NIH Public Access

Author Manuscript
J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Published in final edited form as:
NIH-PA Author Manuscript

J Clin Child Adolesc Psychol. 2014 ; 43(4): 566–578. doi:10.1080/15374416.2013.814541.

Assessing Anxiety in Youth with the Multidimensional Anxiety

Scale for Children (MASC)
Chiaying Wei, M.A.,
Department of Psychology, Weiss Hall, Temple University, 1701 North 13th Street, Philadelphia,
PA 19087, U.S.A

Alexandra Hoff,
Department of Psychology, Weiss Hall, Temple University, 1701 North 13th Street, Philadelphia,
PA 19087

Marianne A. Villabø,
Center for Child and Adolescent Mental Health, Oslo, Norway
NIH-PA Author Manuscript

Jeremy Peterman,
Department of Psychology, Weiss Hall, Temple University, 1701 North 13th Street, Philadelphia,
PA 19087

Philip C. Kendall, Ph.D., ABPP,

Department of Psychology, Weiss Hall, Temple University, 1701 North 13th Street, Philadelphia,
PA 19087, U.S.A

John Piacentini,
University of California Los Angeles, Los Angeles, CA

James McCracken,
University of California Los Angeles, Los Angeles, CA

John T. Walkup,
Cornell University, New York, NY

Anne Marie Albano,

NIH-PA Author Manuscript

Columbia University, New York, NY

Moira Rynn,
Columbia University, New York, NY

Joel Sherrill,
National Institute of Mental Health, Bethesda, MD

Dara Sakolsky,
University of Pittsburgh Medical Center, Pittsburgh, PA

Boris Birmaher,
University of Pittsburgh Medical Center, Pittsburgh, PA

Correspondence should be addressed to Chiaying Wei or Philip C. Kendall, Department of Psychology, Weiss Hall, Temple
University, 1701 North 13th Street, Philadelphia, PA 19087. [email protected] or [email protected].
Wei et al. Page 2

Golda Ginsburg,
The Johns Hopkins University School of Medicine – Psychiatry, 550 N Broadway, Baltimore, MD
21205
NIH-PA Author Manuscript

Courtney Keaton,
The Johns Hopkins University School of Medicine – Psychiatry, 550 N Broadway, Baltimore, MD
21205

Elizabeth Gosch,
Philadelphia College of Osteopathic Medicine, Philadelphia, PA

Scott N. Compton, and

Duke University Medical Center, Durham, NC

John March
Duke University Medical Center, Durham, NC
Chiaying Wei: [email protected]; Alexandra Hoff: [email protected]; Marianne A. Villabø:
[email protected]; Jeremy Peterman: [email protected]; John Piacentini: [email protected];
James McCracken: [email protected]; John T. Walkup: [email protected]; Anne Marie Albano:
[email protected]; Moira Rynn: [email protected]; Joel Sherrill: [email protected]; Dara
Sakolsky: [email protected]; Boris Birmaher: [email protected]; Golda Ginsburg: [email protected]; Courtney
Keaton: [email protected]; Elizabeth Gosch: [email protected]; Scott N. Compton: [email protected]; John
NIH-PA Author Manuscript

March: [email protected]

Abstract
The present study examined the psychometric properties, including discriminant validity and
clinical utility, of the youth self-report and parent-report forms of the Multidimensional Anxiety
Scale for Children (MASC) among youth with anxiety disorders. The sample included parents and
youth (N= 488, 49.6% male) ages 7 – 17 who participated in the Child/Adolescent Anxiety
Multimodal Study (CAMS). Although the typical low agreement between parent and youth self-
reports was found, the MASC evidenced good internal reliability across MASC subscales and
informants. The main MASC subscales (i.e., Physical Symptoms, Harm Avoidance, Social
Anxiety, and Separation/Panic) were examined. The Social Anxiety and Separation/Panic
subscales were found to be significantly predictive of the presence and severity of social phobia
and separation anxiety disorder, respectively. Using multiple informants improved the accuracy of
prediction. The MASC subscales demonstrated good psychometric properties and clinical utilities
in identifying youth with anxiety disorders.
NIH-PA Author Manuscript

Keywords
anxiety; children; assessment; MASC; parent report

Anxiety disorders are prevalent and tend to emerge during childhood (Kessler et al., 2005).
It is estimated that distressing anxiety affects as many as 20% of youth (Costello, Egger, &
Angold, 2005). Children and adolescents with anxiety disorders experience psychosocial
impairment (Langley, Bergman, McCracken, & Piacentini, 2004), heightened risk for mood
disorders, increased substance use problems, educational underachievement (Woodward &
Fergusson, 2001), and increased rates of suicidal ideation (O’Neil, Puleo, Benjamin, Podell,
& Kendall, 2012). Furthermore, youth-onset anxiety disorders are likely to persist into

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 3

adulthood if left untreated (Costello et al., 2003; Pine, Cohen, Gurley, Brook, & Ma, 1998).
Despite the importance of treatment to stave off potential long-term consequences, only a
fraction of youth with anxiety disorders receive treatment for their anxiety; even among
NIH-PA Author Manuscript

adolescents with severe anxiety, fewer than a third report ever receiving treatment
(Merikangas et al., 2011).

Accurate assessment of anxiety disorders is a necessary first step in proper treatment. Semi-
structured diagnostic interviews are sound methods of identifying anxiety disorders in youth
(e.g., Kaufman et al., 1997; Silverman & Albano, 1996) but they are time consuming and
require an intensive interviewer training, and thus may not be feasible in primary care or
school settings where anxious youth may initially be identified. The use of screening
measures (e.g., questionnaires) may foster the identification of anxious youth in these
settings and be useful for ongoing assessment of symptom change during intervention.
Questionnaires are easy to administer and require minimal time and training. In addition,
information from both youth and parents can easily be obtained. Discrepant reports of
anxiety symptoms from youth and their parents are common (Choudhury, Pimentel, &
Kendall, 2003; Comer & Kendall, 2004; Safford, Kendall, Flannery-Schroeder, Webb, &
Sommer, 2005). Current thinking regarding multiple informant assessment suggests that
NIH-PA Author Manuscript

there are benefits from informants’ observation of clinical relevant behaviors in different
settings (De Los Reyes, Thomas, Goodman, & Kundey, 2013), and researchers suggest that
youth and parents each provide unique and valuable information when reporting particular
anxiety symptoms in youth (Villabø, Gere, Torgersen, March, & Kendall, 2012; Wren,
Bridge, & Birmaher, 2004).

The Multidimensional Anxiety Scale for Children (MASC; March, Parker, Sullivan,
Stallings, & Conners, 1997) offers both youth self-report and parent-report versions to
obtain information on youth anxiety symptoms. The 39-item questionnaire assesses
emotional, cognitive, physical, and behavioral symptoms. The MASC produces a total score,
as well as scores from four subscales: Physical Symptoms, Harm Avoidance, Social
Anxiety, and Separation/Panic. Studies using the MASC have reported high retest reliability
(March et al., 1997; March, Sullivan, & Parker, 1999), favorable divergent and convergent
validity (Baldwin & Dadds, 2007; March et al., 1997; Rynn et al., 2006), and good internal
reliability within the four subscales (e.g., Baldwin & Dadds, 2007; Dierker et al., 2001;
March et al., 1997).
NIH-PA Author Manuscript

Research supports the MASC as a measure that discriminates between youth with and
without anxiety disorders. In a school-based sample, the youth self-report MASC predicted a
diagnosis of generalized anxiety disorder (GAD) in females, but it was less successful at
discriminating youth with other individual anxiety disorders (Dierker et al., 2001). In a
university-based anxiety and depression clinic, the youth-reported MASC total score
significantly identified youth with and without anxiety disorders (van Gastel & Ferdinand,
2008). In addition, the Social Anxiety subscale predicted a social phobia (SoP) diagnosis,
the Separation/Panic subscale predicted panic disorder, agoraphobia, and separation anxiety
disorder (SAD), and the Physical Symptoms subscale predicted panic disorder and
agoraphobia (van Gastel & Ferdinand, 2008). In a community outpatient clinical setting, the
youth self-report version of the MASC was moderately accurate at classifying youth with

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 4

and without anxiety disorders, the Social Anxiety subscale significantly predicted a SoP
diagnosis, and the Harm Avoidance subscale significantly predicted a diagnosis of GAD
(Grills-Taquechel, Ollendick, & Fisak, 2008). In an outpatient sample of youth ages 8 to 17,
NIH-PA Author Manuscript

the MASC differentiated between youth with and without an anxiety disorder (Wood,
Piacentini, Bergman, McCracken, & Barrios, 2002). In addition, youth with social phobia
scored higher on the Social Anxiety subscale and youth with separation anxiety disorder
scored higher on the Separation Anxiety and Harm Avoidance subscales, but the MASC
failed to differentiate youth with a diagnosis of GAD in this study (Wood et al., 2002). Thus,
research investigating the utility of the MASC in predicting specific anxiety disorders is less
conclusive than research supporting its overall ability to discriminate youth who have an
anxiety disorder from youth who do not.

The inconsistent results of previous studies in predicting specific anxiety disorder diagnoses
with the MASC may be attributed to the varying prevalence of certain disorders at different
ages. For example, evidence suggests that the prevalence of anxiety disorders in general
increases from childhood to adolescence, but that rates of separation anxiety disorder and
generalized anxiety disorder decrease while social phobia and panic disorder increase in
prevalence (Costello, Copeland, & Angold, 2011). The studies to date examining the
NIH-PA Author Manuscript

diagnostic utility of the MASC had samples with varying age ranges, some of which
encompassed both childhood and adolescence, but those that examined whether the MASC
was differentially efficient at different ages yielded inconsistent results. Van Gastel and
Ferdinand (2008) found that the MASC better predicted panic disorder and specific phobia
in an adolescent age group as opposed to younger children, but others did not find any age
effects (Grills-Taquechel et al., 2008; Wood et al., 2002). Thus, further study is needed to
examine the role of age in the diagnostic utility of the MASC.

Regarding the utility of the parent-report version, Villabø and colleagues (2012) found, in a
Norwegian community mental health sample, that both mothers’ and youth’s reports on the
MASC discriminated between youth (ages 7 to 13) with and without an anxiety disorder
diagnosis. In addition, the Separation/Panic subscale significantly predicted a diagnosis of
SAD based on the youth, mother, and father report, but only mothers’ and fathers’ reports on
the Social Anxiety subscale were predictors of a SoP diagnosis. Overall, one study suggests
that adding parent report to youth self-report significantly improved the diagnostic
efficiency of the MASC (Villabø et al., 2012; Wood et al., 2002). Further study is needed to
NIH-PA Author Manuscript

replicate and extend these findings, especially given the inconsistent results regarding the
ability of the child-report version to predict specific diagnoses. The current study examines
to role of the unique information provided by youth and by parents in regard to specific
anxiety symptoms/diagnoses in youth.

Previous studies suggest that the youth self- and parent-report MASC can discriminate youth
with and without anxiety disorders, but the ability of the MASC to predict severity of
anxiety in samples of anxious youth merits study. Severity of anxiety symptoms, as
determined by questionnaires (i.e., the MASC) may be useful for informing referral
decisions. In addition, the utility of multiple informants warrants further investigation. The
present study investigated the efficiency of the MASC at predicting severity of anxiety in a
large multi-site sample of anxiety-disordered youth, the largest sample yet employed in

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 5

researching the diagnostic efficiency of the MASC. Given the inclusion of both younger
children (ages 7 to12) and adolescents (ages 13 to 17) in the sample, it was possible to
investigate age differences. The study further examined the merits of the addition of parents
NIH-PA Author Manuscript

as informants. We hypothesized that scores on the parent- and youth self-report MASC
would significantly predict the presence and severity of specific anxiety disorders as
determined by a semi-structured diagnostic interview. Specifically, based on previous
findings, we hypothesized that scores on the Separation/Panic subscale of the MASC would
predict the presence and severity of SAD, scores on the Social Anxiety subscale would
predict SoP, and scores on the Harm Avoidance subscale would predict GAD. We also
hypothesized that both parent and child reports would add significantly to the predictive
power of the MASC.

Method
Participants
The sample consisted of 488 youth (49.6% male) ages 7 to 17 years (M = 10.7 years, SD =
2.8; 74.2% children between ages of 7 to 12) and their parents, who participated in the
Child/Adolescent Anxiety Multimodal Study (CAMS; Kendall, et al., 2010; Walkup, et al.,
NIH-PA Author Manuscript

2008). Among the 488 youth, 78.9% self-identified as white, 9.0% as black, 2.9% as Asian,
1.2 % as American Indian, 0.4% as Pacific Islander, and 8.0% as Other. Socioeconomic
status was classified as “low” for 25.4% of participants. Based on the Anxiety Disorder
Interview Schedule for Children (ADIS-C/P; Silverman & Albano, 1996), all youth met
diagnostic criteria for an anxiety disorder as the principal diagnosis: SoP only (8.2%), GAD
only (6.8%), SAD only (3.3%), SAD and SoP (7.8%), SAD and GAD (9.8%), SoP and GAD
(30.7%), or all three anxiety disorders (33.49%). Other comorbid disorders included other
internalizing disorders (i.e., other anxiety disorders and dysthymia; 43.6%), attention-deficit/
hyperactivity disorder (ADHD; 11.9%), oppositional defiant disorder or conduct disorder
(9.4%), and tic disorder (2.7%). Exclusion criteria for CAMS included a diagnosis of major
depressive disorder, bipolar disorder, a psychotic disorder, a pervasive developmental
disorder, untreated ADHD, an eating disorder, or a substance use disorder; a diagnosis of
any other Axis I disorder with a greater severity than the GAD, SAD, or SoP; recent severe
school refusal behavior; suicidality or homicidality; two previous failed trials of an SSRI or
a failed trial of an adequate course of CBT for GAD, SAD, or SoP; intolerance to sertraline;
NIH-PA Author Manuscript

a confounding medical condition; pregnancy; and if the child or adolescent did not speak
English.

Measures
ADIS-C/P (Silverman & Albano, 1996)—The ADIS-C/P is a semi-structured interview
for assessing youth anxiety disorders based on DSM-IV criteria (American Psychiatric
Association, 1994). During separate interviews, youth and parents reported on the youth’s
anxiety symptoms, as well as symptoms of other Axis I disorders. Independent evaluators
(IEs) provided clinician severity ratings (CSRs; Silverman & Albano, 1996) for each
interview, ranging from 0 to 8; 0 = not at all, 4 = some, and 8 = very, very much. A CSR of 4
or above is required for a diagnosis. A composite CSR for each diagnosis was then
determined based on information from both reports by the IE. The ADIS has demonstrated

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 6

favorable psychometric properties, including excellent retest reliability (Silverman,

Saavedra, & Pina, 2001), good inter-rater agreement among diagnosticians (e.g., Chavira,
Stein, Bailey, & Stein, 2004), and good convergent validity based on self-report measures of
NIH-PA Author Manuscript

anxiety (March, Parker, Sullivan, Stallings, & Conners, 1997; Wood et al., 2002).

MASC, Child and Parent Versions (March et al., 1997)—The MASC C/P is a self-
report questionnaire assessing youth anxiety symptoms. Both the youth and parent versions
consist of 39 items and contain four main subscales: Physical Symptoms, Social Anxiety,
Separation Anxiety/Panic, and Harm Avoidance. The item content and scales in each are
identical, except that the items in the child version refer to “I” and those in the parent
version refer to “my child.” It has demonstrated favorable psychometric properties in
previous studies (Baldwin & Dadds, 2007; Dierker et al., 2001; March et al., 1997, 1999;
Rynn et al., 2006), as reviewed in the introduction.

Procedure
Participants were recruited for CAMS (Walkup et al, 2008), a randomized clinical trial
evaluating youth anxiety disorder treatment at six urban sites in the United States, via
community mental health centers and clinics, community organizations, churches, schools,
NIH-PA Author Manuscript

and advertisements in local media. All participants provided written informed parental
consent and youth assent. Participants completed an initial assessment with an IE to
determine eligibility for randomization to treatment. The 488 participants who met
eligibility criteria completed a baseline assessment before beginning treatment. During this
assessment, an IE interviewed youth and parents separately about the youth’s anxiety and
other symptoms using the ADIS-C/P. The IE then assigned a composite CSR for each
diagnosis based on the information obtained from both the youth and parents. Youth
completed the MASC child version and parents completed the MASC parent version.
Analyses for the present study were conducted using these baseline assessments (For
detailed CAMS procedures, see Compton et al., 2010; Walkup et al., 2008).

Statistical Analyses
Preliminary analyses included t tests to examine differences between youth and parent
reports on the four main MASC subscales and the MASC total score. Gender differences of
these reports were also examined. Cronbach’s internal reliability coefficients (α) were
NIH-PA Author Manuscript

reported for the MASC subscales and total score. Pearson correlations were conducted to
examine the informant agreement between parent and youth reports.

Receiver Operating Characteristic (ROC) analyses examined whether the MASC subscales
predicted the presence of a particular anxiety disorder (i.e., SAD, SoP, or GAD). ROC
analysis indicates the strength of the prediction using area under the curve (AUC), which
ranges from 0.50, no association in prediction, to 1, perfect association in prediction (Hanley
& McNeil, 1982). Regression analyses further examined the association between the MASC
subscales scores and the clinical severity of a particular anxiety disorder (as determined by
the ADIS CSRs). Sequential logistic regression examined whether adding parent report to
youth self-report increased the MASC’s ability to predict the presence of an anxiety
diagnosis.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 7

Results
Preliminary Analyses
NIH-PA Author Manuscript

Means and standard deviations for the MASC subscales and total score are presented in
Table 1. No gender differences were observed on the MASC C/P subscales, except youth-
reported Separation/Panic subscale, t(478) = 2.32, p < .05, with boys reporting higher scores
than girls, and with a small effect size as determined by Cohen’s d (Cohen, 1988). MASC
subscales differed across age groups: compared with adolescents (ages 13 to 17), children
(ages 7 to 12) scored lower on the Physical Symptoms and Social Anxiety subscales and
higher on the Separation/Panic subscale based on both youth and parent report. Children
scored significantly higher on the Harm Avoidance subscale based on youth self-report, but
not parent report.

Cronbach’s internal reliability coefficients (α) for youth and parent report of the MASC total
and subscales are presented in Table 1. Internal reliability estimates for the subscales fell in
the acceptable and good range except for the Harm Avoidance subscale, which demonstrated
lower internal reliability. Informant agreement between youth and parent report are
presented in Table 2. Similar to previous studies (Choudhury et al., 2003; Safford et al.,
NIH-PA Author Manuscript

2005; Villabø et al., 2012), Pearson correlations revealed low youth-parent agreement
overall: a small (Harm Avoidance subscale) or medium effect (Physical Symptoms and
Social Anxiety subscales) was observed for youth/parent agreement on most of the
subscales, with a large effect observed only on the youth/parent agreement on the
Separation/Panic subscale.

Discriminant Validity of MASC

Because there were significant age differences across all MASC subscale scores, primary
analyses on the discriminant validity of the MASC were conducted separately for children (n
= 362) and adolescents (n = 162). Comparisons were made between youth with a diagnosis
of SAD versus those without (SAD versus Non-SAD group), youth with a diagnosis of SoP
versus those without (SoP versus Non-SoP group), and youth with a diagnosis of GAD
versus those without (GAD versus Non-GAD group) in the entire sample, using independent
t-tests and ROC analyses.
NIH-PA Author Manuscript

T-tests—T-tests examined disorder versus non-disorder groups, with results presented in

Table 3. For the child group, both children and parents in the SAD group rated higher on the
Separation/Panic and Physical Symptoms subscale compared to the Non-SAD group, with a
greater difference observed on the Separation/Panic subscale. Compared to the Non-SoP
group, both children and parents in the SoP group rated significantly higher on the Social
Anxiety subscale. Only child-reported scores were significantly lower on the Separation/
Panic subscale. Both children and parents in the GAD group rated higher on the Physical
Symptoms subscale compared to the Non-GAD group. Children in the GAD group also
scored significantly higher on the Harm Avoidance subscales, compared to those in the non-
GAD group.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 8

Reports from the adolescent group indicated the SAD and Non-SAD groups differed most
on the Separation/Panic subscale as per adolescent and parent report. The SAD group also
scored higher on parent-reported Physical Symptoms and Harm Avoidance subscales,
NIH-PA Author Manuscript

compared to the Non-SAD group. According to both adolescent and parent reports, the SoP
group scored significantly higher on the Social Anxiety subscale than the Non-SoP group.
Parent-reported scores in the SoP group were significantly lower on the Harm Avoidance
subscale, compared to the Non-SoP group. The GAD group scored significantly higher on
the Physical Symptoms and Harm Avoidance subscales than the Non-GAD group according
to parent report. The GAD group also scored higher on the Physical Symptoms subscale
according to adolescent self-report.

ROC analyses—ROC analyses examined the ability of the MASC to predict diagnoses of
SAD, SoP, and GAD. The AUC indicates the strength of the prediction ranging from .50 (no
prediction power) to 1 (perfect prediction power), with an AUC below .50 indicating an
inverse prediction. The prediction power is generally considered as .50 - .60 = failed, .60 - .
70 = poor, .70 - .80 = fair, .80 - .90 = good, and .90 - 1.00 = excellent. The results of the
ROC analyses are in Table 4.
NIH-PA Author Manuscript

For children, the Separation/Panic subscale significantly predicted SAD with fair prediction
power, according to both child and parent report. Child report on the Physical Symptoms
subscale also significantly predicted SAD, but with poor prediction power. The Social
Anxiety subscale significantly predicted SoP with fair prediction power, according to both
child and parent report. The child report on the Separation/Panic subscale inversely
predicted SoP, but with poor prediction power. Child and parent report on the Physical
Symptoms subscale as well as child report on the Harm Avoidance subscale demonstrated
significant prediction to GAD, but all of them demonstrated poor prediction power.

For adolescents, the Separation/Panic subscale showed the strongest prediction to SAD,
according to both adolescent and parent report, with fair to good prediction power. Parent
report on the Physical Symptoms and the Harm Avoidance subscales also significantly
predicted SAD, but both with poor prediction power. The Social Anxiety subscale
significantly predicted SoP with good prediction power, according to both adolescent and
parent report. Child and parent report on the Physical Symptoms subscale as well as parent
report on the Harm Avoidance subscale demonstrated significant prediction to GAD, but
NIH-PA Author Manuscript

with poor prediction power.

Associations between MASC and severity of disorders

Multiple regression analyses examined whether the scores of MASC subscales predicted the
severity of SAD, SoP, and GAD for all participants, and child and adolescent groups were
examined separately. The severity of each diagnosis was determined by the CSR given by
the IE based on the ADIS C/P. Table 5 presents results from the regression analyses.

For children, both child and parent report on the Separation/Panic subscale evidenced a
significant relationship with the CSR of SAD. Both child and parent report on the Social
Anxiety subscale and parent report on the Harm Avoidance subscale demonstrated an
inverse relationship with the CSR of SAD, yet all with relative small beta values compared

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 9

to the Separation/Panic subscale. Child and parent report on the Social Anxiety subscale
significantly predicted the CSR of SoP. Child report on the Separation/Panic subscale was
negatively associated with the CSR of SoP, though with a smaller beta value, compared to
NIH-PA Author Manuscript

the Social Anxiety subscale. Both child and parent report on the Physical Symptoms
subscale as well as parent report on the Harm Avoidance subscale demonstrated significant
prediction to the CSR for GAD.

Results from the adolescents exhibited a similar pattern. Both adolescent and parent report
indicated that the Separation/Panic subscale was a significant predictor for the CSR of SAD.
With smaller beta values, both adolescent and parent report on the Social Anxiety subscale
were negative predictors, whereas parent report on the Harm Avoidance subscale was a
positive predictor for the CSR of SAD. Adolescent and parent report on the Social Anxiety
subscale significantly predicted the CSR of SoP, whereas parent report on the Harm
Avoidance subscale negatively significantly predicted the CSR of SoP, with scores on the
Social Anxiety subscale demonstrating the greatest beta value. Both adolescent and parent
reports on the Physical Symptoms subscale demonstrated significant prediction for the CSR
for GAD, whereas parent report on the Social Anxiety subscale negatively predicted the
CSR of GAD.
NIH-PA Author Manuscript

Gains from Multiple Informants

Sequential logistic regression analyses examined the potential gain of adding parent report
to youth report for predicting the presence of SAD, SoP, and GAD using the MASC. In the
first step, scores of youth-reported MASC subscales were entered as predictors to the
presence or absence of an anxiety diagnosis (i.e., SAD, SoP, or GAD). Scores of parent-
reported MASC subscales were entered in the second step. The same steps were repeated
each for SAD, SoP, and GAD. Again, separate analyses were carried out for child and
adolescent groups. The results of the sequential logistic regressions are presented in Table 6.
The overall model fits in testing the prediction of the three anxiety disorders were good (all
Hosmer-Lemeshow, ps>.05).

For the child group, initial modeling when including only child report found that increased
scores on the Separation/Panic subscale and decreased scores on the Social Anxiety subscale
predicted the presence of SAD. The presence of SoP was predicted by both increased scores
on the Social Anxiety subscale and decreased scores on the Separation/Panic subscale based
NIH-PA Author Manuscript

on child report. The presence of GAD was predicted by increased scores on the child-
reported Physical Symptoms subscale.

When adding parent report to test the full model, additional benefits, including consistently
increased percentage of correctly classified cases as well as the explained variance, were
found compared to the initial model that included only child report. Compared to the initial
model, the explained variance (Nagelkerke’s R2) in the full models increased (from .20 to .
36 for SAD, .20 to .31 for SoP, and .06 to .10 for GAD).

The results from the full model sequential regressions suggested that parent report
contributed unique information in the prediction of the presence of SAD, SoP, and GAD. In
the full model, increased scores on the Separation/Panic subscale predicted SAD according

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 10

to parent report when controlling the effect of other MASC subscales as well as child report.
Child-reported Separation/Panic subscale scores continued to be a significant predictor of
SAD when parent report was added to the model, suggesting that child report and parent
NIH-PA Author Manuscript

report on the Separation/Panic subscale each have unique contributions to the prediction of
SAD. Child report on the Social Anxiety Subscale was no longer associated with the
presence of SAD. In the full model, increased child-reported scores on the Physical
Symptoms subscale also significantly predicted SAD. Based on parent report, increased
scores on the Social Anxiety subscale significantly predicted the presence of SoP,
controlling for child report and other MASC subscales. After adding parent report, child
report on the Social Anxiety subscale remained a significant predictor of SoP, suggesting
that child and parent report on the Social Anxiety subscale each made unique contributions
to the prediction of SoP. In the full model, decreased child report on the Separation/Panic
subscale was not associated with SoP. Increased parent-reported scores on the Physical
Symptoms subscale significantly predicted a diagnosis of GAD, controlling for the effect of
child report and all other MASC subscales. After adding parent report to the model, the child
report on the Physical Symptoms subscale remained a significant predictor of GAD,
indicating that parent and child report each contributed unique information.
NIH-PA Author Manuscript

For adolescents, initial modeling including only adolescent report revealed that increased
scores on the Separation/Panic and decreased scores on the Social Anxiety subscales
predicted SAD. SoP was predicted only by increased scores on the Social Anxiety subscale.
The presence of GAD was predicted by increased scores on the Physical Symptoms
subscale.

Similar to the child group, when adding parent report to test the full model with adolescents,
benefits were found. Compared to the initial model, Nagelkerke’s R2 in the full models were
increased from .30 to .46 for SAD, .37 to .54 for SoP, and .10 to .19 for GAD.

In the full model, increased scores on the Separation/Panic subscale predicted SAD
according to parent report when controlling for the effect of other MASC subscales as well
as adolescent report. After adding parent report, adolescent report on the Separation/Panic
subscale was no longer associated with SAD, suggesting that parent report on the
Separation/Panic subscale was a stronger predictor for SAD in adolescents. In the full
model, increased adolescent-report scores on the Physical Symptoms subscale predicted
NIH-PA Author Manuscript

SAD. According to parent report, decreased scores on the Harm Avoidance subscale were
associated with SoP after controlling for adolescent report and other MASC subscales.
However, adolescent report on the Social Anxiety subscale remained a significant predictor
in the full model, indicating that adolescent report contributed unique diagnostic
information. No MASC subscale scores predicted the presence of GAD in the full model for
adolescents.

Discussion
The present results support the utility of the MASC to predict the presence and severity of
particular anxiety disorders in children and in adolescents. Consistent with past research
(Grills-Taquechel, et al., 2008; van Gastel & Ferdinand, 2008; Villabø et al., 2012), the

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 11

Separation/Panic subscale predicted a diagnosis of SAD, and the Social Anxiety subscale
predicted SoP. Furthermore, supporting the findings of Villabo et al (2012), the addition of
the parent-report to youth self-report improved the accuracy of identifying youth with an
NIH-PA Author Manuscript

anxiety disorder. Both the parent- and youth self-report versions contributed valuable
information when predicting SAD, SoP, and GAD. Accordingly, a multi-informant approach
is recommended when assessing for anxiety disorders in youth.

The MASC predicted SAD and SoP in both children and adolescents using the Separation/
Panic and Social Anxiety subscales, respectively. These subscales were useful for both
identifying SAD and SoP and for predicting the severity of the disorders. Higher scores on
the Separation/Panic subscale were associated with greater severity of SAD and higher
scores on the Social Anxiety subscale were associated with greater severity of SoP across
informants. These findings make sense given that many of the MASC subscale items are
consistent with the diagnostic criteria for these anxiety disorders.

The findings for GAD were less clear-cut. In line with past research (Grills-Taquechel et al.,
2008; van Gastel & Ferdinand, 2008; Villabø et al., 2012; Wood et al., 2002), the MASC
subscales were less predictive of the presence and/or severity of GAD. Of the four
NIH-PA Author Manuscript

subscales, the Physical Symptoms subscale performed best by predicting the presence and
severity of GAD according to multiple and logistic regression analyses, followed by the
Harm Avoidance subscale. Unlike SAD and SoP, there is not a single subscale of the MASC
designed to depict the full scope of GAD symptoms. It is possible that combining items
across subscales could improve the prediction of GAD. For example, the Physical
Symptoms and Harm Avoidance subscales may be inconsistently associated with GAD
because they each tap only some of the DSM-IV criteria (e.g. somatic symptoms or worry
symptoms) of GAD. Selecting items from both scales (and perhaps others) may capture the
complete array of symptoms, and improve prediction of GAD.

Comorbidity was high in the present sample. Most youth met criteria for more than one
anxiety disorder, and there was no non-anxiety disorder group. Symptoms of different
anxiety disorders overlap to some extent and may make it difficult to distinguish between
specific disorders. This observation may in part explain why few subscales performed better
than fair in their ability to identify the specific anxiety disorders. Moreover, the ROC
analyses relied on information from only one informant while other findings in the present
NIH-PA Author Manuscript

study and by others (Villabø et al., 2012) suggest that combining parent and youth self-
report increases the MASC’s ability to detect specific anxiety disorders.

The addition of parent report to youth self-report improved the accuracy in identifying youth
anxiety disorders. Consistent with Villabø et al (2012), valuable information was gained
when youth and parent reports were combined. Furthermore, there were instances when
particular informants provided unique information, and slightly different patterns emerged
for younger and older youth. For children up to the age of 12, both children and parents
provided unique information on the Separation/Panic subscale that together yielded a more
accurate identification of SAD than relying on information from one informant. Similarly,
combining information from both children and parents resulted in improved identification of
SoP than child-report alone using the Social Anxiety subscale. Although predicting the

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 12

presence of GAD proved more challenging, both children and parents provided unique
information of physical symptoms that were associated with the presence of GAD. These
findings emphasize the importance of using multiple informants in the assessment of anxiety
NIH-PA Author Manuscript

disorders in youth. Though the same subscales provided the strongest associations to the
specific disorders according to both parents’ and youth’s reports, different patterns of
symptoms seemed to be reported. For adolescents, parent report of Separation/Panic
symptoms was of particular importance when identifying SAD compared to adolescent self-
report, whereas adolescent self-report of social anxiety symptoms was of greater importance
when identifying SoP. It may be that adolescents are more reluctant to report symptoms of
separation anxiety, but such symptoms may be more readily observable to parents, who also
may be more willing to report them. Symptoms relating to feeling embarrassed in front of
peers, on the other hand, are increasingly prevalent in adolescence and commonly tied to
situations outside of the home, making self reports of these symptoms particularly
important.

This study is not without limitations. First, the sample contained youth with principal
anxiety diagnoses. The absence of a non-anxious comparison group did not allow for the
testing of the MASC to differentiate between anxiety disordered and non-disordered youth.
NIH-PA Author Manuscript

However, prior studies have consistently found the MASC to discriminate between anxious
and non-anxious samples (Dierker et al., 2001; van Gastel & Ferdinand, 2008; Villabø et al.,
2012), and this study extended these findings to investigate differentiation among anxiety
disorders in a clinical sample, as well as to examine prediction of disorder severity. Second,
the sample was predominately Caucasian middle-to-upper SES and thus, generalizability of
the results may be limited. Third, the present study did not include the information regarding
which parent(s) completed the MASC, even though past research indicates that parents are
often consistent when providing information at the symptom level, and that the addition of a
father to a mother’s report does not significantly improve accuracy (Villabø et al. 2012).
Future research should include teacher report to determine if the addition improves
diagnostic accuracy. In addition to SAD, SoP, and GAD, Wood et al. (2002) found both
parent- and youth-reported MASC scores to successfully discriminate panic disorder from
other anxiety disorders. Future research can examine the ability of the MASC subscales to
predict other anxiety disorders, such as specific phobias and agoraphobia. Finally, work is
needed to improve the accuracy of the MASC in predicting GAD. Efforts to create a
NIH-PA Author Manuscript

subscale by uniting items relating to physiological and emotional/cognitive symptoms may

be useful, as has been done previously using the Child Behavior Checklist (Achenbach,
1991), to detect anxiety disorders (Kendall, et al., 2007).

For researchers and clinicians alike, the findings support the utility of the MASC to predict
SAD and SoP. Given the ease of administration and minimal resources required, the MASC
is a cost-effective screener for particular anxiety disorders in youth. The MASC is a useful
tool when administered to both parents and youth and the reported data are integrated.
Despite the predictive abilities of the MASC, it is not a substitute for structured or semi-
structured interviews (Silverman & Ollendick, 2005). However, as a screener, the MASC
can be useful to clinicians to determine youth who may need additional services. The MASC
may reduce clinicians’ assessment burden and assure that youth receive services in line with
their needs.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 13

Acknowledgments
This research was supported by NIMH grant (MH063747) awarded to Philip C. Kendall.
NIH-PA Author Manuscript

References
Achenbach, TM. Integrative Guide for the 1991 CBCL/4-18, YSR, and TRF. Burlington, VT: Univ. of
Vermont; 1991.
American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4. Arlington,
VA US: American Psychiatric Publishing, Inc; 1994.
Baldwin JS, Dadds MR. Reliability and validity of parent and child versions of the Multidimensional
Anxiety Scale for Children in community samples. Journal of the American Academy of Child &
Adolescent Psychiatry. 2007; 46(2):252–260. [PubMed: 17242629]
Chavira DA, Stein MB, Bailey K, Stein MT. Comorbidity of generalized social anxiety disorder and
depression in a pediatric primary care sample. Journal of Affective Disorders. 2004; 80:163–171.
[PubMed: 15207929]
Choudhury MS, Pimentel SS, Kendall PC. Childhood Anxiety Disorders: Parent–Child (Dis)
Agreement Using a Structured Interview for the DSM-IV. Journal of the American Academy of
Child & Adolescent Psychiatry. 2003; 42(8):957–964. [PubMed: 12874498]
Cohen, J. Statistical Power Analysis for the Behavioral Sciences. 2. Hillsdale, New Jersey: Erlbaum;
1988.
NIH-PA Author Manuscript

Comer JS, Kendall PC. A Symptom-Level Examination of Parent–Child Agreement in the Diagnosis
of Anxious Youths. Journal of the American Academy of Child & Adolescent Psychiatry. 2004;
43(7):878–886. [PubMed: 15213589]
Compton SN, Walkup JT, Albano AM, Piacentini JC, Birmaher B, Sherrill JT, Ginsburg GS, et al.
Child/Adolescent Anxiety Multimodal Study (CAMS): Rationale, design, and methods. Child and
Adolescent Psychiatry and Mental Health. 2010; 4
Costello EJ, Copeland W, Angold A. Trends in psychopathology across the adolescent years: What
changes when children become adolescents, and when adolescents become adults? Journal of Child
Psychology and Psychiatry. 2011; 52(10):1015–1025. [PubMed: 21815892]
Costello EJ, Egger H, Angold A. The developmental epidemiology of anxiety disorders:
phenomenology, prevalence, and comorbidity. Child & Adolescent Psychiatric Clinic of North
America. 2005; 14(4):631–48. vii.10.1016/j.chc.2005.06.003
Costello, E Jane; Mustillo, S.; Erkanli, A.; Keeler, G.; Angold, A. Prevalence and Development of
Psychiatric Disorders in Childhood and Adolescence. Archive of General Psychiatry. 2003; 60(8):
837–844.10.1001/archpsyc.60.8.837
De Los Reyes A, Thomas SA, Goodman KL, Kundey SM. Principles Underlying the Use of Multiple
Informants’ Reports. Annual Review of Clinical Psychology. 2013; 910.1146/annurev-
clinpsy-050212-185617
NIH-PA Author Manuscript

Dierker LC, Albano AM, Clarke GN, Heimberg RG, Kendall PC, Merikangas KR, Lewinsohn PM, et
al. Screening for anxiety and depression in early adolescence. Journal of the American Academy
of Child & Adolescent Psychiatry. 2001; 40(8):929–936. [PubMed: 11501693]
Grills-Taquechel AE, Ollendick TH, Fisak B. Reexamination of the MASC factor structure and
discriminant ability in a mixed clinical outpatient sample. Depression and Anxiety. 2008; 25(11):
942–950. [PubMed: 18008335]
Hanley, James A.; McNeil, Barbara J. The Meaning and Use of the Area under a Receiver Operating
Characteristic (ROC) Curve. Radiology. 1982; 143(1):29–36. [PubMed: 7063747]
Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, Williamson D, et al. Schedule for
Affective Disorders an Schizophrenia for School-Age Children-Present an Lifetime Version (K-
SADS-PL): Initial Reliability and Validity Data. Journal of the American Academy of Child &
Adolescent Psychiatry. 1997; 36(7):980–988. [PubMed: 9204677]
Kendall PC, Compton SN, Walkup JT, Birmaher B, Albano AM, Sherrill J, et al. Clinical
characteristics of anxiety disordered youth. Journal of Anxiety Disorders. 2010; 24(3):360–365.
doi: S0887-6185(10)00026-5. [PubMed: 20206470]

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 14

Kendall PC, Puliafico AC, Barmish AJ, Choudhury MS, Henin A, Treadwell KS. Assessing anxiety
with the Child Behavior Checklist and the Teacher Report Form. Journal Of Anxiety Disorders.
2007; 21(8):1004–1015.10.1016/j.janxdis.2006.10.012 [PubMed: 17270388]
NIH-PA Author Manuscript

Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime Prevalence and Age-
of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication.
Archives of General Psychiatry. 2005; 62(6):593–602. [PubMed: 15939837]
Langley AK, Bergman RL, McCracken J, Piacentini JC. Impairment in Childhood Anxiety Disorders:
Preliminary Examination of the Child Anxiety Impact Scale-Parent Version. Journal of Child and
Adolescent Psychopharmacology. 2004; 14(1):105–114. [PubMed: 15142397]
March, JS. Multidimensional Anxiety Scale for Children 2nd Edition (MASC 2)™. MultiHealth
Systems; Toronto, CA: 2012.
March JS, Parker JDA, Sullivan K, Stallings P, Conners C. The Multidimensional Anxiety Scale for
Children (MASC): Factor structure, reliability, and validity. Journal of the American Academy of
Child & Adolescent Psychiatry. 1997; 36(4):554–565. [PubMed: 9100431]
March JS, Sullivan K, Parker J. Test–retest reliability of the multidimensional anxiety scale for
children. Journal of Anxiety Disorders. 1999; 13(4):349–358. [PubMed: 10504106]
Merikangas, Kathleen Ries; He, J-ping; Burstein, M.; Swendsen, J.; Avenevoli, S.; Case, B.;
Georgiades, K., et al. Service utilization for lifetime mental disorders in U.S. Adolescents: Results
of the National Comorbidity Survey-Adolescent Supplement (NCSA). Journal of the American
Academy of Child & Adolescent Psychiatry. 2011; 50(1):32–45. [PubMed: 21156268]
O’Neil KA, Puleo CM, Benjamin CL, Podell JL, Kendall PC. Suicidal Ideation in Anxiety-Disordered
NIH-PA Author Manuscript

Youth. Suicide & Life-Threatening Behavior. 2012; 42(3):305–317. [PubMed: 22509976]

Pine DS, Cohen P, Gurley D, Brook J, Ma Y. The risk for early-adulthood anxiety and depressive
disorders in adolescents with anxiety and depressive disorders. Archives of General Psychiatry.
1998; 55(1):56–64. [PubMed: 9435761]
Rynn MA, Barber JP, Khalid-Khan S, Siqueland L, Dembiski M, McCarthy KS, Gallop R. The
psychometric properties of the MASC in a pediatric psychiatric sample. Journal of Anxiety
Disorders. 2006; 20(2):139–157. [PubMed: 16464701]
Safford SM, Kendall PC, Flannery-Schroeder E, Webb A, Sommer H. A longitudinal look at parent-
child diagnostic agreement in youth treated for anxiety disorders. Journal of Clinical Child and
Adolescent Psychology. 2005; 34(4):747–757. [PubMed: 16232071]
Silverman, W.; Albano, AM. The Anxiety Disorders Interview Schedule for DSM-IV: Child and
parent versions. San Antonio, TX: Graywind; 1996.
Silverman WK, Ollendick TH. Evidence-Based Assessment of Anxiety and Its Disorders in Children
and Adolescents. Journal Of Clinical Child And Adolescent Psychology. 2005; 34(3):380–
411.10.1207/s15374424jccp3403_2 [PubMed: 16026211]
Silverman WK, Saavedra LM, Pina AA. Test-retest reliability of anxiety symptoms and diagnoses with
anxiety disorders interview schedule for DSM-IV: Child and parent versions. Journal of the
American Academy of Child and Adolescent Psychiatry. 2001; 40:937–944. [PubMed: 11501694]
van Gastel W, Ferdinand RF. Screening capacity of the Multidimensional Anxiety Scale for Children
NIH-PA Author Manuscript

(MASC) for DSM-IV anxiety disorders. Depression and Anxiety. 2008; 25(12):1046–1052.
[PubMed: 18833579]
Villabø M, Gere M, Torgersen S, March JS, Kendall PC. Diagnostic efficiency of the child and parent
versions of the Multidimensional Anxiety Scale for Children. Journal of Clinical Child and
Adolescent Psychology. 2012; 41(1):75–85. [PubMed: 22233247]
Walkup JT, Albano AM, Piacentini J, Birmaher B, Compton SN, Sherrill JT, et al. Cognitive
behavioral therapy, sertraline, or a combination in childhood anxiety. New England Journal of
Medcine. 2008; 359(26):2753–2766. doi: NEJMoa0804633.
Wood JJ, Piacentini JC, Bergman RL, McCracken J, Barrios V. Concurrent validity of the anxiety
disorders section of the Anxiety Disorders Interview Schedule for DSM-IV: Child and parent
versions. Journal of Clinical Child & Adolescent Psychology. 2002; 31:335–342. [PubMed:
12149971]

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Wei et al. Page 15

Woodward LJ, Fergusson DM. Life Course Outcomes of Young People With Anxiety Disorders in
Adolescence. Journal of the American Academy of Child & Adolescent Psychiatry. 2001; 40(9):
1086–1093. [PubMed: 11556633]
NIH-PA Author Manuscript

Wren FJ, Bridge JA, Birmaher B. Screening for Childhood Anxiety Symptoms in Primary Care:
Integrating Child and Parent Reports. Journal of the American Academy of Child & Adolescent
Psychiatry. 2004; 43(11):1364–1371. [PubMed: 15502595]
NIH-PA Author Manuscript
NIH-PA Author Manuscript

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 1

Mean Scores, Cronbach’s Alpha, and between Groups Differences by Age and Gender for Youth Self- and Parent-Reported MASC Subscales

Informant and Subscales Youth Report Parent Report

Wei et al.

All Gender Subgroups Age Subgroups (years) All Gender Subgroups Age Subgroups (years)

Boys Girls t (478) d 7-12 13-17 t (478) d Boys Girls t (480) d 7-12 13-17 t (478) d
n = 238 n = 242 n = 354 n = 126 n = 239 n = 243 n = 354 n = 126
Physical Symptoms
Cronbach’s α 0.85 0.85 0.84 0.82 0.90 0.80 0.80 0.81 0.80 0.80
M 13.21 13.42 13.00 0.62 0.06 12.67 14.72 -2.64* -0.28 11.12 11.64 10.60 1.86 0.17 10.57 12.66 -3.33*** -0.35

SD 7.49 7.58 7.41 7.11 8.32 6.11 6.15 6.05 5.91 6.42
Harm Avoidance
Cronbach’s α 0.64 0.64 0.64 0.60 0.69 0.72 0.68 0.75 0.72 0.71
M 18.94 18.90 18.98 -0.22 -0.02 19.45 17.52 4.51*** 0.47 18.98 19.28 18.67 1.57 0.14 19.19 18.39 1.81 0.19

SD 4.22 4.15 4.29 4.06 4.35 4.27 4.01 4.49 4.19 4.45
Social Anxiety 0.87
Cronbach’s α 0.86 0.87 0.86 0.84 0.89 0.88 0.87 0.87 0.87 0.86
M 13.83 14.37 13.30 1.68 0.15 12.90 16.44 -4.97*** -0.52 18.00 18.09 17.90 0.35 0.03 17.48 19.44 -3.17** -0.33

SD 7.02 7.08 6.94 6.85 6.86 6.02 5.93 6.12 5.98 5.94
Separation/Panic
Cronbach’s α 0.76 0.74 0.78 0.71 0.79 0.80 0.77 0.82 0.74 0.80
M 11.54 12.17 10.92 2.32* 0.21 12.88 7.77 8.92*** 0.93 14.23 14.62 13.84 1.32 0.12 15.80 9.79 9.92*** 1.03

SD 5.96 5.72 6.13 5.54 5.49 6.41 6.11 6.69 5.65 6.38
Total Score
Cronbach’s α 0.88 0.89 0.87 0.87 0.91 0.87 0.87 0.87 0.87 0.88

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
M 57.52 58.87 56.20 1.68 0.07 57.90 56.44 0.81 0.07 62.32 63.63 61.02 1.85 0.08 63.04 60.28 1.72 0.08
SD 17.47 17.60 17.28 16.87 19.10 15.47 15.28 15.58 15.01 16.60

Note. MASC = Multidimensional Anxiety Scale for Children

*
p < .05, two-tailed.
**
p < .01, two-tailed.
***
p < .001, two-tailed.
Page 16
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 2

Concordance between Informants for Youth Self- and Parent-Reported MASC Subscales

All Children Ages 7 – 12 Adolescents Ages 13 – 17

Wei et al.

N = 380 n = 354 n = 126

Youth/ Parent Child/ Parent Adolescent/ Parent

MASC Subscales r P≤ r P≤ r P≤
Physical Symptoms 0.31 0.001 0.26 0.001 0.38 0.001
Harm Avoidance 0.21 0.001 0.16 0.01 0.27 0.01
Social Anxiety 0.39 0.001 0.36 0.001 0.43 0.001
Separation/Panic 0.57 0.001 0.47 0.001 0.57 0.001
Total Score 0.31 0.001 0.29 0.001 0.35 0.001

Note. MASC = Multidimensional Anxiety Scale for Children

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Page 17
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 3

Mean Scores (Standard Deviations) of the MASC by Diagnostic Category for Youth and Parent Report

Children Ages 7 - 12
Wei et al.

MASC Subscale SAD Non-SAD d SoP Non-SoP d GAD Non-GAD d

n = 227 n = 127 n = 270 n = 84 n = 287 n = 67
M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)
Physical Symptoms
Child 13.34 (7.53)* 11.47 (6.12) 0.27 12.95 (7.01) 11.79 (7.40) 0.16 13.24 (7.06)** 10.22 (6.85) 0.43

Parent 11.04 (6.07)* 9.72 (5.55) 0.22 10.55 (5.73) 10.64 (6.50) -0.02 11.07 (5.94)*** 8.47 (5.34) 0.45

Harm Avoidance
Child 19.42 (4.07) 19.59 (4.05) -0.04 19.37 (4.11) 19.69 (3.89) -0.08 19.70 (4.00)* 18.37 (4.13) 0.33

Parent 19.44 (4.11) 18.72 (4.29) 0.17 19.09 (4.15) 19.50 (4.32) -0.10 19.38 (4.13) 18.37 (4.34) 0.24
Social Anxiety
Child 17.18 (5.89) 18.03 (6.11) -0.14 13.96 (6.86)*** 9.51 (5.64) 0.68 13.20 (6.83) 11.64 (6.87) 0.23

Parent 17.40 (5.91) 18.71 (6.10) -0.22 18.74 (5.35)*** 13.40 (6.09) 0.95 17.68 (6.19) 16.63 (4.90) 0.18

Separation/Panic
Child 14.22 (5.10)*** 10.49 (5.50) 0.71 12.47 (5.21)* 14.19 (6.34) -0.31 13.02 (5.47) 12.30 (5.81) 0.13

Parent 17.74 (4.65)*** 12.30 (5.63) 1.05 15.53 (5.62) 16.68 (5.69) -0.20 15.88 (5.73) 15.44 (5.33) 0.08

Adolescents Ages 13 - 17
MASC Subscale SAD Non-SAD SoP Non-SoP GAD Non-GAD
n = 33 n = 93 d n = 113 n = 13 d n = 102 n = 24 d
M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)
Physical Symptoms

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Adolescent 16.73 (8.43) 14.01 (8.21) 0.35 14.81 (8.44) 14.00 (7.52) 0.10 15.57 (8.17)* 11.13 (8.17) 0.55

Parent 14.61 (5.66)* 11.97 (6.56) 0.42 12.53 (6.38) 13.77 (6.93) -0.19 13.40 (6.22)** 9.50 (6.43) 0.63

Harm Avoidance
Adolescent 18.42 (5.23) 17.19 (3.98) 0.29 17.67 (4.17) 16.15 (5.73) 0.35 17.86 (4.23) 16.04 (4.67) 0.43
Parent 20.45 (3.78)** 17.66 (4.46) 0.65 18.11 (4.40)* 20.85 (4.28) -0.65 18.94 (4.21)** 16.04 (4.76) 0.68

Social Anxiety
Page 18
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Children Ages 7 - 12
MASC Subscale SAD Non-SAD d SoP Non-SoP d GAD Non-GAD d
n = 227 n = 127 n = 270 n = 84 n = 287 n = 67
Wei et al.

M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)

Adolescent 15.15 (8.39) 16.89 (6.22) -0.26 17.34 (6.49)*** 8.62 (4.81) 1.30 16.61 (7.01) 15.71 (6.29) 0.13

Parent 18.91 (5.88) 19.63 (5.99) -0.12 20.13 (5.68)*** 13.46 (4.93) 1.20 19.45 (6.19) 19.42 (4.91) 0.01

Separation/Panic
Adolescent 11.24 (7.00)*** 6.54 (4.25) 0.93 7.73 (5.53) 8.15 (5.34) -0.08 8.07 (5.48) 6.50 (5.45) 0.29

Parent 15.06 (6.44)*** 7.91 (5.23) 1.29 9.78 (6.48) 9.85 (5.70) -0.01 10.20 (6.56) 8.04 (5.34) 0.34

Note: MASC = Multidimensional Anxiety Scale for Children; SAD = separation anxiety disorder; SoP = social phobia; GAD = generalized anxiety disorder.
*
p<.05.
**
p<.01.
***
p<.001.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Page 19
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 4

Results of Receiver Operating Characteristic Analyses Testing Diagnostic Discriminant Validity of the Youth and Parent MASC

Physical Symptoms Harm Avoidance Social Anxiety Separation/Panic

Wei et al.

AUC 95% CI p≤ AUC 95% CI p≤ AUC 95% CI p≤ AUC 95% CI p≤

Children Ages 7 - 12
SAD
Child 0.57 [0.51, 0.63] 0.025 0.49 [0.43, 0.55] 0.759 0.48 [0.42, 0.54] 0.520 0.69 [0.63. 0.75] 0.001
Parent 0.56 [0.50, 0.62] 0.057 0.55 [0.49, 0.62] 0.102 0.45 [0.39, 0.52] 0.136 0.78 [0.71, 0.82] 0.001
SoP
Child 0.55 [0.48, 0.63] 0.150 0.48 [0.41, 0.55] 0.526 0.69 [0.63, 0.75] 0.001 0.42 [0.34, 0.49] 0.022
Parent 0.51 [0.44, 0.59] 0.704 0.47 [0.40, 0.54] 0.437 0.74 [0.68, 0.80] 0.001 0.43 [0.36, 0.50] 0.058
GAD
Child 0.62 [0.55, 0.70] 0.002 0.60 [0.52, 0.68] 0.011 0.56 [0.49, 0.64] 0.107 0.55 [0.48, 0.63] 0.173
Parent 0.63 [0.56, 0.70] 0.001 0.56 [0.49, 0.64] 0.105 0.57 [0.51, 0.64] 0.060 0.53 [0.46, 0.61] 0.421

Adolescents Ages 13 - 17
SAD
Adolescent 0.59 [0.48, 0.71] 0.108 0.60 [0.48, 0.73] 0.086 0.44 [0.31, 0.57] 0.319 0.70 [0.57, 0.82] 0.001
Parent 0.62 [0.52, 0.73] 0.038 0.69 [0.58, 0.79] 0.001 0.46 [0.34, 0.58] 0.488 0.80 [0.71, 0.89] 0.001
SoP
Adolescent 0.53 [0.38, 0.68] 0.739 0.59 [0.40, 0.79] 0.284 0.86 [0.78, 0.93] 0.001 0.46 [0.29, 0.64] 0.662
Parent 0.47 [0.31, 0.64] 0.757 0.34 [0.17, 0.50] 0.052 0.82 [0.72, 0.91] 0.001 0.49 [0.35, 0.64] 0.920
GAD
Adolescent 0.65 [0.53, 0.78] 0.022 0.63 [0.51, 0.75] 0.054 0.54 [0.42, 0.66] 0.559 0.60 [0.48, 0.73] 0.117
Parent 0.67 [0.54, 0.79] 0.012 0.68 [0.57, 0.80] 0.006 0.53 [0.41, 0.64] 0.698 0.59 [0.48, 0.71] 0.159

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Note: MASC = Multidimensional Anxiety Scale for Children; AUC = area under the curve; CI = confidence interval; SAD = separation anxiety disorder; SoP = social phobia; GAD = generalized anxiety
disorder.
Page 20
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 5

Multiple Regressions to Test the Prediction of the Clinical Severity of Anxiety Diagnoses Using MASC subscales

Child/Adolescent Report Parent Report

Wei et al.

Dx MASC Subscales R2 B SE B β R2 B SE B β

Children Ages 7 - 12
SAD 0.15 0.28
Physical Symptoms 0.023 0.019 0.072 0.001 0.020 0.002
Harm Avoidance -0.025 0.029 -0.044 -0.057 0.028 -0.106*
Social Anxiety -0.063 0.019 -0.191*** -0.042 0.019 -0.109*
Separation/Panic 0.160 0.023 0.391*** 0.222 0.021 0.554***
SoP 0.14 0.25
Physical Symptoms -0.004 0.016 -0.016 -0.033 0.017 -0.101
Harm Avoidance -0.005 0.025 -0.011 -0.045 0.024 -0.096
Social Anxiety 0.113 0.017 0.398*** 0.173 0.016 0.532***
Separation/Panic -0.061 0.019 -0.172** -0.024 0.018 -0.069

GAD 0.06 0.06

Physical Symptoms 0.055 0.016 0.209*** 0.064 0.019 0.199***
Harm Avoidance 0.038 0.026 0.081 0.054 0.027 0.120*
Social Anxiety 0.009 0.017 0.031 0.006 0.018 0.020
Separation/Panic -0.006 0.020 -0.017 -0.027 0.020 -0.081

Adolescents Ages 13 - 17
SAD 0.30 0.40
Physical Symptoms 0.019 0.024 0.071 -0.014 0.030 -0.041

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Harm Avoidance -0.003 0.052 -0.005 0.086 0.040 0.168*
Social Anxiety -0.116 0.031 -0.353*** -0.099 0.029 -0.259***
Separation/Panic 0.227 0.039 0.550*** 0.208 0.030 0.587***
SoP 0.28 0.26
Physical Symptoms -0.017 0.019 -0.084 -0.022 0.025 -0.082
Harm Avoidance -0.046 0.041 -0.116 -0.101 0.034 -0.260**
Page 21
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Child/Adolescent Report Parent Report

Dx MASC Subscales R2 B SE B β R2 B SE B β
Social Anxiety 0.158 0.024 0.624*** 0.146 0.024 0.502***
Wei et al.

Separation/Panic -0.048 0.030 -0.153 -0.001 0.025 -0.004

GAD 0.13 0.13
Physical Symptoms 0.075 0.022 0.333*** 0.066 0.030 0.224*
Harm Avoidance 0.013 0.049 0.031 0.068 0.040 0.160
Social Anxiety -0.053 0.029 -0.194 -0.071 0.029 -0.225*
Separation/Panic 0.040 0.036 0.118 0.024 0.030 -0.081

Note: Dx = diagnosis; MASC = Multidimensional Anxiety Scale for Children; SAD = separation anxiety disorder; SoP = social phobia; GAD = generalized anxiety disorder.
*
p<.05.
**
p<.01.
***
p<.001.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Page 22
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Table 6

Sequential Logistic Regression to Test the Prediction of Anxiety Disorders Using Youth and Parent Report of the MASC

Child/Adolescent Parent
Wei et al.

Dx MASC Scale OR 95% CI Wald χ2 OR 95% CI Wald χ2 Full Model χ2 R2

Children Ages 7 - 12
SAD Physical Symptoms 1.04 [0.99, 1.08] 2.69 0.96 [0.91, 1.01] 2.53
Harm Avoidance 0.95 [0.89, 1.01] 2.88 0.94 [0.87, 1.01] 3.32
Social Anxiety 0.93 [0.89, 0.97] 10.75*** 0.98 [0.93, 1.03] 0.73

Separation/Panic 1.18 [1.12, 1.25] 36.47*** 1.25 [1.17, 1.33] 40.79*** 106.34*** 0.36

SoP Physical Symptoms 0.99 [0.95, 1.04] 0.18 0.97 [0.92, 1.02] 1.37
Harm Avoidance 0.96 [0.90, 1.03] 1.25 0.93 [0.86, 1.01] 3.02
Social Anxiety 1.16 [1.10, 1.22] 32.22*** 1.17 [1.10, 1.24] 26.42***
Separation/Panic 0.90 [0.85, 0.95] 14.68*** 1.00 [0.94, 1.06] 0.01 81.99*** 0.31

GAD Physical Symptoms 1.07 [1.02, 1.12] 6.60* 1.08 [1.01, 1.15] 5.80*
Harm Avoidance 1.07 [0.99, 1.14] 3.21 1.03 [0.95, 1.11] 0.44
Social Anxiety 1.00 [0.95, 1.04] 0.03 0.99 [0.94, 1.05] 0.06
Separation/Panic 0.99 [0.93, 1.04] 0.29 0.98 [0.92, 1.05] 0.29 21.87** 0.10

Adolescents Ages 13 - 17
SAD Physical Symptoms 1.04 [0.77, 1.04] 1.10 0.92 [0.82, 1.03] 2.29
Harm Avoidance 0.99 [0.86, 1.14] 0.03 1.13 [0.97, 1.30] 2.53
Social Anxiety 0.88 [0.81, 0.96] 8.87** 0.92 [0.81, 1.04] 1.84

Separation/Panic 1.24 [1.11, 1.39] 13.62*** 1.28 [1.11, 1.47] 11.54*** 47.97*** 0.46

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
SoP Physical Symptoms 0.95 [0.87, 1.09] 1.17 0.94 [0.79, 1.12] 1.37
Harm Avoidance 0.95 [0.79, 1.15] 0.26 0.78 [0.63, 0.98] 3.02*
Social Anxiety 1.33 [1.15, 1.54] 14.64*** 1.15 [0.98, 1.35] 26.42

Separation/Panic 0.92 [0.80, 1.05] 1.53 1.16 [0.94, 1.43] 0.01 37.99*** 0.54

GAD Physical Symptoms 1.07 [1.00, 1.15] 4.03* 1.08 [0.97, 1.13] 1.82

Harm Avoidance 1.09 [0.96, 1.25] 1.69 1.12 [0.91, 1.24] 3.20
Social Anxiety 0.95 [0.87, 1.04] 1.18 0.96 [0.86, 1.07] 0.64
Page 23
NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Child/Adolescent Parent

Dx MASC Scale OR 95% CI Wald χ2 OR 95% CI Wald χ2 Full Model χ2 R2

Separation/Panic 0.99 [0.89, 1.12] 0.00 1.01 [0.90, 1.13] 0.01 15.81* 0.19
Wei et al.

Note: Dx = diagnosis; MASC = Multidimensional Anxiety Scale for Children; CI = confidence interval; SAD = separation anxiety disorder; SoP = social phobia; GAD = generalized anxiety disorder.
*
p<.05.
**
p<.01.
***
p<.001.

J Clin Child Adolesc Psychol. Author manuscript; available in PMC 2015 January 01.
Page 24