that (a) they offer a reliable summary of general cognitive ability, and (b) they provide methods for evaluating cognitive strengths and weaknesses (Flanagan & Kaufman, 2004; Kaufman & Linchtenberger, 2002; Matarazzo & Herman, 1985; Sattler & Dumont, 2004), also known as profiles/discrepancies. Of these tests, the FSIQ has the highest reliability and achievement-related validity (Wechsler, 2003b). Thus, FSIQ is considered one of the best summaries for overall cognitive ability. Furthermore, Weiss et al. (2006) suggested that index scores should be considered the primary level of clinical interpretation of the WISC-IV. Wechsler index score profiles have tremendous clinical utility (Allen, Thaler, Donohue, & Mayfield, 2010; Calhoun & Mayes, 2005; Weiss, Prifitera, & Saklofske, 2005). In particular, the discrepancy between two WISC-IV indices, general ability (GAI) and cognitive proficiency (CPI), shows meaningful intrapersonal cognitive characteristics and predictive power (Raiford, Weiss, Rolfhus, & Coalson, 2005; Weiss, Beal, Saklofske, Alloway, & Prifitera, 2008; Weiss & Gabel, 2008). The GAI is a composite of the more highly g loaded subtests within WISC-IV. It provides an estimate of general intellectual ability, with reduced emphasis on working memory and processing speed relative to the FSIQ. The CPI is the counterpart to the GAI. It summarizes the performance of working memory and processing speed; thus, providing a more comprehensive summary of cognitive processing proficiency (Saklofske, Zhu, Coalson, Railford, & Weiss, 2010; Saklofske et al., 2012). Efficient cognitive processing frees cognitive resources for more complex or higher-level tasks. The GAI-CPI discrepancy provides a comparison of a measure of current knowledge and reasoning ability to cognitive proficiency, which could be considered as the one Wechsler cognitive profile/discrepancy with the broadest construct perspective and the smallest measurement error (Zhu, Railford, & Coalson, 2009).

To examine the incremental utility of the cancellation subtest, both FSIQ and GAI-CPI discrepancy should be considered. FSIQ is the most reliable summary for overall cognitive ability, and GAI-CPI discrepancy is the most reliable measure for cognitive variability. These two measures cover the core information the Wechsler battery could provide. If the cancellation subtest provides diagnostic power in addition to that provided by FSIQ and GAI-CPI, its practical utility could be secured.

Utility of the cancellation subtest for types of disabilities has never been examined or reported. Cancellation is a measure of processing speed. The importance of considering the processing speed in a complete intelligence structure is well recognized (Carroll, 1993, 2005; Woodcock, McGrew, & Mather, 2001). Processing speed is found to have a significant effect on higher-level cognitive abilities, such as working memory and fluid reasoning (Fry & Hale, 1996, 2000; Kail, 2007; Nettelbeck & Burns, 2010); thus, providing greater weight to processing speed in the intelligence test battery is justified by neuropsychological findings. Sattler and Dumont (2004) suggested that more research is required to determine whether this new subtest (with random vs. structured tasks) is clinically useful. The purpose of this study is to compare and investigate the cancellation scaled scores of clinical and matched control children. The incremental contribution of the cancellation subtest for diagnosis at control versus clinical groups was also tested. More specifically, the research question was whether inclusion of the cancellation subtest improved the ability of FSIQ and GAI-CPI discrepancy to predict clinical status.

Method

Participants

The WISC-IV scores of 597 clinical and 597 matched control children were analyzed. The heterogeneous clinical sample included 15 groups: children with ADHD (n = 89); children with ADHD and learning disorders (n = 45); children with Autistic disorder (n = 19); children with Asperger’s disorder (n = 27); children with expressive language disorder (n = 27); children with mixed receptive-expressive language disorder (n = 40); children with reading disorder (n = 56); children with reading and written expression disorder (n = 35); children with mathematics disorder (n = 33); children with reading, written expression, and mathematics disorder (n = 42); children with intellectual disabilities of mild severity (n = 63); children with intellectual disabilities of moderate severity (n = 57); children with closed head injuries (n = 27); children with open head injuries (n = 16); and children with motor impairments (n = 21). These children were drawn from a variety of educational and clinical settings, and were accepted for participation in clinical samples based on specified inclusion criteria (see Wechsler, 2003b, pp. 145-149, Appendix D). For meaningful comparisons, 597 matched control children were selected from the WISC-IV standardization sample. These children were matched with the clinical children by demographic variables of age, gender, ethnicity, parental education, and geographic region. Demographics of all samples are presented in Table 1. The mean age and FSIQ of each sample is presented, and followed by percentages of sample representations by gender, race/ethnicity, parental education, and geographic region.

Table 1.

Demographic Data for the Studied Samples.

		Clinical Sample
	Matched Control Sample	ID mild	ID moderate	RD	RWD	MD	RWMD	ADHD/LD	ADHD	ELD	RELD	OHI	CHI	AUT	ASP	MI
N	597	63	57	56	35	33	42	45	89	27	40	16	27	19	27	21
Age
M	11.2	11.8	12.1	10.6	10.9	10.8	10.5	11.1	10.7	9.9	10.5	12.5	13.4	11.7	12.4	10.2
SD	2.4	3.1	2.9	1.4	1.5	1.5	1.4	1.4	1.4	2.9	3.0	2.8	3.1	2.8	1.9	2.3
FSIQ
M	100.6	60.5	46.3	89.1	92.5	88.2	87.5	88.0	97.6	83.0	77.3	92.4	90.0	75.3	99.2	85.7
SD	13.4	9.2	8.4	10.3	11.1	8.6	10.5	12.9	14.0	11.1	12.6	17.8	11.9	19.5	17.7	14.9
Sex
Female	35.2	52.4	50.9	26.8	37.1	48.5	42.9	11.1	32.6	48.2	30.0	31.3	40.7	10.5	7.4	33.3
Male	64.8	47.6	49.1	73.2	62.9	51.5	57.1	88.9	67.4	51.8	70.0	68.7	59.3	89.5	92.6	66.7
Race/ethnicity
White	66.8	61.9	42.1	73.2	80.0	60.6	61.9	71.1	78.7	37.0	55.0	75.0	74.1	84.2	92.6	66.7
African American	18.9	23.8	45.6	10.7	8.6	33.3	31.0	15.6	7.9	25.9	17.5	18.8	7.4	10.5	3.7	14.3
Hispanic	12.9	12.7	10.5	14.3	11.4	6.1	7.1	13.3	11.2	37.0	25.0	6.3	11.1	5.3	3.7	19.1
Asian	0.2	—	—	—	—	—	—	—	—	—	—	—	3.7	—	—	—
Other	1.2	1.6	1.8	1.8	—	—	—	—	2.3	—	2.5	—	3.7	—	—	—
Parental education
≤ 8 years	1.5	3.2	3.5	3.6	—	—	2.4	2.2	1.1	—	—	—	—	—	—	—
9-11 years	12.7	25.4	26.3	7.1	11.4	12.1	14.3	8.9	1.1	37.0	25.0	—	—	—	—	9.5
12 years	36.0	39.7	35.1	51.8	31.4	51.5	28.6	28.9	30.3	40.7	42.5	25.0	22.2	31.6	14.8	61.9
13-15 years	31.2	23.8	21.1	30.4	31.4	30.3	40.5	37.8	39.3	22.2	22.5	50.0	40.7	26.3	33.3	19.1
≥ 16 years	18.6	7.9	14.0	7.1	25.7	6.1	14.3	22.2	28.1	—	10.0	25.0	37.0	42.1	51.9	9.5
Region
Northeast	21.6	14.3	15.8	48.2	14.3	39.4	23.8	22.2	22.5	11.1	15.0	—	—	15.8	18.5	42.9
South	33.2	38.1	47.4	14.3	48.6	27.3	28.6	33.3	36.0	44.4	35.0	50.0	22.2	26.3	22.2	14.3
Midwest	25.1	17.46	26.3	19.6	14.3	30.3	31.0	28.9	18.0	25.9	10.0	43.8	59.3	21.1	37.0	38.1
West	20.1	30.2	10.5	17.9	22.9	3.0	16.7	15.6	23.6	18.5	40.0	6.3	18.5	36.8	22.2	4.8

Note. Except for sample size (N), FSIQ, and age, data are reported as percentages.

FSIQ = Full-scaled IQ; ID Mild = Intellectual disabilities–mild severity; ID Moderate = Intellectual disabilities–moderate severity; RD = Reading disorder; RWD = Reading and written expression disorders; MD = Mathematics disorder; RWMD = Reading, written expression, and mathematics disorders; ADHD/LD = Attention-deficit/hyperactivity disorder and learning disorder; ADHD = Attention-deficit/hyperactivity disorder; ELD = Expressive language disorder; RELD = Mixed receptive-expressive language disorder; OHI = Open head injury; CHI = Closed head injury; AUT = Autistic disorder; ASP = Asperger’s disorder; MI = Motor impairment.

Instrumentation

The WISC-IV offers practitioners several cognitive measures, including an FSIQ, six composite scores (GAI, CPI, VCI, PRI, WMI, PSI), and scores from 15 subtests. This study analyzed FSIQ, GAI-CPI discrepancy, and cancellation scaled scores. Cancellation contains two items, the pictures of objects in Item 1 are arranged in a random pattern, and the objects in Item 2 are arranged in a structured pattern. For each item, the child is asked to scan the arrangements of pictures, and mark the target pictures within a 45-s limit. The reliability for FSIQ, GAI, CPI, GAI-CPI discrepancy, cancellation total (CA), cancellation random (CAR), and cancellation structured (CAS) is 0.97, 0.96, 0.93, 0.84, 0.79, 0.70, and 0.75, respectively.

Statistical Analyses

The statistical analyses examining the clinical utility of the WISC-IV cancellation subtest were conducted using SAS. Repeated measure t-tests determined the differences between each clinical group and its matched control counterpart. Sequential logistic regression (Menard, 2000; Spicer, 2005) determined the incremental diagnostic utility of the cancellation subtest after adjusting for the FSIQ and GAI-CPI discrepancy. For each of the 15 clinical groups, fit index, and the parameters of the two logistic regression models were compared with FSIQ and GAI-CPI discrepancy as independent variables, and the cancellation was used as the third predictor. The logistic regression coefficient and odds ratio for the cancellation subtest, classification hit rate, and model R² were reported.

Results

Comparison of Cancellation Scores

Table 2 presents score differences between clinical and matched control groups. The groups with the most salient deficits were intellectual disabilities, motor impairment, head injuries, Autistic/Asperger’s disorders, ADHD with LD cormorbidity, and mathematics disorder. Table 2 also presents the percentages for each clinical sample, as well as for the corresponding control group, that show impaired performance (lower than two standard deviations below the mean) on the cancellation subtest. Numerous clinical groups showed higher impairment rates. Particularly, the children diagnosed with head injuries, motor impairments, reading disabilities, Autistic spectrum disorder, and intellectual disabilities showed higher impairment rates, which were at least twice more than the rates shown in the corresponding control groups.

Table 2.

Group Differences on Cancellation Subtest Performance.

		Clinical Group			Matched Control Group			Group Mean Comparison
Group Type		M	SD	Impairment Rate (%)	M	SD	Impairment Rate (%)	N	Diff	t	d
ADHD	CA	9.1	2.9	4.49	9.6	3.0	4.49	89	0.49	1.13	0.17
	CAR	8.7	2.9	—	9.6	3.0	—	89	0.90	2.15*	0.30
	CAS	9.5	2.9	—	9.4	2.8	—	89	−0.13	−0.32	−0.05
ADHD/LD	CA	8.5	2.5	6.67	9.6	3.4	4.44	45	1.04	1.77	0.35
	CAR	8.4	2.6	—	9.6	3.0	—	45	1.29	2.21*	0.46
	CAS	8.8	2.9	—	9.3	3.1	—	45	0.49	0.89	0.16
Autistic disorder	CA	6.3	3.6	31.58	8.7	4.1	15.79	19	2.47	2.10	0.64
Autistic disorder	CAR	6.5	3.5	—	8.6	3.7	—	19	2.16	1.89	0.60
	CAS	6.6	3.5	—	9.2	4.1	—	19	2.58	2.25*	0.68
Asperger’s disorder	CA	8.0	2.7	14.81	9.9	3.2	3.70	27	1.93	2.27*	0.65
Asperger’s disorder	CAR	8.4	2.7	—	9.5	3.3	—	27	1.07	1.28	0.35
	CAS	7.5	2.4	—	9.9	2.7	—	27	2.41	3.35**	0.96
Expressive language disorder	CA	9.4	2.2	0.00	9.4	3.5	0.00	27	0.00	0.00	0.00
	CAR	9.3	1.9	—	9.2	3.2	—	27	−0.11	−0.14	−0.04
	CAS	9.5	2.2	—	9.6	3.3	—	27	0.07	0.08	0.02
Mixed language disorder	CA	8.4	3.0	5.00	10.4	3.4	5.00	40	1.98	2.52*	0.62
Mixed language disorder	CAR	9.0	2.7	—	10.3	3.4	—	40	1.30	1.69	0.42
	CAS	8.4	2.8	—	10.2	3.2	—	40	1.85	2.56*	0.61
LD-reading	CA	10.1	3.3	3.57	9.7	2.8	1.79	56	−0.43	−0.78	−0.14
	CAR	10.3	3.0	—	9.7	2.7	—	56	−0.68	−1.26	−0.24
	CAS	9.6	3.3	—	9.5	2.8	—	56	−0.14	−0.29	−0.05
LD-reading/written	CA	9.5	3.0	5.71	9.0	2.9	2.8	35	−0.54	−0.70	−0.18
LD-reading/written	CAR	9.7	2.9	—	9.2	2.6	—	35	−0.54	−0.77	−0.20
	CAS	9.1	3.2	—	8.5	3.0	—	35	−0.66	−0.81	−0.21
LD-mathematics	CA	8.9	3.1	6.06	10.0	3.0	3.03	33	1.06	1.43	0.35
	CAR	9.0	3.0	—	10.3	2.7	—	33	1.27	1.97	0.44
	CAS	8.9	3.0	—	9.6	2.8	—	33	0.70	0.95	0.24
LD-RWM	CA	9.8	3.0	2.38	9.9	2.9	2.38	42	0.07	0.11	0.02
	CAR	9.9	3.2	—	9.7	2.8	—	42	−0.19	−0.31	−0.06
	CAS	9.6	2.6	—	9.9	2.9	—	42	0.29	0.47	0.11
Mild ID	CA	6.2	2.5	17.46	9.9	2.7	1.59	63	3.65	8.05**	1.41
	CAR	7.1	2.3	—	9.8	2.5	—	63	2.62	6.14**	1.08
	CAS	6.0	2.2	—	9.9	2.7	—	63	3.90	8.73**	1.57
Moderate ID	CA	4.4	2.9	52.63	9.4	3.2	12.28	57	5.02	7.95**	1.63
	CAR	5.7	3.1	—	9.2	3.1	—	57	3.56	5.62**	1.15
	CAS	4.3	3.0	—	9.5	3.1	—	57	5.25	8.30**	1.72
Closed head injury	CA	7.9	3.8	25.93	11.3	2.2	0.00	27	3.37	3.64**	1.08
Closed head injury	CAR	8.3	3.2	—	11.1	2.6	—	27	2.85	3.01**	0.99
	CAS	8.2	4.1	—	11.2	2.4	—	27	3.00	3.51**	0.90
Open head injury	CA	8.1	4.0	18.75	10.9	2.7	0.00	16	2.81	2.01	0.82
Open head injury	CAR	8.4	4.1	—	10.6	3.1	—	16	2.13	1.55	0.59
	CAS	8.3	4.0	—	11.1	2.5	—	16	2.81	2.15*	0.85
Motor impairment	CA	5.9	3.0	28.57	9.0	3.0	4.76	21	3.14	3.12**	1.05
Motor impairment	CAR	6.7	2.9	—	9.3	2.6	—	21	2.67	2.86**	0.98
	CAS	5.9	3.2	—	8.9	2.8	—	21	3.00	3.15**	1.00

Note. CA= Cancellation total; CAR = Cancellation random; CAS = Cancellation structured; ADHD = Attention-Deficit/hyperactivity disorder; ADHD/LD = Attention-deficit/hyperactivity disorder and learning disorder; LD = Learning disorder; RWM = Reading, written expression, and mathematics; ID = Intellectual disabilities.

p < .05; **p < .01

Table 3 presents discrepancies in cancellation-random and cancellation-structured scores between clinical and matched control groups. Children with intellectual disabilities, Asperger’s disorder, and ADHD showed the largest discrepancies between random and structured tasks. Children with intellectual disabilities and Asperger’s disorder showed a slight advantage under random conditions compared to structured conditions. However, children with ADHD benefited significantly from the alignment of stimuli under structured conditions.

Table 3.

Group Differences in Cancellation Random Minus Structured Discrepancy Score.

	Clinical Group		Matched Control Group		Group Mean Comparison
Group Type	M	SD	M	SD	N	Diff	t	d
ADHD	−0.8	2.7	0.2	2.5	89	1.03	3.27**	.40
ADHD/LD	−0.5	2.8	0.3	2.1	45	0.80	1.64	.32
Autistic disorder	−0.2	2.4	−0.6	2.3	19	−0.42	−0.54	−.18
Asperger’s disorder	0.9	2.6	−0.4	2.6	27	−1.33	−1.89	−.51
Expressive language disorder	−0.2	1.9	−0.3	2.0	27	−0.19	−0.35	−.10
Mixed language disorder	0.6	2.0	0.1	2.6	40	−0.55	−1.03	−.24
LD-Reading	0.8	2.8	0.2	2.5	56	−0.54	−1.15	−.20
LD-reading/written	0.6	3.0	0.7	2.0	35	0.11	0.21	.04
LD-mathematics	0.1	2.1	0.7	2.0	33	0 .58	1.28	.29
LD-RWM	0.3	2.7	−0.2	2.7	42	−0.48	−0.83	−.18
Mild ID	1.2	1.9	−0.1	2.3	63	−1.29	−3.26**	−.62
Moderate ID	1.4	2.7	−0.3	1.9	57	−1.68	−3.47**	−.71
Closed head injury	0.0	2.2	−0.1	2.7	27	−0.15	−0.28	−.06
Open head injury	0.1	3.3	−0.6	2.7	16	−0.69	−0.75	−.23
Motor impairment	0.8	1.6	0.4	1.8	21	−0.33	−0.82	−.20

Note. ADHD = Attention-Deficit/hyperactivity disorder; ADHD/LD = Attention-deficit/hyperactivity disorder and learning disorder; LD = Learning disorder; RWM = Reading, written expression, and mathematics; ID = Intellectual disabilities.

p < .01

Evaluation of the Incremental Diagnostic Utility of Cancellation Subtest

Cancellation was found to improve the predictive power of FSIQ and GAI-CPI discrepancy scores on four clinical types: reading disorders; mild intellectual disabilities; closed head injuries; and motor impairments. The results of the sequential logistic regression analyses of these four clinical types are presented in Table 4. Nonsignificant results are available by request.

Table 4.

Sequential Logistic Regression Analysis: Predictors of Clinical Status.

Predictors	β	SE	Wald χ²	p	Odds Ratio	95%CI	R²	Hit rate	Δχ²	p
Reading disorder
Step 1							.23	.72
FSIQ	−.11	.02	19.88	< .0001	0.90	0.86-0.94
GAI-CPI	.01	.02	0.37	.5407	1.01	0.97-1.1
Step 2							.28	.74	6.91	.0086
FSIQ	−.12	.03	22.30	< .0001	0.88	0.83-0.93
GAI-CPI	.02	.02	0.96	.3274	1.02	0.98-1.06
Cancellation	.22	.09	6.22	.0126	1.24	1.05-1.47
Intellectual disabilities–mild severity
Step 1							.67	.92
FSIQ	−.29	.07	17.50	< .0001	0.75	0.66-0.86
GAI-CPI	−.03	.05	0.34	.5583	0.97	0.88-1.07
Step 2							.69	.95	6.11	.0134
FSIQ	−.28	.08	13.99	.0002	0.76	0.65-0.88
GAI-CPI	−.05	.06	0.78	.3776	0.95	0.86-1.06
Cancellation	−.38	.17	4.91	.0267	0.68	0.49-0.96
Closed head injury
Step 1							.32	.73
FSIQ	−.11	.03	11.04	.0009	0.90	0.84-0.96
GAI-CPI	.08	.04	4.17	.0411	1.08	1.01-1.17
Step 2							.39	.75	5.74	.0166
FSIQ	−.10	.03	7.78	.0053	0.91	0.85-0.97
GAI-CPI	.06	.04	1.93	.1644	1.06	0.98-1.15
Cancellation	−.27	.13	4.65	.0311	0.76	0.59-0.98
Motor impairment
Step 1							.21	.64
FSIQ	−.05	.03	3.17	.07	0.96	0.91-1.00
GAI-CPI	.08	.04	3.55	.06	1.09	1.00-1.18
Step 2							.28	.67	3.87	.0490
FSIQ	−.03	.03	0.92	.34	0.97	0.92-1.03
GAI-CPI	.06	.04	1.98	.16	1.06	0.98-1.16
Cancellation	−.27	.15	3.30	.07	0.77	0.57-1.02

For the reading disorder group, inclusion of the cancellation subtest statistically improved the model (p = .0086). As the logistic coefficient for cancellation is 0.22, when FSIQ and GAI-CPI discrepancies are statistically controlled, children with one increase in cancellation scaled score have a 0.22 increase in their log odds of being in the reading disabled group. The odds of 1.24 also meant that after FSIQ and GAI-CPI discrepancy scores were adjusted, each increase in cancellation scaled score increased the odds of being in the reading disabled group by 24%. Classification analysis showed that 74% of all cases were correctly classified by the probabilities derived from the model using all three independent variables (compared to an overall hit rate of 72% for the model without the cancellation subtest).

All three predictors significantly improved prediction of mild intellectual disabilities (p = .0134). The logistic coefficient for the cancellation subtest was −0.38 and the odds ratio was 0.68, suggesting that for every one unit increase the cancellation subtest scaled score, the odds of being mildly intellectually disabled decreased by 32%. Of the cases, 95% were correctly classified by the model using all three predictors, of which 3% was contributed by the cancellation subtest.

For the closed head injury group, Δχ² also showed significant differences (p = .0166) between the two- and three-predictor models. The logistic coefficient for the cancellation subtest was −0.27, and the odds ratio was 0.76. This suggested that for every one point increase in the cancellation subtest scaled score, the odds of being in the closed head injury group decreased by 24%. Inclusion of the cancellation subtest raised the classification hit rate from 73% to 75%.

Finally, the motor impairment group showed a significant (p = .0490) difference between the models. The logistic coefficient for the cancellation subtest was −0.27, with odds of 0.77. Thus, for every scaled score increase in the cancellation subtest, the odds of being in the motor impairment group decreased by 23%. Classification analysis showed that 67% of the cases were correctly classified by the probabilities derived from the model using all three independent variables (compared to an overall hit rate of 64% for the model without the cancellation subtest).

Discussion

The results supported the clinical utility of the cancellation subtest. Children with individual disabilities, motor impairments, head injuries, Autistic or Asperger’s disorder, ADHD/ LD, and mathematics disorder showed significant deficits in performance on the cancellation subtest. Children with intellectual disabilities and Asperger’s disorder performed better when stimuli were randomly aligned. Children with ADHD benefited significantly from structured conditions. The cancellation subtest contributed extra predictive power to that explained by the FSIQ and GAI-CPI discrepancy scores for identifying children with reading disorders, mild individual disabilities, closed head injuries, and motor impairments.

The findings that children with intellectual disabilities and head injuries performed poorly on the cancellation subtest was expected because processing speed plays a crucial role in the development of intelligence (Kail, 2000), and is sensitive to brain dysfunction (Allen et al., 2010; Calhoun & Mayes, 2005). Performance on cancellation revealed information on general cognitive development and brain function. Children with motor impairments, Autistic or Asperger’s disorder, and mathematics disorder appeared to have a common characteristic. Many of them have difficulties in managing nonverbal (pictorial) materials and perceptual–motor coordination (Boreas, Rintala, & McCubbin, 2009; Panaoura, 2006; Taylor, Espy, & Anderson, 2009; Wechsler, 2003b). Children with these clinical types may also show delayed developmental coordination (Henderson & Henderson, 2002). As cancellation requires children to scan arrangement of pictures and mark target pictures in a timely fashion, deficit performance on this task may reveal their difficulties in processing pictorial materials in a timely fashion and their poor perceptual–motor coordination.

Structured instruction is recognized as a useful strategy for teaching children with ADHD (Walcott & Landau, 2004; Young & Amarasinghe, 2010). Our results from the cancellation subtest coincided with this line of research (Flanagan & Kaufman, 2004). The reason intellectually disabled children and children with Asperger’s syndrome process simple pictorial figures faster under random conditions was less obvious. When a task requires more cognitive ability, structured guidance is assumed to help. When a task is relatively easy, structured guidance may hinder these children because they are known to have more rigid and less flexible thinking patterns (Mintz, 2008; Polirstok & Houghteling, 2006). In the cancellation subtest, structured alignments may imperceptibly lead these children to process each figure in sequence, resulting in lowered performance. Our results suggest that the effects of various stimuli presentation on these types of children should be more carefully differentiated. Structured alignments may not always work the best.

Profile analyses showed that FSIQ typically represented overall cognitive ability, and GAI-CPI discrepancy provided hints for breadth in ability. With the well-documented predictive power of these two composite-based indices (Glutting, Watkins, Konold, & McDermott, 2006; Lynn & Vanhanen, 2006; Pesta, McDaniel, & Bertsch, 2010; Watkins, Glutting, & Lei, 2007; Weiss & Gabel, 2008), and the quick and easy administration of the cancellation subtest, it was noteworthy that the cancellation subtest contributed statistically significant power to identification. After controlling for the FSIQ and GAI-CPI discrepancy, reduced cancellation performance was associated with greater probability of mild intellectual disabilities, closed head injuries, and motor impairments. In comparison, children with reading disabilities had a higher probability of obtaining higher cancellation subtest scores once the FSIQ and GAI-CPI discrepancy was controlled. The processing speed for stimuli of different modalities (no meaningful symbols vs. meaningful pictorial figures) appeared to have a difference. A follow-up examination revealed that children with reading disabilities showed significant deficits in performance on two other processing speed tests, coding (t = 2.47, p < .05), and symbol search (t = 2.01, p < .05). Both of these subtests required children to process nonmeaningful symbols quickly.

In general, the current study examined the practical usefulness of the WISC-IV cancellation subtest in a broad perspective. However, the limitations deserve attention. First, our study examined 15 major clinical types but was not exclusive. There are other clinical groups that need to be studied. Second, these clinical samples were not randomly selected but were selected based on availability by educational and clinical professionals across states. Certain clinical groups had small sample sizes, and cross-validation was not able to be conducted. Future research is recommended to employ large, representative, randomly selected, and independent samples, with independent validation when possible. Third, the numerous comparisons and the relative marginal levels of significance for certain findings could be a weakness. Since this is the first study in this area, we interpreted the conventional significance levels and effect sizes. More conservative data interpretation is suggested. Finally, we judged the subtest utility primarily by its predictive power. Other clinically meaningful qualitative information from this subtest were not examined, such as the location of first cancellation, systematic slowing with time, and function of target location. This information is useful for assessing people with right hemisphere injuries (Samuelsson, Hjelmquist, Naver, & Blomstrand, 1996) and unilateral spatial neglect (Manly et al., 2009). Therefore, the findings of this study should not be considered as the sole use of the cancellation subtest.

Overall, our findings supported the utility of the WISC-IV cancellation subtest. Children with several clinical types showed reduced performance on this subtest. In addition, this subtest was found to be able to improve the predictive power of FSIQ and GAI-CPI discrepancy in four clinical types. The current results permit future comparisons and interpretations.

Footnotes

Authors’ Note

For readers interested in the details of the inclusion criteria for participants, please contact the corresponding author for a summary.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Allen

D. N.

Thaler

N. S.

Donohue

Mayfield

(2010). WISC-IV profiles in children with traumatic brain injury: Similarities to and differences from the WISC-III. Psychological Assessment, 22(1), 57-64.

Borremans

Rintala

McCubbin

J. A.

(2009). Motor skills of young adults with Asperger Syndrome: A cinoarative study. European Journal of Adapted Physical Activity, 2(1), 21-33.

Calhoun

Mayes

S. D.

(2005). Processing speed in children with clinical disorders. Psychology in the Schools, 42, 333-343.

Carroll

J. B.

(1993). Human cognitive abilities: A survey of factor analytic studies. New York, NY: Cambridge University Press.

Carroll

J. B.

(2005). The three-stratum theory of cognitive abilities. In Flanagan

D. P.

Harrison

P. L.

(Eds.), Contemporary intellectual assessment: Theories, tests, and issues (2nd ed., pp. 69-76). New York, NY: Guilford.

Chen

Keith

Weiss

Zhu

(2010). Testing for multigroup invariance of second-order WISC-IV structure across China, Hong Kong, Macau, and Taiwan. Personality and Individual Differences, 49, 677-682.

Ferber

Karnath

H. O.

(2001). How to assess spatial neglectline bisection of cancellation tasks? Journal of Clinical and Experimental Neuropsychology, 23, 599-607.

Flanagan

D. P.

Kaufman

A. S.

(2004). Essentials of WISC-IV assessment. Hoboken, NJ: Wiley.

Fry

A. F.

Hale

(1996). Processing speed, working memory, and fluid intelligence: Evidence for a developmental cascade. Psychological Science, 7, 237-241.

10.

Fry

A. F.

Hale

(2000). Relationships among processing speed, working memory, and fluid intelligence in children. Biological Psychology, 54, 1-34.

11.

Glutting

J. J.

Watkins

M. W.

Konold

T. R.

McDermott

P. A.

(2006). Distinctions without a difference: The utility of observed versus latent factors from the WISC-IV in estimating reading and math achievement on the WIAT-II. Journal of Special Education, 40, 103-114.

12.

Halligan

P. W.

Marshall

j. c.

Wade

D. T.

(1989). Visuospatial neglect: Underlying factors and test sensitivity. Lancet, 2, 908-911.

13.

Henderson

(2002). Toward an understanding of Developmental Coordination Disorder. Adapted Physical Activity Quarterly, 19, 12-31.

14.

Kail

R. V.

(2000). Speed of information processing: Developmental change and links to intelligence. Journal of School Psychology, 38(1), 51-61.

15.

Kail

R. V.

(2007). Longitudinal evidence that increases in processing speed and working memory enhances children’s reasoning. Psychological Science, 18, 312-313.

16.

Kaufman

A. S.

Linchtenberger

E. O.

(2002). Assessing adolescent and adult intelligence (2nd ed.). Boston, MA: Allen and Bacon.

17.

Keith

T. Z.

Fine

J. G.

Taub

G. E.

Reynolds

M. R.

Kranzler

J. H.

(2006). Higher order, multisample, confirmatory factor analysis of the Wechsler Intelligence Scale for Children-Fourth Edition: What does it measure? School Psychology Review, 35, 108-127.

18.

Lezak

M. D.

Howieson

D. B.

Bigler

E. D.

Tranel

(2012). Neuropsychological assessment. (5th ed.).New York, NY: Oxford University Press.

19.

Lynn

Vanhanen

(2006). IQ and global inequality. Augusta, GA: Washington Summit Publishers.

20.

Manly

Dove

Blows

Teasdale

T. W.

Dodds

C. M.

Fish

Warburton

(2009). Assessment of unilateral spatial neglect: Scoring Star Cancellation performance from video recordings-method, reliability, benefits, and normative data. Neuropsychology, 23, 519-528.

21.

Matarazzo

J. D.

Herman

D. O.

(1985). Clinical uses of the WAIS-R: base rates of differences between VIQ-PIQ in the WAIS-R standardization sample. In Wolman

B. B.

(Ed.), Handbook of intelligence: Theories, measurements, and applications (pp. 899-932). New York, NY: John Wiley.

22.

Menard

(2000). Coefficients of determination for multiple logistic regression analysis. American Statistician, 54(1), 17-24.

23.

Miller

D. C.

Hale

J. B.

(2008). Neuropsychological applications of the WISC-IV and WISC-IV integrated. In Prifitera

Saklofske

D. H.

Weiss

L. G.

(Eds.), WISC-IV clinical assessment and intervention 2e (pp. 445-495). San Diego, CA: Elsevier Academic Press.

24.

Mintz

(2008). Working with children with Asperger’s Syndrome in the mainstream classroom; A psychodynamic take from the chalk face. Psychodynamic Practice, 14, 169-180.

25.

Nettelbeck

Burns

N. R.

(2010). Processing speed, working memory and reasoning ability from childhood to old age. Personality and Individual Differences, 48, 379-384.

26.

Panaoura

(2006). The development of young pupil’s self-representation and mathematical performance in relation to processing efficiency and working memory. Educational Psychology, 26, 643-676.

27.

Pesta

B. J.

McDaniel

M. A.

Bertsch

(2010). Toward an index of well-being for the fifty U.S. states. Intelligence, 38, 160-168.

28.

Polirstok

S. R.

Houghteling

(2006). Asperger Syndrome: A primer for behavioral interventions. Journal of Early and Intensive Behavior Intervention, 3, 187-195.

29.

Prifitera

Saklofske

D. H.

Weiss

L. G.

(2008). WISC-IV clinical assessment and intervention (2nd ed.). San Diego, CA: Academic Press.

30.

Raiford

Weiss

L. G.

Rolfhus

Coalson

(2005). General ability index (WISC-IV technical report no. 4). Retrieved from http://www.pearsonassessments.com/pai/ca/research/publications/techrpts/reslist.htm.

31.

Spicer

(2005). Making sense of multivariate data analysis: An intuitive approach. Thousand Oaks, CA: SAGE.

32.

Saklofske

D. H.

Zhu

Coalson

D. L.

Railford

S. E.

Weiss

L. G.

(2010). Cognitive Proficiency Index for the Canadian edition of the Wechsler Intelligence Scale for Children-Fourth Edition. Canadian Journal of School Psychology, 25, 277-286.

33.

Saklofske

D. H.

Zhu

Miller

J. L.

Weiss

L. G.

Babcock

S. E.

Cayton

T. G.

Coalson

D. L.

(2012). The cognitive proficiency index for the Canadian edition of the Wechsler Adult Intelligence Scale-Fourth Edition. Canadian Journal of Behavioural Science, 44, 117-123.

34.

Samuelsson

Hjelmquist

Naver

Blomstrand

(1996). Visuospatial neglect and an ipsilesional bias during the start of performance in conventional tests of neglect. Clinical Neuropsychologist, 10, 15-24.

35.

Sattler

J. M.

Dumont

(2004). Assessment of children: WISC-IV and WPPSI-III supplement. Le Mesa, CA: Author.

36.

Taylor

H. G.

Espy

K. A.

Anderson

P. J.

(2009). Mathematics deficiencies in children with very low birth weight or very preterm birth. Developmental Disabilities, 15, 52-59.

37.

Vanier

Gauthoer

Lambert

Pepin

E. P.

Robillard

Dubouloz

C. J.

Joanette

(1990). Evaluation of left visuospatial neglect: Norms and discrimination power of two tests. Neuropsychology, 4, 87-96.

38.

Walcott

C. M.

Landau

(2004). The relationship between disinhibition and emotion regulation in boys with ADHD. Journal of Clinical Child and Adolescent Psychology, 33, 772-782.

39.

Watkins

M. W.

Glutting

J. J.

Lei

(2007). Validity of the full-scale IQ when there is significant variability among WISC-III and WISC-IV factor scores. Applied Neuropsychology, 14(1), 13-20.

40.

Wechsler

(2003a). Wechsler Intelligence Scale for Children–Fourth Edition. Administration and scoring manual. San Antonio, TX: Harcourt Assessment.

41.

Wechsler

(2003b). Wechsler Intelligence Scale for Children–Fourth Edition. Technical and interpretive manual. San Antonio, TX: Harcourt Assessment.

42.

Weiss

L. G.

Beal

A. L.

Saklofske

D. H.

Alloway

T. P.

Prifitera

(2008). Interpretation and intervention with WISC-IV in the clinical assessment context. In Prifitera

Saklofske

D. H.

Weiss

L. G.

(Eds.), WISC-IV clinical assessment and intervention 2e (pp. 3-66). San Diego, CA: Elsevier Academic Press.

43.

Weiss

L.G.

Gabel

A. D.

(2008). Using the cognitive proficiency index in psychoeducational assessment (WISC-IV technical report no. 6). Retrieved from http://www.pearsonassessments.com/pai/ca/research/publications/techrpts/reslist.htm.

44.

Weiss

L. G.

Keith

Zhu

Chen

(in press). WISC-IV and Clinical Validation of the Four- and Five-Factor Interpretative Approaches. Journal of Psychoeducational Assessment.

45.

Weiss

L. G.

Prifitera

Saklofske

D. H.

(2005). Interpreting the WISC-IV index scores. In Prifitera

Saklofske

D. H.

Weiss

L. G.

(Eds.), WISC-IV clinical use and interpretation (pp. 71-100). San Diego, CA: Elsevier Academic Press.

46.

Weiss

L. G.

Prifitera

Holdnack

J. A.

Saklofske

D. H.

Rolfhus

Coalson

(2006). The essentials and beyond. In Weiss

L. G.

Saklofske

D. H.

Prifitera

Holdnack

J. A.

(Eds.), WISC-IV advanced clinical interpretation (pp. 1-57). San Diego, CA: Elsevier Academic Press.

47.

Woodcock

R. W.

McGrew

K. S.

Mather

(2001). Woodcock-Johnson III tests of cognitive abilities. Itasca, IL: Riverside.

48.

Young

Amarasinghe

(2010). Practioner review: Non-pharmacological treatments for ADHD: A lifespan approach. Journal of Child Psychology and Psychiatry, 51(2), 116-133.

49.

Zhu

Railford

S. E.

Coalson

D. L.

(2009, August). The what, when, and how of the Wechsler General Ability Index. Paper presented at the 117th annual convention of American Psychological Association, Toronto, Ontario, Canada.