/clinical/,/clinical/cckm-tools/,/clinical/cckm-tools/content/,/clinical/cckm-tools/content/questionnaires/,/clinical/cckm-tools/content/questionnaires/related/,

/clinical/cckm-tools/content/questionnaires/related/name-97140-en.cckm

201606168

page

100

UWHC,UWMF,

Clinical Hub,UW Health Clinical Tool Search,UW Health Clinical Tool Search,Questionnaires,Related

Psychometric Properties of the Vanderbilt ADHD Diagnostic Parent Rating Scale in a Referred Population

Psychometric Properties of the Vanderbilt ADHD Diagnostic Parent Rating Scale in a Referred Population - Clinical Hub, UW Health Clinical Tool Search, UW Health Clinical Tool Search, Questionnaires, Related


Psychometric Properties of the Vanderbilt ADHD Diagnostic Parent
Rating Scale in a Referred Population
Mark L. Wolraich,
1
MD, Warren Lambert,
2
PD, Melissa A. Doffing,
1
MA, Leonard Bickman,
2
PD, Tonya Simmons,
2
BS, and Kim Worley,
2
MD
1
Child Study Center, University of Oklahoma Health Sciences Center, and
2
Center for Mental Health
Policy, Vanderbilt University
Objectives To determine the psychometric properties of the Vanderbilt Attention
Deficit/Hyperactivity Disorder Parent Rating Scale (VADPRS), which utilizes information
based on the Diagnostic and Statistical Manual of Mental Disorders, 4th Ed. (DSM-IV). The
VADPRS was created to collect uniform patient data and minimize the time burden of lengthy
interviews. Methods Participant data (N = 243) was used from the first 2 years of a longi-
tudinal study on communication among physicians, teachers, and parents in diagnosing, treat-
ing, and managing children with attention deficit/hyperactivity disorder (ADHD). The reliabil-
ity, factor structure, and concurrent validity of the VADPRS were evaluated and compared with
ratings of children in clinical and nonclinical samples on the Vanderbilt ADHD Teacher Rating
Scale and the Computerized Diagnostic Interview Schedule for Children–IV, Parent version.
Results The internal consistency and factor structure of the VADPRS are acceptable and con-
sistent with DSM-IV and other accepted measures of ADHD. Conclusion The VADPRS is
a reliable, cost-effective assessment for ADHD in clinical and research settings.
Key words ADHD; parent; behavior; rating scale.
Attention deficit/hyperactivity disorder (ADHD) is the
most common neurobehavioral diagnosis affecting chil-
dren today (Olfson, 1992; Shaywitz & Shaywitz, 1988).
Given the widespread attention that ADHD is receiving
in health care and the media (Angier, 1994; Diller, 1996;
McGinnis, 1997), a uniform process to evaluate children
who present with the core symptoms of inattention, hy-
peractivity, impulsivity, or poor academic achievement is
warranted. Guidelines of both child psychiatry (Dulcan,
1997) and pediatrics (American Academy of Pediatrics,
2000, 2001) encourage clinicians to employ criteria of
the Diagnostic and Statistical Manual of Mental Disorders,4th
Ed. (DSM-IV) (American Psychiatric Association [APA],
1994) in making the diagnosis.
Behavior rating scales have been one method for ob-
taining information from parents and teachers efficiently.
Most earlier scales, such as the Conners Rating Scales
(Goyette, Conners, & Ulrich, 1978) and the Child Be-
havior Checklist (Achenbach & Edelbrook, 1983), differ
from DSM-IV in several ways: (a) They were more broad
based, (b) they did not include all the specific DSM crite-
ria required to make a diagnosis, and (c) they derived
their categories based on deviations from the norm. Scales
specific for ADHD utilizing the 18 core symptoms have
been developed (Conners, Sitarenios, Parker, & Epstein,
1998; DuPaul et al., 1997; Molina, Smith, & Pelham, 2001;
Swanson, Nolan, & Pelham, 1982; Wolraich, Feurer, Han-
nah, Pinnock, & Baumgaertel, 1998) for parents and/or
teachers. In addition to the ADHD core symptoms, some
of the scales include symptoms for at least the other dis-
ruptive behaviors. The Vanderbilt ADHD Teacher Rating
Scale (VADTRS) (Wolraich et al., 1998) is a relatively
simple instrument that follows the DSM-IV criteria for
ADHD. In addition, the VADTRS has included a screen
for some of the mood and anxiety symptoms and a rat-
ing of the child’s performance (Jellinek, Patel, Froehle,
2002). While the psychometric properties of the VADTRS
have been reported (Wolraich et al., 1998), those of a par-
Journal of Pediatric Psychology, Vol.  No. , , pp. – ©  Society of Pediatric Psychology
DOI: 10.1093/jpepsy/jsg046
All correspondence concerning this article should be addressed to Dr. Mark Wolraich, OU Health Sciences Center, Child
Study Center, 1100 NE 13th Street, Oklahoma City, Oklahoma 73117. E-mail: mark-wolraich@ouhsc.edu.

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


ent version have not. The purpose of the current study
was to examine the psychometric properties of a Vander-
bilt ADHD Parent Rating Scale (VADPRS).
Methods
Overview
Data were gathered as part of a longitudinal study of com-
munication among physicians, teachers, and parents in
diagnosing, treating, and managing children with ADHD.
Children were recruited from an urban elementary school
system; data were then collected from their parents, teach-
ers, and physicians. Data used in the present study come
from the first 2 years of a 4-year longitudinal study.
Longitudinal Data Collection Procedures
The data collection for the current study included an ini-
tial time-zero screen and three follow-up data collection
waves. Information during the teacher screening process
was collected anonymously, and parental consent was ob-
tained from all the participants through procedures ap-
proved by the university institutional review board. The
time-zero screen was conducted during the 1998 –1999
school year. The communication study was presented to 61
out of 67 eligible elementary schools. The school system
had an overall total K-12 enrollment of 69,400 students
(46.6% black and 46.0% white). Average K-3 class size
was 20 students. Of the 975 teachers, 317 (at 57 schools)
chose to participate in the study by completing behavior-
rating scales (the VADTRS) on all the students in their
classrooms. According to chi-squared tests, participating
teachers were more likely than nonparticipating to have an
education at the master’s level and above, but did not
differ by sex (p = .60), race (p = .18), or type of teacher
(e.g., Title I, certified, grades K– 4, p = .16). According
to t tests, average age (p = .07) and years experience
(p = .12) were not significantly different for participating
teachers.
There were 6,171 rating scales completed, of which
1,536 students were eligible for the study due to a teacher-
reported clinical diagnosis of ADHD or by meeting rat-
ing scale criteria for ADHD. The teacher-reported ADHD
screen showed a very high rate of 23% possible ADHD
cases. This rate is much higher than the usual 3 –5% esti-
mates (APA, 1994) for two reasons. First, rates of ADHD
were high in this district. The rate of “any ADHD” in a
suburban county measured by the same method was 14%.
Second, the teacher screen doesn’t include all the criteria
required to make a diagnosis. It does not include dura-
tion of symptoms and whether they started before the age
of 7. In addition, the teacher screening does not include the
requirement of dysfunction in more than one setting for
the diagnosis of ADHD (criterion C in DSM-IV). The ad-
ditional requirements are likely to produce lower rates of
ADHD. When ADHD, as measured by the teacher screen,
is used to predict the clinical diagnosis of ADHD (as re-
ported by the teacher), the sensitivity was 63% and the
specificity was 78%.
The first wave of the follow-up included multiple at-
tempts made by the school staff to contact the parents of
the 1,536 eligible students by letters and/or phone calls.
(The project had to remain blind to confidential identi-
fying information until after parents consented to partic-
ipate.) The parents who participated (N = 288) completed
a fully structured Computerized Diagnostic Interview
Schedule for Children (C-DISC-IV; National Institute of
Mental Health, 1997) interview in person by researchers.
In April-November 1999, the VADTRS was sent to the
teachers of the 288 participants; 89.9% of the surveys
were returned (N = 259).
The second wave, approximately 6 months later, in-
cluded a second interview (by phone) utilizing the parent
rating scale (VADPRS) with the parents of 261 of the chil-
dren (90.6%). Reasons for attrition of the parents varied:
4 were no longer interested, 15 had moved out of the area,
and 8 could not be located.
The third wave, 6 months after the second interview,
included a phone interview using the ADHD section of
the C-DISC-IV and the VADPRS with 256 (95.2%) of the
remaining 269 parents. Reasons for attrition varied: 2
moved in with a new caregiver, 4 had moved out of the
area, and 7 could not be located. Additionally, the VADTRS
was mailed to the teachers of the 269 children, with 89.6%
(N = 241) of the surveys returned.
Measures
The VADPRS is the parents’ version of the teacher rating
scale, the VADTRS (Wolraich et al., 1998). It includes all
18 of the DSM-IV criteria for ADHD. In addition, 8 crite-
ria for oppositional defiant disorder (ODD) and 12 crite-
ria for conduct disorder (CD) are included, along with 7
criteria from the Pediatric Behavior Scale (Lindgren &
Koeppl, 1987) that screen for anxiety and depression. The
wording has been simplified so that the reading level is
slightly below third grade. As with the teachers’ form, the
parents are asked to rate the severity of each behavior
on a 4-point scale (“never” to “very often”). The diagno-
sis is considered present if scores of 2 or 3 on a 0 –3 scale
(indicating that a behavior is “often” or “very often” pres-
ent) are checked for the requisite number of criteria based
on the DSM-IV definition of ADHD diagnosis.
The performance section of the VADPRS is an eight-
 Wolraich et al.

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


item scale with four items relating to academic perfor-
mance: (a) overall academic performance, (b) reading,
(c) mathematics, and (d) written expression. Another four
items evaluate relationships: (e) peers, (f) siblings, (g)
parents, and (h) participation in organized activities. The
parent rates each of these on a 5-point scale from “prob-
lematic” to “above average.”
Results
The eligible sample (N = 1,536) shown in Table I includes
children who met the behavior rating scale criteria for
ADHD on the teacher screen or who had a teacher-reported
clinician diagnosis of ADHD. All children were in ele-
mentary school (grades K– 4), and a little more than half
(52%) were African American. The modal age for partic-
ipants was 7 (mean 7.41). As expected, the sample in-
cluded mostly boys (68 –69%). Two hundred forty-three
participants were still in the project after completing the
screening and the three data collection points with parents,
teachers, and physicians. For simplicity, we call this a clin-
ical sample of children with ADHD. Strictly speaking,
though, children positive on the teacher screen did not
necessarily meet the two requirements of DSM-IV: age of
onset and impairment in two or more settings.
In most ways the study sample of 243 children with
Psychometric Properties of the VADPRS 
Table I. Description of 1,536 Children with Attention Deficit/Hyperactivity Disorder (ADHD) (from 6,171 elementary school children)
Screen Positive Study Volunteers
N = 1,293 N = 243
Child Characteristic Mean SD Mean SD p
Child characteristic (range)
Teacher-based N inattention problems (0 –9) 6.84 2.53 6.40 2.79 *
Teacher-based N hyperactive problems (0 –9) 4.90 3.30 4.74 3.23 NS
Teacher-based N ODD-CD problems (0 –10) 1.84 2.80 1.87 2.93 NS
Teacher-based N anxiety problems (0 –7) 1.11 1.91 0.97 1.68 NS
Teacher-based N performance problems (0 –8) 5.52 2.19 5.28 2.32 NS
Child Characteristic, %
Clinician diagnosed with ADHD
a
15.3 24.3 **
On stimulant medication
a
12.2 22.6 **
Ever referred for S-team
a,b
26.2 37.3 **
VADTRS ADHD inattentive type
c
43.8 41.6 NS
VADTRS ADHD hyper/impulsive type
c
14.8 14.0 NS
VADTRS ADHD combined type
c
35.7 31.7 NS
VADTRS ADHD any type
c
94.3 87.2 **
ODD-CD
c
26.3 25.5 NS
Anxiety/depression
c
17.2 13.6 NS
Malea 68.2 68.5 NS
Ethnicity
a
African American 52.4 55.6 NS
White 39.0 40.5 NS
Other 8.6 3.9 *
Grade
a
K 21.7 21.8 NS
First 31.5 36.2 NS
Second 27.6 26.3 NS
Third 7.4 7.4 NS
Fourth 12.8 8.2 NS
Age, y
d
— — 7.41 1.32 —
ODD-CD = oppositional defiant disorder–conduct disorder; VADTRS = Vanderbilt ADHD Teacher Rating Scale. All data from teacher report at original screening (N = 6,171).
a
As reported by teacher.
b
S-team = support team meeting in which school reviews a child with problems.
c
Diagnosis calculated from VADTRS according to Diagnostic and Statistical Manual, 4th Ed. rules. In this context, “with ADHD” means a positive teacher screen or teacher-
reported clinician diagnosis of ADHD.
d
Age not available for nonparticipants.
* p = .05.
** p = .01.

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


ADHD, or ADHD characteristics, resembled the sample
of eligible nonparticipants, e.g., by VADTRS scores, in-
ternalizing and externalizing comorbidity, race, grade in
school. However, significant differences appear in recog-
nition of ADHD, with higher percents of the participant
sample having a clinical diagnosis, being treated with
stimulant medication, or being referred to a special /support
team (S team) for problems in school.
Attrition
For confidentiality, the research team was blinded to iden-
tifying data until after parents volunteered to participate.
Invitations to participate had to be sent out by schools,
leaving two reasons for nonparticipation: (a) parents chose
not to participate, or (b) schools failed to get invitations to
the parents. The higher rates of clinical diagnosis, med-
ication, and S-team referrals in the study sample (N =
243) suggest that parents who chose to participate were
somewhat more likely than nonparticipants (N = 1,293)
to have children with recognized ADHD problems.
Analysis
Analysis of the VADPRS appears under five headings: (a)
internal consistency reliability, (b) item analysis, (c) fac-
tor structure, (d) concurrent validity, and (e) comorbid
scales for factor structure, reliability, and validity.
Internal Consistency Reliability. This analysis compares
the VADPRS with the VADTRS and C-DISC-IV using avail-
able samples, including teacher and parent ratings of clin-
ical samples and a large screening sample of school chil-
dren. The main question concerns the reliability of the
VADPRS, but the various samples also let us determine
whether the DSM-IV criteria are reliable in low-end
samples (school children without ADHD) and high-end
samples (children identified as having ADHD).
Whole scale internal consistency reliabilities for the
VADPRS, VADTRS, and C-DISC-IV appear in Table II. Re-
sults show good internal consistency in all methods and
samples, with the overall Cronbach’s alpha ≥.90 in every
case.
Item Analysis. The next analysis examines item corre-
lations to determine whether some of the symptoms prove
unreliable when used for parent reports in a clinical
sample. Item reliabilities were evaluated with part-whole
correlations, the correlation of a given item with the sum
of all other items. Low part-whole correlations indicate
items that are unreliable in a given sample, perhaps be-
cause of the restricted range.
Figure 1 compares item correlations for the different
instruments (VADPRS, VADTRS, and C-DISC-IV). Rank-
ing the part-whole correlations for each instrument sample
made it possible to compare the overall quality of item
correlations from best (highest) to worst (lowest). These
item correlations show no items with inadequate part-
whole correlations, which would be anticipated for the
well-researched symptoms of ADHD. In this respect the
parent rating scale is no worse, and possibly a bit better,
than the VADTRS and C-DISC-IV.
The best and worst items are identified on the chart,
showing for example that speech symptoms (excessive
talking, blurting out of answers) have the lowest correla-
tions with other symptoms of ADHD.
Factor Structure. Factor analysis of the VADPRS ADHD
scales, as opposed to the comorbidity scales, is done to
assess consistency with the DSM-IV measurement model
for ADHD, as shown in Figure 2. This model views ADHD
as comprising two separate but correlated components:
inattention and hyperactivity/impulsivity.
The measurement model in Figure 2 is based on DSM-
IV. Symptoms 1–9 measure inattention, symptoms 10 –
18 measure hyperactivity/impulsivity, and the two groups
of symptoms are correlated. In addition to showing the
 Wolraich et al.
Table II. Internal Consistency Reliability for Ratings of 18 Attention Deficit/Hyperactivity Disorder (ADHD) Symptoms
Alpha
a
N Rater Sample Instrument
b
Time
0.94 255 Parent Clinical study sample V Wave 2
0.95 253 Parent Clinical study sample V Wave 3
0.92 4,582 Teacher School screen, ADHD– V Time zero screen
0.91 1,268 Teacher School screen, ADHD+ no follow-up V Time zero screen
0.90 240 Teacher School screen, ADHD+ with follow-up V Time zero screen
0.93 251 Teacher Clinical study sample V Wave 1
0.94 234 Teacher Clinical study sample V Wave 3
0.93 286 Parent Clinical study sample D Wave 1
0.94 245 Parent Clinical study sample D Wave 3
a
Raw Cronbach alphas. Standardized alphas were identical to two significant figures because all scales used the same 0 –3 ratings.
b
V = Vanderbilt ADHD Teacher or Parent Rating Scale; and D = Computerized-Diagnostic Interview Schedule for Children, 4th Ed. “in the last month.”

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


hypothesized model, Figure 2 shows the empirical load-
ings, which will be explained later.
The first step in the factor analysis was the examina-
tion of the eigenvalues of the correlations matrix in a scree
plot (Figure 3), which provides an initial estimate of how
many factors might reasonably be extracted.
For comparison, eigenvalues for the teacher screen
were included, and the analysis was limited to the 222
children who had all three measures, the VADPRS at Wave
2, the VADPRS at Wave 3, and the VADTRS at Wave 3.
In all 3 samples, the third eigenvalue was less than 1.0,
which suggests that the correlation matrix can be ex-
plained by two factors and that adding a third would be of
little use.
According to confirmatory factor analysis with EQS
software (Multivariate Software, Inc., Encino, California;
Bentler, 1985; Bentler & Bonett, 1980), the VADPRS
is consistent with the two-factor model of ADHD. In
both parent samples, the fit between parent-reported
data and the two-factor model was satisfactory (> .90),
as shown in Table III. The correlation between inatten-
tion and hyperactivity/impulsivity was high (r = .75 to
.79).
The confirmatory model assumes that each symptom
is due to its latent source (inattentive or hyperactive/
impulsive ADHD) and error; the standardized coefficients
in Figure 2 are quite uniform and consistent with a two-
factor model, again indicating the maturity of the DSM-IV
symptom list used in the VADPRS.
Concurrent Validity. The next analysis examines the cor-
relation between VADPRS and the C-DISC-IV in order to
evaluate the concurrent validity of the VADPRS, i.e., its
correlation with an instrument of established reliability
and validity. The VADPRS and ADHD section of the C-
DISC-IV ratings were completed on the same occasion.
Concurrent validity of the item total of the VADPRS is
high (r = .79), suggesting that it measures much the same
thing that the C-DISC-IV does. Internal consistency reli-
abilities of the C-DISC-IV and VADPRS were .93 or higher;
a reliability of .93 can be interpreted as an expectation of
r = .93 for two parallel forms of a given test. Since the ob-
served correlation is lower than .93, we conclude that the
VADPRS and the C-DISC-IV are very similar, but are not
parallel forms of the same test.
Comorbid Scales: Factor Structure, Reliability, and Validity.
In addition to the two ADHD scales, the VADPRS has two
comorbidity scales to assess internalizing problems (anx-
iety and depression) and externalizing problems (ODD
and CD) that often complicate ADHD. In this sample, 8%
of the children were positive for internalizing comorbid-
ity, and 23% were positive for externalizing comorbidity.
Histograms for criteria counts appear in Figure 4. Both
Psychometric Properties of the VADPRS 
Figure 1. Sorted Part-Whole Item
Correlations for Attention Deficit/
Hyperactivity Disorder Symptoms.
Part–whole correlations are lower
for “bad items” that are unreliable in
the sense of measuring something
different from the other items. Item
correlations are adequate in all
samples and with all variants of the
ADHD symptom list in this figure.
These item correlations are no worse
for parent ratings on the VADPRS,
compared to teacher or parent rat-
ings based on the C-DISC-IV.

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


comorbidity counts are highly skewed, having modes of
zero and a few very high scores.
Since the teacher form has a four-factor structure
(Wolraich, Hannah, Baumgaertel, & Feurer, 1998), we
tested this structure in confirmatory factor analyses us-
ing the same methods as noted earlier in this paper, ex-
cept for the addition of 20 items for ODD or CD and 7
for anxiety or depression. Factors 1 and 2 were again
the nine symptoms of inattention and nine symptoms
of hyperactivity/impulsivity. When the 28 comorbid
items were forced into a single Factor 3, the model’s fit
was inadequate (comparative fit index [CFI] = .88). Fit
was satisfactory for a model with four factors (inatten-
tion, hyperactivity/impulsivity, ODD-CD, and anxiety-
depression). In this case CFI = .93. This result suggests that
the VADPRS, like the related teacher scale, can be con-
sidered a four-factor test instrument.
Reliability analysis suggested that the comorbid scales’
 Wolraich et al.
Figure 2. Factor Model Based on the Diagnostic and Statistical Manual of
Mental Disorders, 4th Ed. The model states that ADHD is measured by 18
items in two factors (inattention and hyperactivity/impulsivity), and that
the scores for these two factors are correlated.
internal consistency was adequate. Cronbach’s alpha for
ODD-CD (20 items) was .91; and for anxiety-depression
(7 items), .79. The comorbidity scales showed some va-
lidity in correlations with hypothesized events and mea-
sures, but not enough to claim that convergent and dis-
criminant validity (Campbell & Fiske, 1959; Fiske &
Campbell, 1992) were well established. For example, the
internalizing problem count had a significant correlation
with scores on the Columbia Impairment Scale (r = .43,
p < .0001) and children above the internalizing cut score
were significantly more likely to have an S-team (p = .01)
or school suspension (p = .002). However, they were no
more likely to have contact with a mental health provider
(p = .15) or detention in school (p = .18) or to miss more
days of school (r = .06, p = .40).
The externalizing problem count had significant cor-
relations with both Columbia Impairment Scale scores
(r = .52, p < .0001) and days of school missed during the
year (r = .24, p < .0001), and children above the cut score
for externalizing comorbidity were significantly more
likely to have been suspended (p = .003) and have contact
with a mental health provider (p = .003). However, they
were no more likely for detention in school (p = .07) or
having an S-team (p = .13).
Discussion
While the VADPRS uses the well-tested ADHD symptom
list from the DSM-IV, it still had to be evaluated empirically
before field use. This was done in the present study. The re-
liability, factor structure, and concurrent validity of the
VADPRS were evaluated and compared with VADTRS and
C-DISC-IV ratings of children in clinical and nonclinical
samples. Results suggest that the internal consistency and
factor structure of the VADPRS are acceptable and con-
sistent with DSM-IV and other accepted measures of
ADHD. Table II (internal consistency in nine samples)
suggests that the criteria are robust, having good reliabil-
ity under a variety of conditions, including different
respondents (parent and teacher), different methods
(Vanderbilt checklists vs. C-DISC-IV interview), and dif-
ferent severities (ranging from the 4,582 ADHD-negative
children to the various clinical samples). This adequacy
is not unexpected. Since the VADPRS, VADTRS, and C-
DISC-IV all use the same 18-symptom list from the
DSM-IV, large differences in reliability or validity would not
be anticipated. Confirmatory factor analysis and reliabil-
ity assessment also confirm the structure of the two co-
morbid dimensions of internalizing and externalizing con-
ditions.
The present study suggests that the VADPRS has good

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


psychometric properties in a high-risk sample; however, a
study of a referred sample cannot establish the sensitivity
and specificity of diagnostic method. It is possible that
the VADPRS would perform differently on a low-risk
sample with different characteristics. However, when the
same items are used with teachers, confirmatory factor
analysis showed that urban and suburban samples shared
the same factor structure, and that the factor structure
showed only small differences by gender, age, school grade,
or severity of ADHD symptom counts (Wolraich et al.,
2002). Like the VADTRS, the VADPRS makes no attempt
to enforce the DSM-IV’s ≥2-setting requirement, so preva-
lence rates may be higher than the true rates. Wolraich
et al. (in press) found that parent-teacher agreement was
low, with symptom count correlations ranging from r =
.27–.34 and ADHD diagnostic agreement ranging from
Kappa = .11 to .15. Strict enforcement of the ≥2-setting re-
quirement led to rates of diagnosis much lower than par-
ent or teacher alone. The VADPRS is not a substitute for
clinical diagnosis. Together, the VADPRS and VADTRS
can be useful adjuncts to an evaluation.
The VADPRS is easy to complete (having a reading
level slightly below third grade), easy to score following the
criteria required to make a DSM-IV diagnosis, and in-
cludes both the core symptoms and a rating of perfor-
mance as well as providing a screen for common comor-
bid conditions. Its ease of administration and scoring as
well as its provision of specific DSM-IV information make
it a useful clinical tool in evaluating patients and a use-
ful research tool for evaluating ADHD in a cost-efficient
manner.
Rating scales such as the VADPRS are systematic and
require less professional time than do interviews to de-
termine the presence of specific behaviors. The high cor-
relations between rating scales and direct interview in an
urban inner-city sample suggest that rating scales may be
used as accurate measures of a child’s behavior for the
purposes of diagnosing ADHD, even with a sample of par-
ents with low reading levels. A rating scale can reduce the
time a clinician requires to gather information and thereby
allow the clinician to focus the clinical interview on other
issues. However, it is important to note that the scales are
Psychometric Properties of the VADPRS 
Table III. Results of Confirmatory Analysis of Measurement Model
Wave 2 Wave 3
Method or Result VADPRS VADPRS
N parental ratings 242 242
Item scores missing none none
Bentler Comparative Fit Index (CFI) 0.923 0.914
Bentler robust CFI 0.941 0.940
Correlation r (inattention hyperactivity) .792 .753
VADPRS = Vanderbilt Attention Deficit/Hyperactivity Disorder Parents Rating
Scale.
CFI’s from EQS software (Bentler & Wu, 1993; Hu & Bentler, 1995) indicate
how well model fits observed data. CFI > .90 considered adequate (Byrne, 1994,
p. 55). Robust estimate corrects for possible non-normal distribution of item
scores. Sample includes same children, each rated twice with no missing items.
Figure 3. Eigenvalues of Three Corre-
lation Matrices

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


not substitutes for clinical interviews to obtain information
from parents and children in evaluating a child for possible
ADHD.
Rating scales such as the VADPRS are important for re-
search. They can obtain information on large groups of
children at low costs. Its reasonable correlation with the C-
DISC-IV makes it possible to attain DSM-IV diagnostic
information for ADHD without having to employ more
costly structured interviews. Such a process could reduce
the reliance on convenience-based samples, which is cur-
rently the norm in many studies, and allow for samples
that better reflect the overall population. This change will
provide more accurate prevalence estimates.
The present study had limitations. The rate of par-
ticipation among eligible families was low. However, there
were few differences between participants and nonpartic-
ipants. Another limitation was the absence of test-retest re-
liability. While we were able to use a high-risk sample to
compare the VADPRS to a gold standard for ADHD, the C-
DISC-IV, and we could demonstrate a relationship be-
tween the comorbid conditions and measures of impair-
ment, we could not specifically assess the concurrent
validity of the comorbid conditions and performance. A
study to obtain this assessment is currently in progress.
Acknowledgement
This study was supported by a grant from the National
Institute of Mental Health (HS/MH 0905).
 Wolraich et al.
Received August 19, 2002; first revisions received November
11, 2002; second revisions received January 27, 2003; ac-
cepted March 21, 2003
References
Achenbach, T. M., & Edelbrook, C. (1983). Manual for
the Child Behaviour Checklist and revised Child Be-
haviour Profile. Burlington: University of Vermont.
American Academy of Pediatrics (2000). Clinical prac-
tice guideline: Diagnosis and evaluation of the child
with attention-deficit/hyperactivity disorder. Pedi-
atrics, 105, 1158 –1170.
American Academy of Pediatrics (2001). Clinical prac-
tice guideline: Treatment of the school-aged child
with attention-deficit/hyperactivity disorder. Sub-
committee on ADHD and Committee on Quality
Improvement. Pediatrics, 108, 1033 –1044.
American Psychiatric Association (1994). Diagnostic
and statistical manual of mental disorders. 4th ed.
Washington, DC: Author.
Angier, N. (1994, July 24). The debilitating malady
called boyhood. New York Times, section 4, pp. 1, 4.
Bentler, P. (1985). EQS structural equations program
manual [Computer software]. Encino, CA: Multi-
variate Software, Inc.
Bentler, P., & Bonett, D. G. (1980). Significance tests
and goodness-of-fit in the analysis of covariance
structures. Psychological Bulletin 88, 588 –600.
Figure 4. Counts of Comorbid
Criteria

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m


Bentler, P. M., & Wu, E. J. C. (1993). EQS/Windows
User’s Guide [Computer software]. Los Angeles:
BMDP Statistical Software, Inc.
Byrne, B. M. (1994). Structural Equation Modeling
with EQS and EQS/Windows [Computer software].
Thousand Oaks, CA: Sage Publications.
Campbell, D. T., & Fiske, D. W. (1959). Convergent
and discriminant validation by the multitrait-
multimethod matrix. Psychological Bulletin, 56,
81–105.
Conners, C. K., Sitarenios, G., Parker, J. D., & Epstein,
J. N. (1998). Revision and restandardization of the
Conners Teacher Rating Scale (CTRS-R): Factor
structure, reliability, and criterion validity. Journal
of Abnormal Child Psychology, 26, 279 –291.
Diller, L. H. (1996). The run on Ritalin: Attention
deficit disorder and stimulant treatment in the
1990’s. Hastings Center Report, 26, 12 –18.
Dulcan, M. (1997). Practice parameters for the assess-
ment and treatment of children, adolescents, and
adults with attention-deficit /hyperactivity disorder.
Journal of American Academy of Child and Adolescent
Psychiatry, 36(Suppl 10), 85S–121S.
DuPaul, G. J., Power T. J., Anastopoulos, A. D., Reid, R.,
McGoey, K. E., Ikeda, M. J. (1997). Teacher ratings
of attention deficit hyperactivity disorder symp-
toms: Factor structure and normative data. Psycho-
logical Assessment, 9, 436 – 444.
Fiske, D. W., & Campbell, D. T. (1992). Citations do
not solve problems. Psychological Bulletin, 112,
393 –395.
Goyette, C. H., Conners, C. K., & Ulrich, R. F. (1978).
Normative data on revised Conners Parent and
Teacher Rating Scales. Journal of Abnormal Child
Psychology, 6, 221–236.
Hu, L. T., & Bentler, P. M. (1995). Evaluating model fit.
In H. H. Rick (Ed.), Structural equation modeling:
Concepts, issues, and applications (pp. 76 –99).
Thousand Oaks, CA: Sage Publications, Inc.
Jellinek, M. J., Patel, B. P., Froehle, M. C. (2002). Bright
futures in practice: Mental health. Arlington, VA:
National Center for Education in Maternal and
Child Health, Georgetown University.
Lindgren, S., & Koeppl, G. (1987). Assessing child
behavior problems in a medical setting: Develop-
ment of the Pediatric Behavior Scale. In R. J. Prinz
(Ed.), Advances in behavioral assessment of children
and families (Vol. 3, pp. 57–90). Greenwich, CT:
JAI Press.
Molina, B. S., Smith, B. H., & Pelham, W. E. (2001).
Factor structure and criterion validity of secondary
school teacher ratings of ADHD and ODD. Journal
of Abnormal Child Psychology, 29, 71–82.
McGinnis, J. (1997, September 18). Attention deficit
disaster. Wall Street Journal, p. A14.
National Institute of Mental Health (1997). Computer-
ized Diagnostic Interview Schedule for Children, IV
(NIMH C-DISC IV). New York: Author, Columbia
University.
Olfson, M. (1992). Diagnosing mental disorders in
office-based pediatric practice. Journal of Develop-
mental and Behavioral Pediatrics, 13, 363 –365.
Shaywitz, B. A., & Shaywitz, S. E. (1988). Attention
deficit disorder: Current perspectives. In J. F. Kava-
nagh & T. J. Truss (Eds.), Learning disabilities: Pro-
ceedings of the national conference (pp. 369 –523).
Parkton, MD: York Press.
Swanson, J., Nolan, W., & Pelham, W. E. (1982). SNAP
rating scale. Educational Resources Information
Center (ERIC).
Wolraich, M. L., Feurer, I., Hannah, J. N., Pinnock,
T. Y., & Baumgaertel, A. (1998). Obtaining system-
atic teacher reports of disruptive behavior disorders
utilizing DSM-IV. Journal of Abnormal Child Psy-
chology, 26, 141–152.
Wolraich, M. L., Hannah, J. N., Baumgaertel, A.,
& Feurer, I. D. (1998). Examination of DSM-IV
critieria for attention deficit/hyperactivity disorder
in a county-wide sample. Journal of Developmental
and Behavioral Pediatrics, 19, 162 –168.
Wolraich, M. L., Lambert, E. W., Baumgaertel, A.,
Garcia-Tornel, S., Feurer, I. D., Bickman, L., et al.
(in press). Teachers’ screening for attention deficit/
hyperactivity disorder: Comparing multinational
samples on teacher ratings of ADHD. Journal of
Abnormal Child Psychology, 31, 445 – 455.
Wolraich, M. L., Lambert, E. W., Bickman, L., Simmons,
T., Doffing, M. A., & Worley, K. A. (in press). As-
sessing the impact of parent and teacher agreement
on diagnosing ADHD. Journal of Developmental and
Behavioral Pediatrics.
Psychometric Properties of the VADPRS 

a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m



a
t

U
n
i
v
e
r
s
i
t
y

o
f

W
i
s
c
o
n
s
i
n
-
M
a
d
i
s
o
n

L
i
b
r
a
r
i
e
s

o
n

F
e
b
r
u
a
r
y

2
,

2
0
1
5
h
t
t
p
:
/
/
j
p
e
p
s
y
.
o
x
f
o
r
d
j
o
u
r
n
a
l
s
.
o
r
g
/
D
o
w
n
l
o
a
d
e
d

f
r
o
m