Applied Neuropsychology: Child

ISSN: 2162-2965 (Print) 2162-2973 (Online) Journal homepage: https://www.tandfonline.com/loi/hapc20

Examination of five- and four-subtest short form

IQ estimations for the Wechsler Intelligence Scale
for Children-Fifth edition (WISC-V) in a mixed
clinical sample

John W. Lace, Zachary C. Merz, Erin E. Kennedy, Dylan J. Seitz, Tara A. Austin,
Bradley J. Ferguson & Michael D. Mohrland

To cite this article: John W. Lace, Zachary C. Merz, Erin E. Kennedy, Dylan J. Seitz, Tara A.
Austin, Bradley J. Ferguson & Michael D. Mohrland (2020): Examination of five- and four-subtest
short form IQ estimations for the Wechsler Intelligence Scale for Children-Fifth edition (WISC-V) in
a mixed clinical sample, Applied Neuropsychology: Child, DOI: 10.1080/21622965.2020.1747021

To link to this article: https://doi.org/10.1080/21622965.2020.1747021

Published online: 16 Apr 2020.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=hapc20
APPLIED NEUROPSYCHOLOGY: CHILD
https://doi.org/10.1080/21622965.2020.1747021

Examination of five- and four-subtest short form IQ estimations for the Wechsler
Intelligence Scale for Children-Fifth edition (WISC-V) in a mixed clinical sample
John W. Lacea,b, Zachary C. Merzc, Erin E. Kennedyb, Dylan J. Seitzb, Tara A. Austinb, Bradley J. Fergusonb,d,e,
and Michael D. Mohrlandb,e
a
Department of Psychology, Saint Louis University, Saint Louis, MO, USA; bDepartment of Health Psychology, Univerisity of Missouri,
Columbia, MO, USA; cDepartment of Physical Medicine and Rehabilitation, University of North Carolina, Chapel Hill, NC, USA; dDepartment
of Radiology, University of Missouri, Columbia, MO, USA; eThompson Center for Autism & Neurodevelopmental Disorders, Columbia,
MO, USA

ABSTRACT KEYWORDS
Evaluating general cognitive ability (i.e., intelligence) is common in neuropsychological practice, Psychological assessment;
and identifying abbreviated assessments of intelligence is often advantageous. Despite the WISC-V; short form; IQ;
Wechsler Intelligence Scale for Children, Fifth Edition’s (WISC-V) widespread contemporary use, lim- test validation
ited research has identified clinically useful short form (SF) full scale IQ (FSIQ) estimations in clin-
ical samples. This study sought to address this gap in the literature. Two hundred sixty-eight
pediatric participants (M age ¼ 9.79; 69% male; 88% Caucasian/White) who underwent psycho-
logical/neuropsychological evaluation were included. Mean scores for WISC-V scores fell in the low
average-to-average ranges, consistent with the clinical nature of this sample (e.g., M FSIQ ¼ 85.3).
Ten unique SF combinations with five (pentad) and four (tetrad) subtests, each intentionally
selected to represent a breadth of domains subsumed by WISC-V FSIQ, were described by sum-
ming subtest age-corrected scaled scores. Regression-based and prorated FSIQ estimates were cal-
culated, and mean differences suggested some prorated estimates should be arithmetically
adjusted. Both regression-based and prorated/adjusted methods provided FSIQ estimates that
were accurate within five Standard Score points of true FSIQ for approximately 81–92% (pentad)
and 65–76% (tetrads) of participants. Prorated/adjusted estimates appeared to provide somewhat
better accuracy than regression-based estimates. Relationships between SFs and true FSIQ did not
appear to be moderated by participant age, gender, nor how many WISC-V subtests were adminis-
tered to participants within this archival sample (i.e., 7 vs. 10). Implications of these findings,
including benefits, detriments, and other considerations of each SF combination, in addition to
limitations of this study, are discussed in detail.

The assessment of general cognitive ability (i.e., intelligence) Identifying psychometrically appropriate short forms
as part of psychological and neuropsychological evaluation is (SFs) for routinely used assessment batteries is a prudent
clinically valuable (Flanagan & McDonough, 2018), particu- task for several reasons. First, SFs reduce clinical time spent
larly for children and adolescents (Weiss et al., 2015), given measuring a single cognitive construct (Donders, 2001).
its relationship with various neurocognitive abilities (Lung Given that obtaining WISC-V FSIQ may require
et al., 2018; Seidenberg et al., 1983), academic achievement “administration time of approximately 65-80 minutes”
(Mayes et al., 2009; Mayes & Calhoun, 2007), and other (Doyle & Mattson, 2015, p. 176), minimizing testing time
domains of life success (Bergman et al., 2014; Firkowska- may result in reduced costs to the examinee or their payor,
Mankiewicz, 2011). While some may critique the inclusion which is an important consideration in many neuropsycho-
of measures of intelligence in neuropsychological assessment logical settings (Ricker, 1998; Soldner et al., 2009).
(Ardila, 1999; Lezak, 1988), others argue that it is critical Moreover, by reducing time spent assessing one cognitive
and provides rich, clinically meaningful data in psycho- construct, more attention may be paid to administering tests
logical evaluations (Keith, 1994; Pfeiffer et al., 2000). assessing other clinically relevant domains not thoroughly
Notably, derivations of the Wechsler Intelligence Scale for assessed by the WISC (e.g., executive functioning, learning/
Children (WISC; see Cormier et al., 2016, Kaufman et al., memory) that may be specifically related to the examinee’s
2006, and Na & Burns, 2016 for histories and reviews of the presenting concerns or to be used to expand the clinical
WISC and its various iterations) remain the most frequently interview (Donders, 2001). Second, given that “long cogni-
used assessment tools designed specifically for use with chil- tive tests … can be stressful to [examinees] suffering from,
dren and adolescent examinees (Egeland et al., 2016; Rabin for example, concentration and attention problems, chronic
et al., 2016). physical fatigue, or brain damage due to hereditary defects

CONTACT John W. Lace John.Lace@slu.edu Department of Psychology, Saint Louis University, Saint Louis, MO 63103, USA.
ß 2020 Taylor & Francis Group, LLC
2 J. W. LACE ET AL.

or traumatic events” (Emons et al., 2007, p. 105), reducing above 125 (Aubry & Bourdin, 2018). Importantly, these SFs
the testing burden on individuals may reduce excessive have marked limitations and possibly low generalizability, as
fatigue and prevent diminished tolerance toward testing. they have not been validated in clinical samples. Moreover,
These burden reductions may be particularly beneficial for Piovesana and colleagues’ (2019) formula includes scores
examinees with notable behavioral disturbances (e.g., exter- from subtests that do not normally contribute to WISC-V
nalizing behavior) or other difficulties engaging in testing FSIQ (i.e., Visual Puzzles and Letter-Number Sequencing)
and those with gross cognitive impairments for whom pro- and altogether omits all subtests corresponding to processing
longed neuropsychological testing may pose problems speed. Aubry and Bourdin (2018) formulas utilized simu-
(Calhoun & Newson, 1984; Stuss et al., 1996). lated data for IQs not representative of the general popula-
Given the mainstream contemporary use of short or tion. Overall, the identification and validation of clinically
quick IQ tests (Thompson et al., 2004), extensive work has applicable SFs in the WISC-V remains a notable gap in
been dedicated to the topic of identifying appropriate SFs,
the literature.
with decades of literature focusing on this topic (Aubry &
Of note, the WISC-V underwent meaningful alterations
Bourdin, 2018; Campbell, 1998; Goh, 1980; Merz et al.,
compared to previous versions (see Na & Burns, 2016 for a
2019). Specifically regarding early versions of the WISC,
review), including notable changes to available subtests (e.g.,
several studies have sought to describe, validate, and critique
adding Picture Span), number of index scores (i.e., five
various SFs in diverse samples (Carleton & Stacey, 1954;
instead of four, though this structure is strongly disputed;
Dirks et al., 1980; Kaufman, 1976; Ortiz & Gonzalez,
1989; Ryan, 1981; Silverstein, 1970; Yalowitz & Armstrong, Canivez et al., 2016, 2017; Canivez et al., 2018; Dombrowski
1955; Yudin, 1966). Similarly, the WISC-III received exten- et al., 2018), and an alteration in the calculation of FSIQ.
sive attention in the context of SF development, with several That is, the WISC-V uses seven subtests to calculate FSIQ
possible combinations proposed and validated during and (Similarities, Vocabulary, Block Design, Matrix Reasoning,
beyond the WISC-III’s period of contemporary use (Beal Figure Weights, Digit Span, and Coding) rather than a
et al., 1996; Comninel & Bordieri, 2001; Connery et al., weighted sum of index scores (which required administra-
1996;; Donders, 1997; 2001; Donders & Warschausky, 1996; tion of 10 subtests) in its predecessor. Practitioners may
Dumont & Faro, 1993; Kaufman et al., 1996; Mark et al., employ clinical preference in deciding to administer either
1998; van Ool et al., 2018). only these seven or more subtests (Wechsler, 2014).
Regarding the WISC-IV, which required the administra- Nonetheless, there appears to be notable scholarly and
tion of 10 subtests in deriving FSIQ, Crawford et al. (2010) clinical interest in the identification and validation of SF
described a seven-subtest SF (including Vocabulary, FSIQ estimates for WISC iterations to increase efficiency
Similarities, Block Design, Matrix Reasoning, Digit Span, across settings. The interest in streamlined assessment proc-
Coding, and Symbol Search) from WISC-IV normative data. esses, perhaps in part, led to the creation of the Wechsler
Later research validated this seven-subtest SF in samples of Abbreviated Scale of Intelligence (WASI) and its newer ver-
children referred for possible intellectual disability sion (WASI-II; Wechsler, 2011), which are distinct, compen-
(McKenzie et al., 2014) and in those with epilepsy, the latter dious instruments in obtaining an estimate of global
suggesting that 85% of FSIQ estimates were within 5 points cognitive ability. Unfortunately, the WASI-II, while efficient,
of true FSIQ (Hrabok et al., 2014). Hrabok and colleagues may not be psychometrically equivalent to longer form
(20144) also suggested that a five-subtest SF (Block Design, Wechsler measures of intelligence (Adaki, 2009; Axelrod,
Similarities, Coding, Vocabulary, and Letter-Number 2002) as it captures only certain aspects of cognitive ability
Sequencing) provided similar accuracy, as 81% of FSIQ esti- (i.e., verbal compression and perceptual reasoning) while
mates were correct within 5 points. Donders et al. (2013)
ignoring the contributions of the domains of working mem-
described an eight-subtest SF that yielded excellent compar-
ory and processing speed. Importantly, working memory
ability with the full-length FSIQ and demonstrated clinical
and processing speed may be particularly sensitive to neuro-
utility in identifying children with intellectual disability
developmental, neurological, and psychiatric dysfunction
(Murray et al., 2016). In a sample of Spanish children, Dasi
and may be important to consider when estimating global
et al. (2014) reported that prorated sums of scores for the
four-subtest SF of Vocabulary, Block Design, Letter-Number cognitive ability (Allen et al., 2010; Calhoun & Mayes, 2005;
Sequencing, and Coding provided strong accuracy compared De Clercq-Quaegebeur et al., 2010; San Miguel Montes
to true FSIQ. et al., 2010). Thus, it is important to be mindful of domain
Regarding the WISC-V, relatively less work has investi- coverage and subtest selection when considering validity and
gated SFs within this most recent iteration. Review of extant utility of derived FSIQ estimations.
literature identified only two articles exploring SFs in this Unfortunately, to date, no previous work has sought to
regard. One described a motor-free, six-subtest SF identify SF FSIQ estimations in the WISC-V that provide
(Vocabulary, Similarities, Visual Puzzles, Figure Weights, broad domain coverage and utility in a clinical sample. As
Digit Span, and Letter-Number Sequencing; Piovesana et al., such, the goal of the present study was to examine the clin-
2019) based on normative data. Another identified adjusted ical utility of 10 unique five- and four-subtest short forms
formulas for two- and four-subtest SFs using simulated data that offer appropriate domain/construct coverage for the
in the context of intellectually gifted children with IQs at or WISC-V in a mixed clinical sample of pediatric examinees.
 APPLIED NEUROPSYCHOLOGY: CHILD 3

Method & Canivez, 2018). While its five-factor structure is reported

in the standardization sample (Chen et al., 2015; Wechsler,
Participants
2014), other empirical work has suggested it better repre-
Participant data were extracted from an IRB-approved arch- sents four factors combining VSI and FRI (Canivez et al.,
ival dataset of examinees who were referred for and com- 2016, 2018; Dombrowski et al., 2018). In the present study,
pleted a psychological or neuropsychological evaluation at FSIQ was calculated according to Wechsler (2014), such that
the Thompson Center for Autism and Neurodevelopmental the sum of scaled scores (Ms ¼ 10, SDs ¼ 3) for
Disorders between 2015 and 2019. The Thompson Center is Vocabulary, Similarities, Block Design, Matrix Reasoning,
an outpatient, pediatric clinic at the University of Missouri. Figure Weights, Digit Span, and Coding was converted to a
Participants with complete data for WISC-V FSIQ and the FSIQ (Wechsler, 2014, Table A.7) represented by a Standard
seven subtests comprising it were of interest. Of 288 avail- Score (M ¼ 100, SD ¼ 15).
able cases, 268 participants were included in the final data-
set. Twenty cases with questionable or noncredible
performance (as determined by either Test of Memory Procedures
Malingering Trial 2 scores <45 or Rey Fifteen Item Test The present study was approved Health Sciences
recall scores <9; Green et al., 2014; Rees et al., 1998) were Institutional Review Board at the University of Missouri. All
excluded; of note, clinical judgment was used to decide participants were referred for and completed a neuropsycho-
whether or not to include a standalone performance validity logical evaluation between 2015 and 2019. Participants were
test, and not every examinee was administered one. While administered either the seven- or 10-subtest WISC-V as part
consistent with trends in contemporary clinical practice of clinical assessment by either a trained psychometrist, doc-
(DeRight & Carone, 2015; MacAllister et al., 2019), this limi- toral level clinical psychology graduate student, predoctoral
tation is discussed in greater detail in the discussion section. intern, or postdoctoral neuropsychology fellow proficient in
Relatedly, reliable digit span was not considered as an indi- WISC-V administration under supervision by a licensed
cator of performance validity as previous work has identified psychologist. Those who were administered either seven or
high levels of false positives in clinical samples when consid-
10 subtests did not significantly differ in terms of age
ering this variable (Blaskewitz et al., 2008; Ventura et al.,
[t(266) ¼ 1.20, p ¼ .233], gender (Fisher’s Exact Test p ¼
2019; Welsh et al., 2012).
.144), nor racial/ethnic identity (coded as White or
The final sample included 185 males and 83 females.
Nonwhite; Fisher’s Exact Test p ¼ .848). Thus, all partici-
Participants ages ranged from 6 to 16 (M ¼ 9.79, SD ¼ 2.48).
pants were collapsed. All participants had complete data for
Most participants were Caucasian/White (88.4%), with 5.6%
the seven subtests comprising FSIQ.
Black/African-American, 1.1% Hispanic/Latinx, 0.4% Asian/
Asian-American, and 2.6% reporting other racial/ethnic
identity. Regarding diagnoses, 7.8% did not receive any clin- Statistical analyses
ical diagnosis, 19.8% received only one diagnosis, 22.8%
received two diagnoses, 24.3% received three diagnoses, and Data were analyzed using SPSS 25.0. All analyses were per-
25.4% received four or more diagnoses. For brevity and suc- formed using standardized scores for WISC-V subtests and
cinctness, regarding primary diagnoses (i.e., those listed first FSIQ. Prior to performing analyses, data were examined for
in examinee’s charts), 30.9% were autism spectrum disor- univariate outliers (i.e., > j3.29j SD from mean; Field, 2013).
ders, 19.4% attention-deficit/hyperactivity disorders, 16.8% No data points for FSIQ or subtest scores were identified as
specific learning disorders, 7.8% psychiatric (e.g., depressive/ univariate outliers.
anxiety/trauma- and stressor-related) disorders, 6.3% other Ten unique SF combinations were created by summing
medical/neurocognitive/neurodevelopmental disorder (e.g., scaled scores for included subtests. Five-subtest SFs (i.e.,
cerebral palsy, chromosomal abnormality, seizures), and pentads) were identified, each of which included one subtest
5.2% intellectual disability or borderline intellectual contributing to FSIQ from each WISC-V index. Given that
functioning. WMI and PSI only contain one subtest which contribute to
the FSIQ calculation (i.e., Digit Span and Coding, respect-
ively), only four unique pentad combinations were possible.
Measures Next, in the context of previous work suggesting that the
WISC-V WISC-V may better fit a four-factor model merging VSI
The Wechsler Intelligence Scale for Children, Fifth Edition and FRI (Canivez et al., 2016, 2018; Dombrowski et al.,
(WISC-V; Wechsler, 2014) is the most widely used cognitive 2018), several four-subtest SFs (i.e., tetrads) were created
assessment for children (Rabin et al., 2016). It provides using one subtest contributing to FSIQ from each factor
scores representing full scale intelligence (FSIQ), Verbal (i.e., VCI, VSI/FRI, WMI, PSI). Again, because the latter
Comprehension (VCI), Visual Spatial (VSI), Fluid Reasoning two indices contribute only one subtest each to the FSIQ,
(FRI), Working Memory (WMI), and Processing Speed only six unique tetrad combinations were possible. Of note,
(PSI). Good reliability and validity have been reported for three- and two-subtest SFs were not considered as part of
the WISC-V across various populations (Flanagan & the present study, as they would likely underrepresent the
Alfonso, 2017; Kaufman et al., 2015; Watkins, Dombrowski, breadth of domain coverage assessed by FSIQ and may yield
4 J. W. LACE ET AL.

inaccurate estimations (Herrera-Graf et al., 1996; Wheaton approximately one standard deviation below the normative
& Vandergriff, 1987). mean (see Table 1). These relatively low scores (compared
FSIQ estimates (expressed in Standard Scores, M ¼ 100, to normative means) appeared consistent with the present
SD ¼ 15) for each SF were calculated via separate direct study’s inclusion of a mixed clinical sample. Correlations
entry linear regression equations; each SF sum of scaled between subtests and FSIQ are displayed in Table 2.
scores was entered as the sole predictor and WISC-V FSIQ Consistent with the excellent psychometric properties by the
was always entered as the outcome (Merz et al., 2019). WISC-V, all correlations between subtests and FSIQ were
Additionally, prorated sums of scaled scores for each SF significant at p < .01.
combination were calculated according to how many subt-
ests were included in each SF (Glass et al., 2008). That is,
pentad SFs were multiplied by 7/5 and tetrads by 7/4.These Estimating IQ from short forms
prorated sums of scaled scores were converted to FSIQ esti- Table 3 lists short form abbreviations and the subtests com-
mates according to WISC-V scoring manual (see Wechsler, prising each. Individual linear regression models were per-
2014, Table A.7). Paired-samples t-tests (with Holm- formed for each possible SF. Using these models, regression
Bonferroni-corrected p values, as needed) were calculated to equations were calculated, allowing for the sums of scaled
determine if SF estimates for regression-based and prorated scores for each SF to convert to a Standard Score (M ¼ 100,
methods differed significantly from true FSIQ scores. SD ¼ 15) akin to traditional FSIQ. Formulas derived from
Adjustments to select prorated FSIQ estimates were each of these individual linear regression analyses were pro-
described, as needed, according to t-test results. Tables vided in Table 3 for readers to calculate their own regres-
describing conversions between SFs and estimated FSIQ for sion-based estimated FSIQ according to these results.
both methods were provided. Frequencies of discrepancies Further results may be made available upon reasonable
between estimated and true FSIQ were calculated for each request from the first author. Additionally, sums of scaled
SF for both regression-based and prorated/adjusted esti- scores for each SF were prorated; that is, the sum of scaled
mates. Several demographic variables (e.g., participant age, scores for pentad and tetrad SFs were multiplied by 7/5 and
gender, and 7- vs. 10-subtest WISC-V administration) were 7/4 respectively. These prorated values were then used to
examined as possible moderator variables via hierarchical estimate FSIQ for each participant according to normative
regressions. Acronyms for SF combinations were created data from the WISC-V manual (Wechsler, 2014, Table A.7).
according to the subtests included within each. For example, As expected, Pearson’s correlations between true FSIQ and
SBMDC is the five-subtest SF comprised of the sum of both regression-based (rs ranged from .94 to .98) and pro-
scaled scores for Similarities, Block Design, Matrix rated estimated FSIQ (rs ranged from .94 to .98) were
Reasoning, Digit Span, and Coding. very strong.
The possible moderating effects of three demographic
Results variables on the relationship between SFs and true FSIQ
were examined. The possible moderating variables of interest
Descriptive results and correlations were: 7- vs. 10-subtest WISC-V administration, participant
age, and participant gender. Statistically, these variables were
Overall, average performance in this sample across WISC-V
examined as moderating factors via a traditional, two-block,
FSIQ, indices, and subtests appeared to be in the low aver-
hierarchical regression method. That is, true FSIQ was
age-to-average ranges, with mean FSIQ scores falling
always entered as the outcome variable. Dependent variables
Table 1. WISC-V full scale, index, and subtest level descriptive data.
were entered in two blocks: the first included the SF com-
N M SD Range (Min.–Max.)
bination of interest and the moderator variable; the second
FSIQ 268 85.34 15.91 43–128
included a new variable equal to the product of the SF com-
VCI 268 89.58 16.91 45–127 bination of interest and moderator variable of interest (i.e.,
Similarities 268 8.03 3.28 1–16
Vocabulary 268 8.21 3.34 1–16
Table 2. Correlations between WISC-V FSIQ and subtests.
VSI 158 90.41 15.07 57–129
Block design 268 8.17 3.29 1–18 SIM VOC BD VP MR FW DS PS CD SS
Visual puzzles 158 9.03 3.04 3–16 FSIQ .80 .75 .72 .66 .77 .77 .75 .63 .52 .44
FRI 267 90.43 16.27 55–147 Similarities (SIM) — .74 .47 .46 .47 .54 .58 .43 .25 .30
Matrix reasoning 268 7.85 3.10 1–17 Vocabulary (VOC) — .42 .54 .46 .47 .49 .50 .25 .27
Figure weights 268 8.77 3.16 2–19 Block Design (BD) — .53 .58 .50 .44 .38 .24 .18
WMI 157 84.67 15.28 51–130 Visual Puzzles (VP) — .57 .58 .40 .44 .24 .15^
Digit span 268 7.12 2.94 1–16 Matrix Reasoning (MR) — .64 .50 .47 .28 .31
Picture span 157 7.98 3.36 1–19 Figure Weights (FW) — .49 .54 .34 .28
PSI 161 85.48 15.23 45–123 Digit Span (DS) — .58 .36 .40
Coding 268 7.15 3.05 1–16 Picture Span (PS) — .32 .42
Symbol search 159 7.86 2.97 1–14 Coding (CD) — .47
FSIQ: full scale IQ; VCI: verbal comprehension index; VSI: visuospatial index; Symbol Search (SS) —
FRI: fluid reasoning index; WMI: working memory index; PSI: processing FSIQ: Full Scale IQ. Correlations were conducted with pairwise exclusion, such
speed index. Subtests contributing to FSIQ calculation. FSIQ and Index that the n for each bivariate correlation is equivalent to the lower n for
scores are expressed as Standard Scores (M ¼ 100, SD ¼ 15) and subtest either variable indicated in Table 2. All correlations are significant at p < .05
scores are expressed as scaled scores (M ¼ 10, SD ¼ 3). except for that marked by ^.
 APPLIED NEUROPSYCHOLOGY: CHILD 5

an interaction term). Moderation occurs “if (and only if)

Cohen’s d

These formulas are derived from individual linear regression analyses. x in each formula equals the sum of scaled scores for each short form combination and the solution is expressed in Standard Score units (i.e.,
M ¼ 100, SD ¼ 15). bThis difference is the average difference expressed in Standard Score units between true FSIQ and prorated estimated FSIQ. Positive values in this column indicate that the prorated estimated FSIQ
underestimated true FSIQ. SIM: Similarities scaled score; BD: Block Design scaled score; MR: Matrix Reasoning scaled score; DS: Digit Span scaled score; CD: Coding scaled score; VOC: Vocabulary Scaled Score; FW:
.46
.12
.39
.04
.37

.15
.50

.29
.41
.10
this [interaction] term is significant” (Dawson, 2014, p. 2).
Given that 30 total hierarchical regression analyses were per-
formed to explore possible moderating effects of these three
variables for each of the 10 SFs, significances for each inter-
action term were compared to a conservative critical p value
of .01, which was chosen to minimize Type 1 error.
t (Bonferroni-adjusted p)

Regarding 7- vs. 10-subtest WISC-V administration, none

7.45 (< .001)

6.35 (< .001)

6.02 (< .001)

8.25 (< .001)

4.83 (< .001)

6.83 (< .001) of the interaction terms in the 10 moderation analyses
1.87 (.126)

.61 (.542)

2.43 (.032)

1.60 (.111) emerged as statistically significant (ps ranged from .024 to

.830); only one interaction term in the moderation analysis
for VBFDC trended toward significance (p ¼ .024), though
the interaction term explained no additional variance within
the model and moderation was not identified. Similarly,
regarding age, none of the interaction terms in the 10 mod-
eration analyses emerged as significant (ps ranged from .018
(i.e., True FSIQ—Prorated Estimated FSIQ)

to .678); only one interaction term in the moderation ana-

lysis for SBDC trended toward significance, though
explained negligible additional variance (.002%) to the
regression model, and moderation was not identified.
Mean differenceb

Finally, sex did not emerge as a moderator for the relation-

1.62
.41
1.40
.12
1.88
2.41
.79
1.49
2.04
.50

ship between SF and true FSIQ in any moderation analysis

(ps ranged from .060 to .964). In sum, neither 7- vs 10-subt-
est WISC-V administration, participant age, nor participant
gender appeared to moderate the relationship between SF
combination and true FSIQ. To preserve journal space, pres-
entation of these null moderation analyses and specific
regression results are not presented herein, though may be
made available upon request from the first author.1
Formulas to calculate Regression-based Estimated FSIQa

Comparing estimated FSIQs to true FSIQ

Regarding regression-based estimates, paired-samples t-tests
revealed no significant differences between true FSIQ and
estimated FSIQ for all SFs (ps > .99). Regarding prorated
FSIQ estimates for pentad SFs, paired-samples t-tests (dis-
33.476 þ 1.354(x)
32.640 þ 1.343(x)
32.796 þ 1.365(x)
31.442 þ 1.367(x)
35.747 þ 1.628(x)

35.705 þ 1.598(x)
35.749 þ 1.645(x)

34.792 þ 1.649(x)
35.279 þ 1.651(x)
34.528 þ 1.626(x)

played in Table 3) revealed significant differences between

true FSIQ and prorated FSIQ for SBMDC and VBMDC (ps
< .001; Cohen’s ds ranged from .39 to .46). Regarding tetrad
Table 3. Paired-samples t-tests between true FSIQ and prorated estimates.

SFs, paired-samples t-tests revealed significant differences

between SBDC, SMDC, SFDC, VBDC, and VMDC (ps <
.001; Cohen’s ds ranged from .15 to .50). Of note, Holm-
Bonferroni stepdown procedure was used to correct p values
for the set of pentad SFs and tetrad SFs separately. Analysis
of mean difference scores indicated that estimated FSIQ for
these six prorated SFs appeared to underestimate true FSIQ
by 0.79  2.41 Standard Score points, on average. As such,
to adjust for these underestimations, prorated and arithmet-
ically adjusted estimated FSIQs were additionally calculated
VBMDC (VOC þ BD þ MR þ DS þ CD)
SBMDC (SIM þ BD þ MR þ DS þ CD)

VBFDC (VOC þ BD þ FW þ DS þ CD)

SBFDC (SIM þ BD þ FW þ DS þ CD)

for each of these SFs. That is, an integer was added the pro-
Figure Weights Scaled Score.

rated FSIQ derived from WISC-V manual (see Wechsler,

VMDC (VOC þ MR þ DS þ CD)
Estimated FSIQ combinations

SMDC (SIM þ MR þ DS þ CD)

VFDC (VOC þ FW þ DS þ CD)

VBDC (VOC þ BD þ DS þ CD)
SFDC (SIM þ FW þ DS þ CD)
SBDC (SIM þ BD þ DS þ CD)

2014, Table A.7) for SBMDC (þ2), VBMDC (þ1), SBDC

(þ2), SMDC (þ2), SFDC (þ1), VBDC (þ1), and VMDC

1
The authors thank an anonymous reviewer for their suggestion to perform
these moderation analyses and describe their results. The relationships
between SF combinations and true FSIQ did not appear to be moderated by
how many subtests were administered, participant age, nor
participant gender.
a
6 J. W. LACE ET AL.

(þ2). For example, consider an examinee whose sum of Second, the SFs described above provided various possibil-
scaled scores for SBMDC is 40, which converts to prorated ities for estimating FSIQ. Importantly, each SF combination
sum of scaled scores of 56. The unadjusted FSIQ derived represents a broad range of cognitive abilities, each of which
from the WISC-V manual is 85, and the prorated/adjusted are likely important to consider when estimating general cog-
estimated FSIQ is 87 (i.e., 85 þ 2). nitive ability (Wechsler, 2014). Regarding the five-subtest SFs,
A method for converting sums of scaled scores to pro- VBFDC (Vocabulary þ Block Design þ Figure Weights þ
rated sums of scaled scores in order to obtain estimated Digit Span þ Coding) appeared to provide the most consist-
FSIQ is provided in Table 4. Regarding accuracy of regres- ently accurate performance regardless of estimation method
sion-based, prorated, and prorated/adjusted FSIQ estimates, used (i.e., regression-based or prorated/adjusted), with
for each participant and SF, concordance rates between true approximately 85–92% of examinees’ FSIQ being estimated
and estimated FSIQ were calculated and these results are correctly within 5 Standard Score points. Similarly, SBMDC
displayed in Table 5. For pentads, most participants’ regres- (Similarities þ Block Design þ Matrix Reasoning þ Digit
sion-based estimated FSIQ (81.3  85.4%) fell within 5 Span þ Coding) estimated FSIQ provided good accuracy with
Standard Score points of true FSIQ and 100% fell within 81% to 88% of examinees’ estimates within 5 Standard Score
one standard deviation (i.e., 15 Standard Score points) for points. Regarding tetrad SFs, the best performing tetrads
each pentad. Prorated and prorated/adjusted estimated FSIQ appeared to be VMDC (Vocabulary þ Matrix Reasoning þ
appeared to provide somewhat improved accuracy, with Digit Span þ Coding), SMDC (Similarities þ Matrix
85.4–91.8% of estimated FSIQ falling within 5 Standard Reasoning þ Digit Span þ Coding), and VFDC
Score points. VBFDC and SBMDC (with a two-point adjust- (Vocabulary þ Figure Weights þ Digit Span þ Coding), with
ment to the latter) appeared to be the best perform- approximately 75% of examinees’ prorated or prorated/
ing pentads. adjusted FSIQ falling within 5 Standard Score points of true
Regarding tetrads, most participants’ regression-based FSIQ. These accuracies are consistent with values reported in
estimated FSIQ (64.9  70.1%) fell within 5 Standard Score other SF validation studies with similar aims (Hrabok et al.,
points of true FSIQ and 99.6–100% fell within one standard 2014; Hwang & Oh, 2017; Merz et al., 2019).
deviation. Similarly, prorated and prorated/adjusted esti- Of course, the cognitive domains represented within the
mated FSIQ appeared to be somewhat more accurate, with pentads (i.e., including one subtest from every WISC-V
69.4–76.1% of estimated FSIQ falling within 5 Standard index) are broader than those represented by tetrads.
Score points. VMDC, SMDC (with two-point adjustments to Additionally, the use of five subtests in pentads understand-
both), and VFDC appeared to be the best perform- ably accounts for more variance in FSIQ and yields
ing tetrads. improved accuracy. As such, it follows that the most accur-
ate FSIQ estimation can be obtained when considering per-
formance on more subtests, especially those across all five
Discussion
indices. Unfortunately, given that pentads require five subt-
Despite the widespread contemporary use of the newest iter- ests for its calculation, they provide the least abbreviated
ation of the Wechsler Intelligence Scale for Children (i.e., and most time-consuming option of those presented. In
WISC-V; Wechsler, 2014) in clinical practice (Rabin et al., contrast, the tetrads SFs, while providing more abbreviated
2016), limited research to date has explored clinically applic- options as they require the fewest subtests, yielded lower
able short form (SF) combinations for estimating full scale accuracies, perhaps (at least in part) because they represent
IQ (FSIQ). Only two previous articles have explored this a relatively narrower selection of available cognitive
topic (Aubry & Bourdin, 2018; Piovesana et al., 2019), domains. Of note, while these results suggested that the
though each pose methodological and practical limitations regression-based estimation methods may provide adequate/
including not utilizing mixed clinical samples. To address acceptable accuracy in the present sample, the prorated and
this gap in the literature, the present study identified several prorated/adjusted estimates generally appeared to provide
SF combinations for the WISC-V, each of which represented greater concordance with true FSIQ, and it may be that the
an intentionally broad range of cognitive abilities, within a latter method provides greater generalizability to
mixed clinical sample. Several findings deserve fur- other samples.
ther discussion. Importantly, as described at the outset of this paper, the
First, regarding the pattern of performance on WISC-V primary benefit of using abbreviated or shortened forms of
indices and subtests, mean scores appeared to be between tests is twofold. First, reducing time spend measuring a sin-
two-thirds and one standard deviation below the normative gle cognitive construct (i.e., general cognitive ability) may
mean within the low average-to-average ranges. Given the allow for increased clinical time assessing other neurocogni-
present study’s inclusion of examinees referred for clinical tive domains of interest. For example, the utilization of a
purposes, most of whom received neurodevelopmental diag- pentad to estimate FSIQ instead of the seven- or 10-subtest
noses, such mean scores are commensurate with those WISC-V could easily save well over 15 min of test adminis-
reported in other clinical samples with various versions of tration time by eliminating, for example, Similarities and
the WISC (Bremner et al., 2011; Mayes & Calhoun, 2006; Matrix Reasoning (Ryan et al., 2007) or those two subtests
Nader et al., 2015; Serenius et al., 2016; Toffalini plus Visual Puzzles, Picture Span, and Symbol Search.
et al., 2017). Tetrad combinations logically provide even greater clinical
Table 4. Prorated sums of scaled scores for pentads and tetrads.
Sum of pentad Prorated sum of Sum of pentad Prorated sum of Sum of tetrad Prorated sum of Sum of tetrad Prorated sum of
Scaled Scores scaled scoresa Scaled Scores scaled scoresa Scaled Scores scaled scoresa Scaled Scores scaled scoresa
5 7 51 71 4 7 50 88
6 8 52 73 5 9 51 89
7 10 53 74 6 11 52 91
8 11 54 76 7 12 53 93
9 13 55 77 8 14 54 95
10 14 56 78 9 16 55 96
11 15 57 80 10 18 56 98
12 17 58 81 11 19 57 100
13 18 59 83 12 21 58 102
14 20 60 84 13 23 59 103
15 21 61 85 14 25 60 105
16 22 62 87 15 26 61 107
17 24 63 88 16 28 62 109
18 25 64 90 17 30 63 110
19 27 65 91 18 32 64 112
20 28 66 92 19 33 65 114
21 29 67 94 20 35 66 116
22 31 68 95 21 37 67 117
23 32 69 97 22 39 68 119
24 34 70 98 23 40 69 121
25 35 71 99 24 42 70 123
26 36 72 101 25 44 71 124
27 38 73 102 26 46 72 126
28 39 74 104 27 47 73 128
29 41 75 105 28 49 74 130
30 42 76 106 29 51 75 131
31 43 77 108 30 53 76 133
32 45 78 109 31 54
33 46 79 111 32 56
34 48 80 112 33 58
35 49 81 113 34 60
36 50 82 115 35 61
37 52 83 116 36 63
38 53 84 118 37 65
39 55 85 119 38 67
40 56 86 120 39 68
41 57 87 122 40 70
42 59 88 123 41 72
43 60 89 125 42 74
44 62 90 126 43 75
45 63 91 127 44 77
46 64 92 129 45 79
47 66 93 130 46 81
48 67 94 132 47 82
49 69 95 133 48 84
50 70 49 86
a
These prorated sums of scaled scores was calculated by multiplying the sum of scaled scores for pentad SFs by 7/5 and for tetrad SFs by 7/4. The prorated sum of scaled scores may be converted to an FSIQ estimate
according to WISC-V manual Table A.7 (Wechsler, 2014) and adjusted as necessary according to Table 5.
APPLIED NEUROPSYCHOLOGY: CHILD
7
8 J. W. LACE ET AL.

Table 5. Percentage of regression-based and prorated/adjusted FSIQ estimates within 5, 10, and 15 standard score points of true FSIQ.
D±5 D ± 10 D ± 15 r True FSIQ
Regression-based estimates
SBMDCa 81.3 99.6 100 .97
SBFDC 83.6 99.6 100 .97
VBMDC 83.6 100 100 .97
VBFDC 85.4 99.3 100 .98
SBDC 64.9 94.0 100 .95
SMDC 68.3 97.8 100 .95
SFDC 64.9 94.4 99.6 .94
VBDC 64.9 96.3 99.6 .95
VMDC 70.1 95.9 100 .95
VFDC 68.7 95.9 99.6 .95
Prorated and prorated/adjusted estimates
SBMDC (þ2) 83.6 (88.1) 98.9 (99.6) 100 (100) .98
SBFDC 86.2 99.3 100 .97
VBMDC (þ1) 81.3 (85.4) 100 (100) 100 (100) .95
VBFDC 91.8 99.6 100 .98
SBDC (þ2) 70.1 (71.3) 95.5 (94.8) 99.6 (100) .95
SMDC (þ2) 69.4 (75.4) 96.3 (97.4) 99.6 (100) .95
SFDC (þ1) 71.3 (69.4) 94.4 (95.5) 99.3 (99.6) .94
VBDC (þ1) 67.2 (72.4) 97.4 (97.0) 99.6 (100) .95
VMDC (þ2) 68.3 (76.1) 94.8 (97.0) 100 (100) .95
VFDC 73.9 95.6 99.6 .95
N ¼ 268. For short forms followed by parentheses, concordance rates for the prorated/unadjusted estimated FSIQ are presented in plain
text and adjusted concordance rates corresponding to prorated/adjusted estimated FSIQ are presented in parentheses.
a
These acronyms are consistent with those displayed in Table 1.

time saving, though at a cost of possible estimation accur- such that these tetrads may be more applicable with examin-
acy. Nonetheless, such time savings could allow practitioners ees for whom manipulating blocks poses marked difficulties,
to focus more assessment time on individually relevant cog- including those with gross/fine motor or other coordination
nitive domains (e.g., memory, executive functioning, aca- deficits, or for whom suspected decrements in completion of
demic achievement, language) that are not properly assessed timed tasks may possibly markedly misidentify cognitive
by the WISC-V (Donders, 2001). Second, abbreviating test- ability. Additionally, SF’s that include Matrix Reasoning
ing time results in less stressful and cognitively draining over Figure Weights may reduce the linguistic demand on
experiences for examinees, especially those who are particu- examinees and be more culturally appropriate in certain
larly sensitive to fatigue (Emons et al., 2007). Minimizing cases (Cormier et al., 2016). While not an exhaustive discus-
the testing burden on such individuals may prevent reduced sion of comparisons between subtests included in each SF,
tolerance toward testing, reduce suspicions of decreased test these examples illustrate the possible benefits and detriments
engagement throughout the evaluation, and lessen the and required considerations when choosing to implement
chance of problem behaviors (e.g., temper tantrums) during one SF over another. Importantly, clinical judgment should
testing. The routine administration of an abbreviated SF for be used to determine which of these SF combinations, if
estimating WISC-V FSIQ may be a natural extension of any, is appropriate in each individual case, and the authors
other accommodated assessment procedures and appropriate refrain from making generalized recommendations in this
for clinical populations that are traditionally “hard-to-test” regard. Furthermore, given that none of the three variables
(Calhoun & Newson, 1984). examined as possible moderators (i.e., 7- vs. 10-subtest
Each SF described in the present study may provide dif- WISC-V administration, age, gender) emerged as such, it is
ferential pragmatic clinical utility given the subtests repre- possible and perhaps likely that each SF combination may
sented within each. As noted above, pentads may provide be applicable within many diverse clinical settings for both
the broadest and most representative estimate of FSIQ given boys and girls across the WISC-V’s age span.
they sample subtests across each of WISC-V’s domains. Of note, some may critique the need for identifying short
Furthermore, for example, VBFDC’s inclusion of Vocabulary forms in the WISC-V for two reasons. First, some may
may make it less influenced by marked deficits in verbally argue that the revision to FSIQ’s calculation in the WISC-V
mediated executive functioning or abstraction ability (in (i.e., using sum of scaled scores for seven subtests) negates
comparison to Similarities) which may be more negatively the need for identifying short forms. That is, according to
impacted by neurological insult (Ong et al., 1998). the WISC-V manual, FSIQ is now calculated from seven
Additionally, VBFDC’s inclusion of Figure Weights may subtests rather than 10 (as is needed in previous WISC iter-
provide utility over Matrix Reasoning given its timed nature ations and in the most recent adult-focused analogous meas-
(and possible additive time-saving properties) and relation- ures [i.e., WAIS-IV]), and it may be argued that the
ships with quantitative reasoning-related outcomes, includ- administration of seven subtests is sufficiently brief for most
ing academic achievement, creativity, and standardized test clinical settings. However, the authors believe that identify-
performance (Wechsler, 2014). Regarding tetrads (which ing subtest combinations that offer even more abbreviated
merged Visual Spatial and Fluid Reasoning Indices), VMDC estimates of FSIQ is in line with remote and recent research
and SMDC included Matrix Reasoning over Block Design, (e.g., Aubry & Bourdin, 2018; Clements, 1965; Piovesana
 APPLIED NEUROPSYCHOLOGY: CHILD 9

et al., 2019; Ryan, 1981) and appropriate to continue to do proportions of examinees of racial/ethnic minorities.
so with this measure. Moreover, as stated above, the ability Additionally, nearly one-third of participants had a primary
to estimate FSIQ using an abbreviated battery is clinically diagnosis of autism spectrum disorder, which may limit the
relevant, particularly given the opportunity to reduce testing generalizability to samples with higher proportions of
burden on patients for whom fatigue or behavioral distur- acquired neurological injury (e.g., traumatic brain injuries,
bances are paramount and to focus clinical assessment time strokes, tumors). Relatedly, high levels of diagnostic comor-
on other cognitive domains more important to the present- bidity were noted, with nearly half of participants receiving
ing concerns of each examinee (e.g., memory, executive three or more diagnoses. Nonetheless, the authors purport
functioning, academic achievement, language). As such, the that this is characteristic of a mixed clinical sample and, in
practice of describing appropriate SFs is in line with recom- fact, represents likely ecological validity of these findings.
mendations to use abbreviated tests as part of routine clin- Regardless, future research should attempt to recruit samples
ical assessment whenever possible (Donders, 2001) and to that are representative of the demographic and diagnostic
maximize brevity more broadly within the pediatric neuro- diversity (or homogeneity, depending on future research
psychological assessment process (Donders, 1999). goals) seen across geographic and clinical settings.
Additionally, some may suggest that identifying short Third, variability in performance validity test use was
forms in the WISC-V is redundant or unnecessary given the noted in the archival dataset, such that not every examinee
existence of instruments such as the Wechsler Abbreviated was administered a standalone measure of test engagement;
Intelligence Scale, Second Edition (WASI-II; Wechsler, 2011). this observation was consistent with noted trends in con-
However, while the end goal of the WASI-II and the WISC-V temporary clinical practice (DeRight & Carone, 2015;
are similar (i.e., to quantify general cognitive ability), the MacAllister et al., 2019). However, while it is possible that
methods employed by each test are unique. That is, the WASI- some examinees included in the final dataset demonstrated
II, while streamlined and efficient, may assess a psychometric- noncredible performance, this statement cannot be made
ally distinct construct from WISC-V FSIQ, and these variables definitively. The authors chose not to consider reliable digit
cannot be equated (McCrimmon & Smith, 2013). For span or other digit span-derived variables when determining
example, the WASI-II’s FSIQ calculation represents only subt- credibility of scores given the potential for high false posi-
ests purportedly tapping dimensions of verbal comprehension tives (e.g., Blaskewitz et al., 2008; Welsh et al., 2012) and to
(i.e., Vocabulary and Similarities) and perceptual reasoning avoid any statistical ramifications resulting from possible
(i.e., Block Design and Matrix Reasoning) while altogether restriction of range in Digit Span subtest scaled scores.
omitting tests of working memory (e.g., digit or picture span Future work may consider prospectively including well-vali-
paradigms) and processing speed (e.g., digit-symbol coding or dated measures of performance validity to ensure examinees
cancelation tests); the construct assessed by the WASI-II may are adequately screened for test engagement excluded
be more akin to WISC-V General Ability Index. In contrast, as needed.
the WISC-V’s FSIQ calculation includes subtests across the Fourth, from a statistical perspective, correlations
breadth of these cognitive domains (regardless of the WISC- reported in Tables 2 and 5, while not the primary focus of
V’s disputed four- or five-factor structure; Canivez et al., 2016, the present study, were uncorrected for shared error vari-
2018; Dombrowski et al., 2018; Lecerf & Canivez, 2018). ance and may represent an overestimate of the true relation-
Importantly, aspects of working memory and processing speed ships between subtest-FSIQ and estimated FSIQ-true FSIQ
are particularly fragile in the context of neurodevelopmental, bivariate pairs. Future research that primarily focuses on
neurological, and psychiatric insult in children (e.g., Allen correlational results may choose to incorporate more sophis-
et al., 2010; Calhoun & Mayes, 2005; De Clercq-Quaegebeur ticated statistical corrections (e.g., Girard & Christensen,
et al., 2010; San Miguel Montes et al., 2010) and may be 2008; Levy, 1967). Relatedly, the present study’s focus on 10
important to consider when describing intellectual ability in distinct pentad and tetrad combinations may limit other
children presenting for neuropsychological assessment. Thus, combinations of other subtests which do not regularly con-
the present study’s emphasis on SF combinations that tribute to FSIQ or combinations with fewer than four subt-
included subtests across many cognitive domains representa- ests. Future studies may choose to specifically consider
tive of FSIQ is noteworthy. otherwise predetermined SF combinations of interest to the
Various limitations deserve mentioning. First, the utiliza- researchers or adopt other statistical methodologies. Finally,
tion of a mixed clinical sample may limit generalizability to future researchers may choose to investigate diagnosis as a
other populations of interest, including those of specialized potential moderating variable; given the archival nature of
or homogenous clinical presentations or those with high these data and the high rates of diagnostic comorbidity rep-
rates of examinees presumably at the upper end of IQ distri- resented in the present sample, this line of inquiry, unfortu-
butions. Future work may seek to replicate and extend these nately, could not be meaningfully pursued.
findings and SF combinations into other samples of interest.
Second, and relatedly, given these data were extracted from
Conclusion
a clinic in the rural Midwestern United States and that an
overwhelming majority of participants identified as White/ In conclusion, the present study is among the first to pre-
Caucasian, these results may not adequately generalize to sent short form combinations for estimating full scale IQ in
nationally representative samples or those with higher the WISC-V specifically within a mixed clinical sample.
10 J. W. LACE ET AL.

Regression-based estimated FSIQ appeared to offer accept- and secondary subtests. Psychology in the Schools, 55(7), 741–769.
able accuracy, although prorated/adjusted estimated FSIQ https://doi.org/10.1002/pits.22138
Canivez, G. L., Watkins, M. W., & Dombrowski, S. C. (2016). Factor
scores appeared somewhat better, as the latter tended to
structure of the Wechsler intelligence scale for children-fifth edition:
demonstrate good overall accuracy (generally >70% within Exploratory factor analyses with the 16 primary and secondary subt-
5 Standard Score points of true FSIQ) using short forms ests. Psychological Assessment, 28(8), 975–986. https://doi.org/10.
with five and four subtests. Each described short form cer- 1037/pas0000238
tainly presents its own benefits, detriments, and other con- Canivez, G. L., Watkins, M. W., & Dombrowski, S. C. (2017).
siderations. Clinicians may seek to integrate these findings Structural validity of the Wechsler intelligence scale for children-
fifth edition: Confirmatory factor analyses with the 16 primary and
with their clinical practice, and researchers may utilize these secondary subtests. Psychological Assessment, 29(4), 458–472.
short form combinations to quickly and adequately estimate https://doi.org/10.1037/pas0000358
cognitive ability in research samples. Future work should Carleton, F. O., & Stacey, C. L. (1954). Evaluation of selected short forms
validate and critique these and other WISC-V short form of the Wechsler Intelligence Scale for Children (WISC). Journal of
combinations in demographically and diagnostically Clinical Psychology, 10(3), 258–261. https://doi.org/10.1002/1097-
4679(195407)10:3<258::AID-JCLP2270100314>3.0.CO;2-U
diverse samples.
Chen, H., Zhang, O., Raiford, S. E., Zhu, J., & Weiss, L. G. (2015).
Factor invariance between genders on the Wechsler intelligence scale
for children–fifth edition. Personality and Individual Differences, 86,
References 1–5. https://doi.org/10.1016/j.paid.2015.05.020
Adaki, K. (2009). The validity of the WASI in predicting former WISC- Clements, G. R. (1965). An abbreviated form of the Wechsler
IV scores for re-evaluation purposes in a sample of special education Intelligence Scale for Children. Journal of Consulting Psychology,
students [Doctoral dissertation]. Farleigh Dickinson University. 29(1), 92. https://doi.org/10.1037/h0020970
Allen, D. N., Thaler, N. S., Donohue, B., & Mayfield, J. (2010). WISC- Comninel, M. E., & Bordieri, J. E. (2001). Estimating WISC-III scores
IV profiles in children with traumatic brain injury: Similarities to for special education students using the Dumont-Faro short form.
and differences from the WISC-III. Psychological Assessment, 22(1), Psychology in the Schools, 38(1), 11–16. https://doi.org/10.1002/
57–64. https://doi.org/10.1037/a0016056 1520-6807(200101)38:1 < 11::aid-pits2 > 3.0.co;2-p
Ardila, A. (1999). A neuropsychological approach to intelligence. Connery, S., Katz, D., Kaufman, A. S., & Kaufman, N. L. (1996).
Neuropsychology Review, 9(3), 117–136. https://doi.org/10.1023/ Correlations between two short cognitive tests and a WISC-III short
a:1021674303922 form using a sample of adolescent inpatients. Psychological Reports,
Aubry, A., & Bourdin, B. (2018). Short forms of Wechsler Scales 78(3_suppl), 1373–1378. https://doi.org/10.2466/pr0.1996.78.3c.1373
assessing the intellectually gifted children using simulation data. Cormier, D. C., Kennedy, K. E., & Aquilina, A. M. (2016). Test review:
Frontiers in Psychology, 9, 830. https://doi.org/10.3389/fpsyg.2018. Wechsler intelligence scale for children, fifth edition: Canadian 322
00830 (WISC-VCDN) by D. Wechsler. Canadian Journal of School
Axelrod, B. N. (2002). Validity of the Wechsler abbreviated scale of Psychology, 31(4), 322–334. https://doi.org/10.1177/08295735166
intelligence and other very short forms of estimating intellectual 48941
functioning. Assessment, 9(1), 17–23. https://doi.org/10.1177/ Cormier, D. C., Wang, K., & Kennedy, K. E. (2016). Linguistic
1073191102009001003 demands of the oral directions for administering the WISC-IV and
Beal, A. L., Dumont, R., Cruse, C. L., & Branche, A.-E. H. (1996). WISC-V. Canadian Journal of School Psychology, 31(4), 290–304.
Practical implications of differences between the American and https://doi.org/10.1177/0829573516643314
Canadian norms for WISC-III and a short form for children with Crawford, J. R., Anderson, V., Rankin, P. M., & MacDonald, J. (2010).
learning disabilities. Canadian Journal of School Psychology, 12(1), An index-based short-form of the WISC-IV with accompanying
7–14. https://doi.org/10.1177/082957359601200103 analysis of the reliability and abnormality of differences. British
Bergman, L. R., Corovic, J., Ferrer-Wreder, L., & Modig, K. (2014). Journal of Clinical Psychology, 49(Pt 2), 235–258. https://doi.org/10.
High IQ in early adolescence and career success in adulthood: find- 1348/014466509X455470
ings from a Swedish Longitudinal Study. Research in Human Dasi, C., Soler, M. J., Bellver, V., & Ruiz, J. C. (2014). Short form of
Development, 11(3), 165–185. https://doi.org/10.1080/15427609. Spanish version of the WISC–IV for intelligence assessment in elem-
2014.936261 entary school children. Psychological Reports, 115(3), 784–793.
Blaskewitz, N., Merten, T., & Kathmann, N. (2008). Performance of https://doi.org/10.2466/03.PR0.115c32z7
children on symptom validity tests: TOMM, MSVT, and FIT. Dawson, J. F. (2014). Moderation in management research: What, why,
Archives of Clinical Neuropsychology, 23(4), 379–391. https://doi. when, and how. Journal of Business and Psychology, 29(1), 1–19.
org/10.1016/j.acn.2008.01.008 https://doi.org/10.1007/s10869-013-9308-7
Bremner, D., McTaggart, B., Saklofske, D. H., & Janzen, T. (2011). De Clercq-Quaegebeur, M., Casalis, S., Lemaitre, M.-P., Bourgois, B.,
WISC-IV GAI and CPI in psychoeducational assessment. Canadian Getto, M., & Vall
ee, L. (2010). Neuropsychological Profile on the
Journal of School Psychology, 26(3), 209–219. https://doi.org/10. WISC-IV of French children with dyslexia. Journal of Learning
1177/0829573511419090 Disabilities, 43(6), 563–574. https://doi.org/10.1177/00222194103
Calhoun, S. L., & Mayes, S. D. (2005). Processing speed in children 75000
with clinical disorders. Psychology in the Schools, 42(4), 333–343. DeRight, J., & Carone, D. A. (2015). Assessment of effort in children:
https://doi.org/10.1002/pits.20067 A systematic review. Child Neuropsychology, 21(1), 1–24. https://
Calhoun, M. L., & Newson, E. (1984). Parents as experts: An assess- doi.org/10.1080/09297049.2013.864383
ment approach for hard-to-test children. Diagnostique, 9(4), Dirks, J., Wessels, K., Quarforth, J., & Quenon, B. (1980). Can short-
239–244. https://doi.org/10.1177/073724778400900406 form WISC-R 1Q.
Campbell, J. M. (1998). Internal and external validity of seven Dombrowski, S. C., Canivez, G. L., & Watkins, M. W. (2018). Factor
Wechsler Intelligence Scale for Children–Third Edition short forms structure of the 10 WISC-V primary subtests across four standard-
in a sample of psychiatric inpatients. Psychological Assessment, 10(4), ization age groups. Contemporary School Psychology, 22(1), 90–104.
431–434. https://doi.org/10.1037/1040-3590.10.4.431 https://doi.org/10.1007/s40688-017-0125-2
Canivez, G. L., Dombrowski, S. C., & Watkins, M. W. (2018). Factor Donders, J. (1997). A short form of the WISC–III for clinical use.
structure of the WISC-V in four standardization age groups: Psychological Assessment, 9(1), 15–20. https://doi.org/10.1037/1040-
Exploratory and hierarchical factor analyses with the 16 primary 3590.9.1.15
 APPLIED NEUROPSYCHOLOGY: CHILD 11

Donders, J. (1999). Pediatric neuropsychological reports: Do they really Kaufman, A. S. (1976). A four-test short form of the WISC-R.
have to be so long? Child Neuropsychology, 5(1), 70–78. https://doi. Contemporary Educational Psychology, 1(2), 180–196. https://doi.
org/10.1076/chin.5.1.70.7071 org/10.1016/0361-476X(76)90022-9
Donders, J. (2001). Using a short form of the WISC-III: Sinful or Kaufman, A. S., Flanagan, D. P., Alfonso, V. C., & Mascolo, J. T.
smart? Child Neuropsychol, 7(2), 99–103. https://doi.org/10.1076/ (2006). Test review: Wechsler intelligence scale for children, fourth
chin.7.2.99.3126 edition (WISC-IV). Journal of Psychoeducational Assessment, 24(3),
Donders, J., Elzinga, B., Kuipers, D., Helder, E., & Crawford, J. R. 278–295. https://doi.org/10.1177/0734282906288389
(2013). Development of an eight-subtest short form of the WISC-IV Kaufman, A. S., Kaufman, J. C., Balgopal, R., & McLean, J. E. (1996).
and evaluation of its clinical utility in children with traumatic brain Comparison of three WISC-III short forms: Weighing psychometric,
injury. Child Neuropsychol, 19(6), 662–670. https://doi.org/10.1080/ clinical, and practical factors. Journal of Clinical Child Psychology,
09297049.2012.723681 25(1), 97–105. https://doi.org/10.1207/s15374424jccp2501_11
Donders, J., & Warschausky, S. (1996). Validity of a short form of the Kaufman, A. S., Raiford, S. E., & Coalson, D. L. (2015). Intelligent test-
WISC-III in children with traumatic head injury. Child ing with the WISC-V. John Wiley & Sons.
Neuropsychology, 2(3), 227–232. https://doi.org/10.1080/09297049 Keith, T. Z. (1994). Intelligence is important. School Psychology
608402255 Quarterly, 9(3), 209–221. https://doi.org/10.1037/h0088286
Doyle, L. R., & Mattson, S. N. (2015). Neurobehavioral disorder associ- Lecerf, T., & Canivez, G. L. (2018). “Complementary exploratory and
ated with prenatal alcohol exposure (ND-PAE): Review of evidence confirmatory factor analyses of the French WISC-V: Analyses based
and guidelines for assessment. Current Developmental Disorders on the standardization sample”: Correction to Lecerf and Canivez
Reports, 2(3), 175–186. https://doi.org/10.1007/s40474-015-0054-6 (2018). Psychological Assessment, 30(8), 1009. https://doi.org/10.
Dumont, R., & Faro, C. (1993). A WISC-III short form for learning- 1037/pas0000638
disabled students. Psychology in the Schools, 30(3), 212–219. https:// Levy, P. (1967). The correction for spurious correlation in the evalu-
doi.org/10.1002/1520-6807(199307)30:3<212::aid-pits2310300303>3. ation of short-form tests. Journal of Clinical Psychology, 23(1),
0.co;2-v 84–86. https://doi.org/10.1002/1097-4679(196701)23:1 < 84::AID-
Egeland, J., Lovstad, M., Norup, A., Nybo, T., Persson, B. A., Rivera, JCLP2270230123 > 3.0.CO;2-2
D. F., … Arango-Lasprilla, J. C. (2016). Following international Lezak, M. D. (1988). IQ: R.I.P. Journal of Clinical and Experimental
trends while subject to past traditions: neuropsychological test use Neuropsychology, 10(3), 351–361. https://doi.org/10.1080/
in the Nordic countries. The Clinical Neuropsychologist, 30(sup1), 01688638808400871
Lung, F. W., Chen, P. F., & Shu, B. C. (2018). Performance of
1479–1500. https://doi.org/10.1080/13854046.2016.1237675
Wisconsin card sorting test in five-year-old children in Taiwan:
Emons, W. H., Sijtsma, K., & Meijer, R. R. (2007). On the consistency
Relationship to intelligence and cognitive development. PLOS One.,
of individual classification using short scales. Psychological Methods,
13(8), e0202099. https://doi.org/10.1371/journal.pone.0202099
12(1), 105–120. https://doi.org/10.1037/1082-989x.12.1.105
MacAllister, W. S., Vasserman, M., & Armstrong, K. (2019). Are we
Field, A. (2013). Discovering statistics using IBM SPSS statistics. Sage.
documenting performance validity testing in pediatric neuropsycho-
Firkowska-Mankiewicz, A. (2011). Adult Careers: Does childhood IQ
logical assessments? A brief report. The Journal of Pediatric
predict later life outcome? Journal of Policy and Practice in
Neuropsychology, 25(8), 1035–1042. https://doi.org/10.1080/
Intellectual Disabilities, 8(1), 1–9. https://doi.org/10.1111/j.1741-
09297049.2019.1569606
1130.2011.00281.x
Mark, R., Beal, A. L., & Dumont, R. (1998). Validation of a WISC-III
Flanagan, D. P., & Alfonso, V. C. (2017). Essentials of WISC-V assess-
short-form for the identification of Canadian gifted students.
ment. John Wiley & Sons.
Canadian Journal of School Psychology, 14(1), 1–10. https://doi.org/
Flanagan, D. P., & McDonough, E. M. (Eds.). (2018). Contemporary
10.1177/082957359801400101
intellectual assessment: Theories, tests, and issues (4th ed). The
Mayes, S. D., & Calhoun, S. L. (2006). WISC-IV and WISC-III profiles
Guilford Press. in children with ADHD. Journal of Attention Disorders, 9(3),
Girard, T. A., & Christensen, B. K. (2008). Clarifying problems and
486–493. https://doi.org/10.1177/1087054705283616
offering solutions for correlated error when assessing the validity of Mayes, S. D., & Calhoun, S. L. (2007). Wechsler Intelligence Scale for
selected-subtest short forms. Psychological Assessment, 20(1), 76–80. Children-Third and -Fourth Edition predictors of academic achieve-
https://doi.org/10.1037/1040-3590.20.1.76 ment in children with attention-deficit/hyperactivity disorder. School
Glass, L. A., Ryan, J. J., Bartels, J. M., & Morris, J. (2008). Estimating Psychology Quarterly, 22(2), 234–249. https://doi.org/10.1037/1045-
WISC-IV indexes: proration versus linear scaling. Journal of Clinical 3830.22.2.234
Psychology, 64(10), 1175–1180. https://doi.org/10.1002/jclp.20518 Mayes, S. D., Calhoun, S. L., Bixler, E. O., & Zimmerman, D. N.
Goh, D. S. (1980). Note on selection of WISC-R short forms for differ- (2009). IQ and neuropsychological predictors of academic achieve-
ent uses. Journal of Clinical Psychology, 36(1), 319–321. https://doi. ment. Learning and Individual Differences, 19(2), 238–241. https://
org/10.1002/1097-4679(198001)36:1 < 319::aid-jclp2270360144 > 3.0. doi.org/10.1016/j.lindif.2008.09.001
co;2-z McCrimmon, A. W., & Smith, A. D. (2013). Review of the Wechsler
Green, C. M., Kirk, J. W., Connery, A. K., Baker, D. A., & Kirkwood, abbreviated scale of intelligence, second edition (WASI-II). Journal
M. W. (2014). The use of the Rey 15-item test and recognition trial of Psychoeducational Assessment, 31(3), 337–341. https://doi.org/10.
to evaluate noncredible effort after pediatric mild traumatic brain 1177/0734282912467756
injury. Journal of Clinical and Experimental Neuropsychology, 36(3), McKenzie, K., Murray, A. L., Murray, K. R., & Murray, G. C. (2014).
261–267. https://doi.org/10.1080/13803395.2013.879096 Assessing the accuracy of the WISC-IV seven-subtest short form
Herrera-Graf, M., Dipert, Z. J., & Hinton, R. N. (1996). Exploring the and the child and adolescent intellectual disability screening ques-
effective use of the vocabulary/block design short form with a spe- tionnaire in identifying intellectual disability in children. Child
cial school population. Educational and Psychological Measurement, Neuropsychology, 20(3), 372–377. https://doi.org/10.1080/09297049.
56(3), 522–528. https://doi.org/10.1177/0013164496056003014 2013.799642
Hrabok, M., Brooks, B. L., Fay-McClymont, T. B., & Sherman, E. M. Merz, Z. C., Lace, J. W., Eisenstein, A. M., & Grant, A. F. (2019).
(2014). Wechsler Intelligence Scale for Children-fourth edition Examination of Wechsler adult Intelligence Scale-Fourth Edition
(WISC-IV) short-form validity: A comparison study in pediatric epi- short-form IQ estimations in an outpatient psychoeducational sam-
lepsy. Child Neuropsychol, 20(1), 49–59. https://doi.org/10.1080/ ple. Applied Neuropsychology Adult. Advance online publication.
09297049.2012.741225 https://doi.org/10.1080/23279095.2019.1687480
Hwang, K., & Oh, S. W. (2017). Validity of K-WISC-IV short forms. Murray, A. L., McKenzie, K., & Murray, G. C. (2016). An evaluation of
Korean Journal of Clinical Psychology, 36(3), 381–390. https://doi. the performance of the WISC-IV eight-subtest short form with chil-
org/10.15842/kjcp.2017.36.3.008 dren who may have an intellectual disability. Journal of Intellectual
12 J. W. LACE ET AL.

and Developmental Disability, 41(1), 50–53. https://doi.org/10.3109/ Silverstein, A. B. (1970). Reappraisal of the validity of WAIS, WISC,
13668250.2015.1084611 and WPPSI short forms. Journal of Consulting and Clinical
Na, S. D., & Burns, T. G. (2016). Wechsler intelligence scale for chil- Psychology, 34(1), 12–14. https://doi.org/10.1037/h0028680
dren-V: Test review. Applied Neuropsychology: Child, 5(2), 156–160. Soldner, J. L., Upton, T. D., & Dunlap, P. N. (2009). Acquired brain
https://doi.org/10.1080/21622965.2015.1015337 injury rehabilitation and managed care: Implications for rehabilita-
Nader, A.-M., Jelenic, P., & Soulieres, I. (2015). Discrepancy between tion administrators. Journal of Rehabilitation Administrators, 33(1),
WISC-III and WISC-IV cognitive profile in autism spectrum: What 45–57.
does it reveal about autistic cognition? PlOs One, 10(12), e0144645. Stuss, D. T., Meiran, N., Guzman, D. A., Lafleche, G., & Willmer, J.
https://doi.org/10.1371/journal.pone.0144645 (1996). Do long tests yield a more accurate diagnosis of dementia
Ong, L., Chandran, V., Zasmani, S., & Lye, M. (1998). Outcome of than short tests?: A comparison of 5 neuropsychological tests.
closed head injury in Malaysian children: Neurocognitive and Archives of Neurology, 53(10), 1033–1039. https://doi.org/10.1001/
behavioural sequelae. Journal of Paediatrics and Child Health, 34(4), archneur.1996.00550100119021
363–368. https://doi.org/10.1046/j.1440-1754.1998.00239.x Thompson, A. P., LoBello, S. G., Atkinson, L., Chisholm, V., & Ryan,
Ortiz, V. Z., & Gonzalez, A. (1989). Validation of a short form of the J. J. (2004). Brief intelligence testing in Australia, Canada, the
WISC-R with accelerated and gifted hispanic students 1. Gifted United Kingdom, and the United States. Professional Psychology:
Child Quarterly, 33(4), 152–155. https://doi.org/10.1177/ Research and Practice, 35(3), 286–290. https://doi.org/10.1037/0735-
001698628903300405 7028.35.3.286
Pfeiffer, S. I., Reddy, L. A., Kletzel, J. E., Schmelzer, E. R., & Boyer, Toffalini, E., Giofre, D., & Cornoldi, C. (2017). Strengths and weak-
L. M. (2000). The practitioner’s view of IQ testing and profile ana- nesses in the intellectual profile of different subtypes of specific
lysis. School Psychology Quarterly, 15(4), 376–385. https://doi.org/ learning disorder: A study on 1,049 diagnosed children. Clinical
10.1037/h0088795 Psychological Science, 5(2), 402–409. https://doi.org/10.1177/
Piovesana, A. M., Harrison, J. L., & Ducat, J. J. (2019). The develop- 2167702616672038
ment of a motor-free short-form of the Wechsler Intelligence Scale van Ool, J. S., Hurks, P. P. M., Snoeijen-Schouwenaars, F. M., Tan,
for Children–Fifth Edition. Assessment, 26(8), 1564–1572. https:// I. Y., Schelhaas, H. J., Klinkenberg, S., Aldenkamp, A. P., &
doi.org/10.1177/1073191117748741 Hendriksen, J. G. M. (2018). Accuracy of WISC-III and WAIS-IV
Rabin, L. A., Paolillo, E., & Barr, W. B. (2016). Stability in test-usage short forms in patients with neurological disorders. Developmental
practices of clinical neuropsychologists in the United States and Neurorehabilitation, 21(2), 101–107. https://doi.org/10.1080/
17518423.2016.1277799
Canada over a 10-year period: a follow-up survey of INS and NAN
Ventura, L. M., DeDios-Stern, S., Oh, A., & Soble, J. R. (2019). They’re
Members. Archives of Clinical Neuropsychology, 31(3), 206–230.
not just little adults: The utility of adult performance validity meas-
https://doi.org/10.1093/arclin/acw007
ures in a mixed clinical pediatric sample. Applied Neuropsychology:
Rees, L. M., Tombaugh, T. N., Gansler, D. A., & Moczynski, N. P.
Child. Advance online publication. https://doi.org/10.1080/21622965.
(1998). Five validation experiments of the Test of Memory
2019.1685522
Malingering (TOMM). Psychological Assessment, 10(1), 10–20.
Watkins, M. W., Dombrowski, S. C., & Canivez, G. L. (2018).
https://doi.org/10.1037/1040-3590.10.1.10
Reliability and factorial validity of the Canadian Wechsler intelli-
Ricker, J. H. (1998). Traumatic brain injury rehabilitation: is it worth
gence scale for children–fifth edition. International Journal of School
the cost? Applied Neuropsychology, 5(4), 184–193. https://doi.org/ & Educational Psychology, 6(4), 252–265. https://doi.org/10.1080/
10.1207/s15324826an0504_3 21683603.2017.1342580
Ryan, J. J. (1981). Clinical utility of a WISC-R short form. Journal Wechsler, D. (2011). Wechsler abbreviated scale of intelligence (2nd ed).
of Clinical Psychology, 37(2), 389–391. https://doi.org/10.1002/1097- Pearson.
4679(198104)37:2<389::aid-jclp2270370229>3.0.co;2-# Wechsler, D. (2014). WISC-V: Technical and interpretive manual.
Ryan, J. J., Glass, L. A., & Brown, C. N. (2007). Administration time Pearson.
estimates for Wechsler Intelligence Scale for Children-IV subtests, Weiss, L. G., Saklofske, D. H., Holdnack, J. A., & Prifitera, A. (2015).
composites, and short forms. Journal of Clinical Psychology, 63(4), WISC-V assessment and interpretation: Scientist-practitioner perspec-
309–318. https://doi.org/10.1002/jclp.20343 tives. Elsevier.
San Miguel Montes, L. E., Allen, D. N., Puente, A. E., & Neblina, C. Welsh, A. J., Bender, H. A., Whitman, L. A., Vasserman, M., &
(2010). Validity of the WISC–IV Spanish for a clinically referred MacAllister, W. S. (2012). Clinical utility of reliable digit span in
sample of Hispanic children. Psychological Assessment, 22(2), assessing effort in children and adolescents with epilepsy. Archives
465–469. https://doi.org/10.1037/a0018895 of Clinical Neuropsychology, 27(7), 735–741. https://doi.org/10.1093/
Seidenberg, M., Giordani, B., Berent, S., & Boll, T. J. (1983). IQ level arclin/acs063
and performance on the Halstead-Reitan Neuropsychological Test Wheaton, P. J., & Vandergriff, A. F. (1987). Short Forms of the
Battery for Older Children. Journal of Consulting and Clinical Mexican version of the WISC—R in the identification of gifted stu-
Psychology, 51(3), 406–413. https://doi.org/10.1037/0022-006X.51.3. dents. Psychological Reports, 61(1), 259–264. https://doi.org/10.2466/
406 pr0.1987.61.1.259
Serenius, F., Ewald, U., Farooqi, A., Fellman, V., Hafstr€ om, M., Yalowitz, J. M., & Armstrong, R. G. (1955). Validity of short forms of
Hellgren, K., Mars
al, K., Ohlin, A., Olhager, E., Stjernqvist, K., the Wechsler Intelligence Scale for Children (WISC). Journal of
Str€
omberg, B., Åd
en, U., & K€all
en, K., for the Extremely Preterm Clinical Psychology, 11(3), 275–277. https://doi.org/10.1002/1097-
Infants in Sweden Study Group. (2016). Neurodevelopmental out- 4679(195507)11:3<275::AID-JCLP2270110306>3.0.CO;2-P
comes among extremely preterm infants 6.5 years after active peri- Yudin, L. W. (1966). An abbreviated form of the WISC for use with
natal care in Sweden. JAMA Pediatrics, 170(10), 954–963. https:// emotionally disturbed children. J Consult Psychol, 30(3), 272–275.
doi.org/10.1001/jamapediatrics.2016.1210 https://doi.org/10.1037/h0023312