Sei sulla pagina 1di 12

J Psychopathol Behav Assess (2009) 31:178–189

DOI 10.1007/s10862-008-9119-8

A Psychometric Analysis of the Child Behavior


Checklist DSM-Oriented Scales
Brad J. Nakamura & Chad Ebesutani & Adam Bernstein &
Bruce F. Chorpita

Published online: 9 December 2008


# Springer Science + Business Media, LLC 2008

Abstract The Child Behavior Checklist for Ages 6–18 2002; Jensen et al. 1999; Manassis et al. 1997). The Child
(CBCL/6-18) possesses newly developed DSM-Oriented Behavior Checklist (CBCL; Achenbach 1991; Achenbach
Scales, constructed through expert clinical judgment to and Rescorla 2001), the parent version of the Youth Self
match selected categories for behavioral/emotional prob- Report (YSR; Achenbach 1991; Achenbach and Rescorla
lems as described in the DSM-IV. The present investigation 2001), is among the most widely used parent-report
examined the basic psychometric properties for all six measures of youth symptoms, assessing a wide range of
DSM-Oriented Scales (i.e., Affective, Anxiety, Somatic, problems. Beginning with its first edition (Achenbach and
Attention-Deficit/Hyperactivity, Oppositional, and Conduct Edelbrock 1983), the CBCL has provided empirically
Scales) in a large clinical sample of children and adoles- derived Syndrome and Competence and Adaptive Scales,
cents (N=673). Findings from the present study provide as well as Internalizing, Externalizing, and Total Scales.
strong evidence for the reliability, as well as convergent and The extensive data behind the CBCL’s standardized scores
discriminative validity, of these scales. It appears that the and clinical cutoffs have allowed for its use in a wide
DSM-Oriented Scales may provide accurate supplementary variety of settings and have aided in current understandings
information that may be considered when formulating of youth psychopathology (e.g., Goodman and Scott 1999).
clinical diagnoses. Within clinical settings, the CBCL has demonstrated
remarkable utility, particularly with respect to being able to
Keywords Child behavior checklist . Diagnostic distinguish between referred and nonreferred populations
and statistical manual . Psychometrics . Test validity . (Achenbach 1991; Chen et al. 1994; Drotar et al. 1995). For
Test reliability example, a recent meta-analytic study found that scores
from the CBCL Internalizing Scale could discriminate
between youths with and without an anxiety disorder, as
The large research literature on psychological measurement well as youths with anxiety disorders and youths with
highlights the value of including multiple informants, externalizing disorders (Seligman et al. 2004). Concerning
particularly parents, in youth assessment (Barbosa et al. the more domain-specific CBCL Syndrome Scales, research
suggests significant and often clinically useful associations
with both broad-based (e.g., anxiety and affective disorders
groups) and specific (e.g., Attention Deficit Disorder and
Conduct Disorder) diagnostic groups (Edelbrock and
B. J. Nakamura (*)
Costello 1988; Kasius et al. 1997; Kazdin and Heidish
Department of Psychology, University of Hawai‘i at Mānoa,
2430 Campus Road, 1984; Eiraldi et al. 2000).
Honolulu, HI 96822, USA Although the CBCL Syndrome Scales have evidenced
e-mail: bradn@hawaii.edu several strengths, their empirical derivation via factor
C. Ebesutani : A. Bernstein : B. F. Chorpita
analytic methods has yielded a long-recognized discordance
University of California, with nosology from the Diagnostic and Statistical Manual
Los Angeles, CA, USA of Mental Disorders (e.g., American Psychiatric Associa-
J Psychopathol Behav Assess (2009) 31:178–189 179

tion 2000).1 This discordance has been supported by represent GAD, SAD and Simple Phobia and an absence
several studies that suggest that the CBCL Syndrome of somatic-related items). Accordingly, Kendall et al. (2007)
Scales yield only modest associations with DSM-IV constructed a modified CBCL-based DSM-Oriented Anxiety
disorders, have limited positive predictive values, and scale with 10 additional items. These authors found that the
seemingly do not map well onto specific diagnoses modifications led to an increased ability for mothers’ reports
(Brunshaw and Szatmari 1988; Jensen et al. 1993; Kasius (but not fathers’ reports) to predict anxiety disorder status
et al. 1997). This lack of concordance may be of concern among adolescent youths when compared with the CBCL’s
given that the DSM-IV classification system forms the basis official DSM-Oriented Anxiety Scale, the Internalizing Total
for most psychopathology research and treatment protocols, Scale, and the Anxious/Depressed Syndrome Scale.
and that mental health services reimbursement eligibility in In summary, although the CBCL DSM-Oriented Scales
America is largely tied to DSM-IV diagnoses (Achenbach have been criticized for not mapping strongly to certain
and Dumenci 2001; Doucette 2002). DSM diagnoses, these scales have nonetheless received
In an effort to provide closer linkage with prevailing some support in initial investigations (e.g., Achenbach et al.
DSM nosology, Achenbach et al. (2003) developed the 2003). Additionally, their congruence with DSM nosology
CBCL DSM-Oriented Scales to supplement the CBCL offers the potential for specific categorical applications that
Syndrome Scales. Whereas the Syndrome Scales were may not be possible with the Syndrome scales. However, the
derived empirically via factor analytic methods, DSM- strengths and weaknesses of their psychometric properties
Oriented Scales were constructed through agreement in are not yet well substantiated in large clinical samples. It is
experts’ ratings of the preexisting items’ consistency with thus useful to further investigate all CBCL DSM-Oriented
DSM-IV diagnostic criteria. Due to their recent construc- Scales for future clinical applications.
tion, research on the CBCL DSM-Oriented scales is The present study evaluated numerous psychometric
limited. Among the investigations conducted to date, properties of all six DSM-Oriented Scales in a large, clinic-
Achenbach et al. (2003) reported that the psychometric referred sample of youth. Four hypotheses were examined.
properties of the DSM-Oriented Scales were generally First, it was predicted that each DSM-Oriented Scale would
similar to those of the Syndrome Scales when comparing evidence an acceptable level of reliability as demonstrated
among the same large national sample of referred and non- through internal consistency analyses. Second, concerning
referred children. For instance, internal consistency and convergent validity, each DSM-Oriented Scale was
test-retest reliability for these scales were found to be good, expected to demonstrate convergence (i.e., significant
with Cronbach Alphas ranging from .75 to .84 and test- positive correlations) with all available construct-consistent
retest coefficients ranging from .78 to .88. instrument scales. Third, with respect to divergent validity,
Other research, however, has provided mixed results, each DSM-Oriented Scale was expected to demonstrate
warranting further investigations. For instance, van Lang et divergence (i.e., lack of a significant correlation) with
al. (2005) examined the CBCL’s child-report counterpart, construct-inconsistent instrument scales. Fourth, concerning
the YSR, and found that a measure of Major Depressive the DSM-Oriented Scales’ discriminative properties, it was
Disorder (MDD) corresponded more closely with the YSR predicted that these scale scores would significantly
DSM-Oriented Affective Problems Scale than with either discriminate between youths with and without the mental
the YSR Anxious/Depressed or Withdrawn/Depressed health diagnosis relevant to each DSM-Oriented Scale.
Syndrome Scale. However, these investigators also found
that measures of Generalized Anxiety Disorder (GAD) and
Separation Anxiety Disorder (SAD) corresponded more Method
closely with the YSR Anxious/Depressed Syndrome Scale
than with the YSR DSM-Oriented Anxiety Problems Scale. Participants
Vreugdenhil et al. (2006) also found low concordance
between the YSR DSM-Oriented scales and DSM-IV The present investigation was conducted across a clinical
DISC-C diagnoses among incarcerated adolescents. Kendall sample of youths, whose caregivers completed either the
et al.( 2007) also argued that the CBCL DSM-Oriented CBCL/4-18 or CBCL/6-18. Participants were selected on
Anxiety Scale could be improved upon due to some the basis of the availability of completed CBCL/4-18 (n=
problematic features (e.g., only six items collectively 555) and CBCL/6-18 (n=118) data from 813 consecutive
referrals made to the Center for Cognitive Behavioral
Therapy (CCBT), at the University of Hawai‘i at Mānoa,
1
It is important to note that no “gold standard” exists for determining
diagnoses. Moreover, the question of whether the described disorders
for a mental health assessment. The CCBT utilized the
are in fact categorical (as in the DSM) or dimensional (as measured by CBCL/4-18 for all intake assessments up until approxi-
CBCL scale scores) is one of active debate (e.g., Haslam, 2003). mately May 2003 and switched to the use of the CBCL/6-
180 J Psychopathol Behav Assess (2009) 31:178–189

18 shortly thereafter. This form changeover occurred after provided following the assessment. Additionally, while the
the authors became aware of the new CBCL/6-18 form. sample includes more youths with principal diagnoses of an
Participants were excluded from analyses if they were externalizing (e.g., Disruptive Behavior and Attention/
missing more than (a) eight items on their entire CBCL, (b) Hyperactivity) than internalizing (e.g., Anxiety and De-
two items from their DSM-Oriented Attention-Deficit/ pression) nature, the proportion of internalizing-type diag-
Hyperactivity Problems Scale (CBCL/4-18 youths would noses exceeds that typically found in youth disorder
have at least two of their five items missing due to version prevalence studies (e.g., Garland et al. 2001). Many of the
differences explained below) or (c) 20% of their item- assessment instruments utilized in the study (see Measures
responses necessary for calculating any other DSM-Oriented below) thus examine internalizing symptoms.
scale. These rationally-derived rules for listwise deletion Numerous analytic strategies were applied to investigate
allowed for maximum use of available data without whether there were significant differences between CBCL/
excessive reliance on mean-substitution procedures (see 4-18 and CBCL/6-18 youths. One, eight, and seven
Procedure below) and are more stringent than missing separate chi-square tests, respectively, examined the rela-
item-level data rules in previous CBCL studies (cf. Galera tionships between CBCL version (CBCL/4-18 and CBCL/
et al. 2005). The mean age of the final combined sample (N= 6-18) and (a) gender, (b) principal diagnosis, and (c) parent-
673) of youths was 12.3 years (SD=3.2; range=4.2 to 19.7), reported child ethnicity (based on major categories in
and the group consisted of 454 boys (67.5%) and 216 girls Table 1). Additionally, a one-way ANOVA was conducted
(32.1%; gender data were missing for three participants). to determine if there were significant age differences
A wide range of ethnicities and DSM-IV (1994) between CBCL/4-18 and CBCL/6-18 participants. All
diagnoses were represented in the present sample (see analyses were performed using a 99.7% confidence interval
Table 1). Notably, 120 participants did not receive a (alpha of .003 after a Bonferroni correction for the 17 tests
diagnosis. However, the absence of a diagnosis does not above performed using a 95% confidence interval). Only
indicate that these youths did not have mental health one of these 17 tests emerged significant. Concerning
concerns or that meaningful recommendations were not principal diagnosis, a one-sample chi-square test for

Table 1 Demographic
information for all study CBCL/4-18 CBCL/6-18 Combined
participants n=555 n=118 N=673

Age:
Minimum 4.2 5.8 4.2
Maximum 19.7 18.0 19.7
Mean (SD) 12.3 (3.1) 12.6 (3.5) 12.3 (3.2)
Gender
Boy 375 79 454
Girl 179 37 216
Missing 1 2 3
Ethnicity
Multiethnic 276 46 322
Asian 85 12 97
White 85 10 95
Pacific Islander 37 7 44
African American 9 2 11
Latino / Hispanic 7 1 8
Native American 0 0 0
Other 45 2 47
Missing 11 38 49
Principal diagnoses
Disruptive behavior 152 50 202
Attention/hyperactivity 124 16 140
Anxiety 112 18 130
Depressive 25 4 29
Adjustment 20 4 24
Youths with co-principal Substance 7 3 10
diagnoses are included in totals Other 56 9 65
for each of their principal None 98 22 120
diagnoses
J Psychopathol Behav Assess (2009) 31:178–189 181

assessing the relationship between CBCL version (CBCL/ ranging from 0 (never true) to 3 (always true). Internal
4-18 versus CBCL/6-18) and whether or not a youth had a consistency and favorable validity have been demonstrated
“Disruptive Behavior Diagnoses” was found significant, χ2 for the AFARS’ three scales (Chorpita et al. 2000a;
(1, N=673)=10.4 p=.001, suggesting a greater proportion Daleiden et al. 2000).
of “Disruptive Behavior Diagnoses” associated with the
CBCL/6-18. All other tests were non-significant, suggest- Child Behavior Checklist for Ages 6–18 (CBCL/6-18;
ing no other systematic differences between the study’s two Achenbach and Rescorla 2001) The 113 items on this
sub-samples. measure are rated as Not True (0), Somewhat or Sometimes
Primary caregivers consisted of 525 biological mothers True (1), or Very True or Often True (2). Validity and
(78.0%), 70 biological fathers (10.4%), 24 grandmothers reliability are excellent, and extensive normative data are
(3.6%), 17 other caregivers (2.5%), 16 foster mothers available for children ranging from 6 to 18. The CBCL/6-
(2.4%), 12 stepfathers (1.8%), 7 grandfathers (1.0%), and 18 is a slightly modified revision of the Child Behavior
2 stepmothers (0.3%). Marital status among primary care- Checklist for Ages 4–18 (CBCL/4-18; Achenbach 1991).
givers was 306 married (45.5%), 191 divorced or separated Changes from the CBCL/4-18 to the CBCL/6-18 include
(28.4%), 104 single (15.5%), and 19 widowed (2.8%) new age norms, the replacement of ineffective items (i.e.,
(marital status data were missing for 53 caregivers). items 2, 4, 5, 28, 78, 99), and the creation of six DSM-
Oriented Scales. The six DSM-Oriented Scales include: (a)
Measures Affective Problems (Dysthymic and Major Depressive
Disorders), (b) Anxiety Problems [Generalized Anxiety
Anxiety Disorders Interview Schedule for DSM-IV, Child Disorder (GAD), Separation Anxiety Disorder (SAD), and
and Parent Versions (ADIS-IV-C/P; Silverman and Albano Specific Phobia (SPEC)], (c) Attention/Deficit/Hyperactivity
1996) The ADIS-IV-C/P is a semi-structured clinical Problems (Hyperactive-Impulsive and Inattentive subtypes),
interview for parents and children ages 7 to 17 that is (d) Conduct Problems [Conduct Disorder (CD)], (e) Oppo-
specifically designed for DSM-IV diagnoses of childhood sitional Defiant Problems [Oppositional Defiant Disorder
anxiety, mood, and behavioral disorders. Assessment (ODD)], and (f) Somatic Problems (Somatization and
involves two interviews, one with the caregiver(s) and one Somatoform Disorders).
with the child. The ADIS-IV-C/P is a revision of the
Anxiety Disorders Interview Schedule for Children (Silverman Children’s Depression Inventory (CDI; Kovacs 1981) The
1991), which has demonstrated satisfactory reliability across a CDI is a 27-item self-report measure designed to assess
range of ages and other parameters (Silverman and Eisen cognitive, behavioral, and affective symptoms of depres-
1992). In the present study, interviews were conducted by Ph. sion. Each item is scored from 0 to 2, and the sum of all
D.-level clinical child psychologists and senior doctoral item scores yields the total CDI score, which ranges from 0
students in clinical psychology who had been trained to to 54, with higher scores indicating more depressive
reliability using the ADIS-IV-C/P. Training first involved symptoms. The CDI has extensive support for its reliability
observation of three interviews conducted by trained inter- and validity (e.g., Saylor et al. 1984).
viewers. The trainee then conducted a series of five inter-
views while being observed by a criterion-trained interviewer. Dimensional Ratings (Chorpita et al. 2000b) Dimensional
An interviewer is considered trained after he/she matches the Ratings, also known as Clinical Severity Ratings, are
experienced interviewer on all clinical diagnoses in three of clinician-reported ratings based on the interference rating
the five interviews. In a subset of 16 randomly selected scale (i.e., a functional impairment rating scale) developed
families in this sample, interrater agreement for principal previously for the ADIS-C/P (Silverman and Nelles 1988).
diagnoses between trainees who administered the ADIS-IV- Dimensional Ratings range from 0 to 8 and are provided for
C/P and trainers who observed the administration was found each ADIS-IV-C/P diagnostic area on the basis of informa-
to be excellent (kappa=.77). tion obtained during each the child and parent interview.
Larger numbers reflect greater stress and impairment, but
Affect and Arousal Scale (AFARS; Chorpita et al. 2000a; not necessarily the presence or absence of a diagnosis (e.g.,
Daleiden et al. 2000) The AFARS is a 27-item self-report it is possible to get a high Conduct Disorder dimensional
questionnaire designed to measure (Cronbach’s alpha rating without receiving this diagnosis because other
coefficients for present sample): positive affect (PA; .83), diagnostic symptom criteria were not met). The presence
negative affect (NA; .81), and physiological hyperarousal of a diagnosis, however, meant that the associated Dimen-
(PH; .86) in children and adolescents. The AFARS requires sional Rating was at least a “4,” indicating clinically
respondents to rate how true each item is with respect to significant impairment in that area. Each child in the
their usual feelings. Items are scored on a 4-point scale present investigation was assigned two different ratings
182 J Psychopathol Behav Assess (2009) 31:178–189

for each diagnostic area, one based on the child report database through the established operating procedures of
alone, and one based on the parent report alone. Dimen- the CCBT clinic, and subsequently extracted electronically
sional ratings have been found to demonstrate good to from the system for analyses.
excellent interrater reliability and convergent validity (Gray Item-level changes between the CBCL/4-18 and CBCL
et al. 2001). The dimensional ratings relevant to the current 6-18 did not affect score-calculations for the DSM-Oriented
study included: (a) Child and Parent (Mean) Anxiety, Anxiety Problems Scale (items 11, 29, 30, 45, 50, 112),
comprised of constituent Separation Anxiety, Social Anx- Somatic Problems Scale (items 56a, 56b, 56c, 56d, 56e,
iety, Specific Phobia, Generalized Anxiety, Panic, Agora- 56f, 56g), or Oppositional Problems Scale (items 3, 22, 23,
phobia, Obsessions/Compulsions, and Posttraumatic Stress 86, 95) because the items comprising these scales were
ratings, (b) Parent Oppositional, and (c) Parent Conduct identical across CBCL forms. Item-level changes did,
dimensional ratings. however, affect (a) one item (item 5) on the DSM-Oriented
Affective Problems Scale (items 5, 14, 18, 24, 35, 52, 54,
Revised Child Anxiety and Depression Scales (RCADS; 76, 77, 91, 100, 102, 103), (b) two items (items 4, 78) on
Chorpita et al. 2000c) The RCADS is a 47-item revision of the DSM-Oriented Attention-Deficit/Hyperactivity Prob-
the Spence Children’s Anxiety Scale (SCAS; Spence 1998). lems Scale (items 4, 8, 10, 41, 78, 93, 104), and (c) one
Children rate the extent to which each item is true of them item (item 28) on the DSM-Oriented Conduct Problems
on a 0 to 3 scale, corresponding to the anchors of “never,” Scale (items 15, 16, 21, 26, 28, 37, 39, 43, 57, 67, 72, 81,
“sometimes,” “often,” and “always.” Factor analytic inves- 82, 90, 97, 101, 106). Achenbach’s ASEBA (2001) manual
tigations suggest the following six subscales (Cronbach’s recommends using raw scores when conducting analyses on
alpha coefficients for present sample): Separation Anxiety the narrow-band Syndrome scales in order to account for
Disorder (SAD; .80), Social Phobia (SOC; .85), General- the full range of variation in these scales. Consistent with
ized Anxiety Disorder (GAD; .83), Obsessive-Compulsive this recommendation, all analyses in the present investiga-
Disorder (OCD; .81), Panic Disorder (PD; .85), and Major tion were conducted on DSM-Oriented raw scale scores.
Depressive Disorder (MDD; .84). These subscales have DSM-Oriented Scale scores unaffected by item-level
demonstrated good factorial validity, internal consistency, changes (Anxiety, Somatic, and Oppositional) between the
one-week test-retest reliability, and convergent and discrim- CBCL/4-18 and CBCL/6-18 versions were calculated by
inant validity (Chorpita, et al. 2000c). summing all relevant items within the scale. Consistent
with Achenbach’s (2001) recommendations, DSM-Oriented
Revised Children’s Manifest Anxiety Scale (RCMAS; Reynolds Scale scores affected by version differences (Affective,
and Richmond 1978) The RCMAS is as 37-item self-report Attention-Deficit/Hyperactivity, and Conduct) were calcu-
measure designed to assess the presence of cognitive, lated by treating changed problem items as missing.
behavioral, or affective symptoms of anxiety and negative Specifically, these scale scores were calculated by summing
affect. Of the 37 items in total, 28 dichotomous (i.e., yes/ the remaining items and multiplying that score by the total
no) items are summed to yield an Anxiety Scale score, number of items divided by the total number of items minus
ranging from 0 to 28, with higher scores representing the number of missing items. Following listwise deletion
greater anxiety. The other 9 items are summed to assess logic above (see "Participants"), scale scores for other
social desirability (i.e., “Lie scale”). Its constituent scales measures were excluded from analyses in a pairwise
include (Cronbach’s alpha coefficients for present sample): fashion if more than 20% of the item-responses comprising
Anxiety (.90), Worry/Oversensitivity (.86), Physiological the scale were missing, with missing items handled with the
(.69), Social Concerns/Concentration (.72), and Lie (.75) same procedure described immediately above.
components. This instrument has been one of the most
widely used instruments for assessing childhood anxiety, Analytic Strategy
and has been demonstrated to be reliable across different
gender, racial, and age groups in children from age 7 to 17 To test the first hypothesis, Cronbach’s alpha coefficients
(Reynolds and Paget 1983). were calculated for each DSM-Oriented Scale. Hypothesis
two was evaluated by calculating zero-order bivariate
correlations between each DSM-Oriented Scale and several
Procedure construct-consistent instruments scales. See bolded text in
Tables 3 and 4 for all a priori hypothesized DSM-Oriented
All participants and their legal guardian(s) underwent Scale by measure convergent (i.e., significant and positive)
standardized Institutional Review Board-approved notice correlations. Towards the goal of evaluating divergent
of privacy and consent procedures prior to any data validity, the third hypothesis involved calculating zero-
collection. All data were gathered and entered into a central order bivariate correlations between each scale and several
J Psychopathol Behav Assess (2009) 31:178–189 183

construct-inconsistent instrument scales. See italicized text performing follow-up tests to evaluate pairwise mean
in Tables 3 and 4 for all a priori hypothesized DSM- differences among the three types of groups above, Tukey’s
Oriented Scale by measure divergent (i.e., non-significant) HSD tests were used when Levene’s test of equality of error
correlations (cf. Chorpita et al. 2005). Note that Child variances was found non-significant. If equal variances
Oppositional and Delinquent Dimensional Ratings are not could not be assumed, Dunnett’s C test was utilized for post
obtained during ADIS-IV-C/P administration and, there- hoc comparisons. Given the preliminary nature of the
fore, were not available for analysis. When a significant and literature on the DSM-Oriented Scales’ diagnostic discrim-
positive relationship unexpectedly did emerge, Fisher’s z- inative properties, youths with the specific disorder under
tests (Meng et al. 1992) were calculated to determine if the examination were hypothesized to score higher than youths
DSM-Oriented Scale under examination was correlated with no related diagnosis, but not necessarily higher than
significantly more with construct-consistent validity indices youths with a disorder from the same category. In order to
than with the construct-inconsistent scale. control the experiment-wise error rate, all analyses were
In order to evaluate each scales’ discriminative proper- performed using a moderately conservative 99% confidence
ties, the fourth hypothesis involved performing one-way interval (alpha level of .01).
ANOVAs to evaluate the relationship between the presence
or absence of various DSM-IV (1994) diagnoses (anywhere
in a child’s diagnostic profile; e.g., primary, secondary, or Results
tertiary, etc.) and relevant DSM-Oriented Scale scores.
Capitalizing on the diagnostic breadth afforded by this Reliability
sample, the over-arching strategy within each DSM-
Oriented Scale by diagnosis analysis first involved using Cronbach’s alpha coefficients are presented in Table 2. All
an ANOVA to detect significant differences between three reliability coefficients were favorable, ranging from .71
groups of children: (a) youths with a specific disorder under (Somatic Problems items) to .89 (Conduct Problems items)
examination, (b) youths with a disorder from the same in the present sample.
diagnostic category as the specific disorder under exami-
nation, but without that specific disorder under examina- Convergent Validity
tion, and (c) youths that did not have the specific disorder
under examination or any disorder in that same diagnostic As predicted, the DSM-Oriented Affective Problems Scale
category. For example, concerning the DSM-Oriented correlated significantly and positively with all convergent
Affective Problems Scale, it was hypothesized that an validity criterion measures (see Table 3 for these results).
ANOVA would detect differences between (a) youths with The DSM-Oriented Anxiety Problems Scale also correlated
Dysthymic Disorder (DD), (b) youths without DD but with significantly and positively with all but one (i.e., RCMAS
another type depressive disorder [i.e., any type of Major Total Anxiety Scale) of its convergent validity criterion
Depressive Disorder (MDD) or Depressive Disorder Not measures. The DSM-Oriented Somatic Problems Scale
Otherwise Specified], and (c) and youths with no depres- correlated significantly and positively with the AFARS-
sive disorder. See Table 5 for all a priori hypothesized PH Scale, but did not significantly correlate with the
DSM-Oriented Scale by disorder analyses. RCMAS Physiological Scale. Correlations of the CBCL
The strength of the initial ANOVA was assessed by η2, DSM-Oriented Attention-Deficit/Hyperactivity, Opposi-
with .01, .06, and .14 and interpreted as small, medium, and tional, and Conduct Problems scales with convergent
large effect sizes, respectively (Green and Salkind 2005). In validity criterion measures are displayed in Table 4. As

Table 2 Alpha coefficients for CBCL/4-18 and CBCL/6-18 DSM-oriented scale scores

CBCL/4-18 sample CBCL 6-18 sample Combined sample

Scale Number of items n Alpha n Alpha n Alpha

Conduct problemsa 17 – – 109 .89 – –


ADHD problemsa 7 – – 116 .85 – –
Oppositional problems 5 542 .79 116 .86 658 .81
Affective problemsa 13 – – 110 .81 – –
Anxiety problems 6 533 .77 116 .79 649 .77
Somatic problems 7 537 .77 115 .71 652 .76
a
As these scales contained new items on the CBCL/6-18, alpha coefficients for these scales could not be computed for those youths with a CBCL/4-18
184 J Psychopathol Behav Assess (2009) 31:178–189

Table 3 Correlation matrix for child behavior checklist DSM-oriented scales and selected internalizing measures

DSM-oriented scales

Affect Anxiety Somatic ADHD ODD CD

Depression measures
Parent depression dimensional rating (n=542) .54** .30** .28** .02 .17** .19**
Child depression dimensional rating (n=525) .41** .17** .20** −.01 .07 .10
RCADS, depression scale (n=619) .33** .16** .25** .07 .09 −.01
Children’s depression inventory, total scale (n=423) .34** .09 .17** .07 .12* .11
Anxiety measures
Parent anxiety (mean) dimensional rating (n=536) .42** .59** .32** .05 .13* .23*
Child anxiety (mean) dimensional rating (n=518) .26** .27** .23** −.01 −.07 −.02
RCADS, total anxiety scale (n = 610) .20** .18** .22** .10 .01 −.05
RCMAS, total Scale (n = 167) .16 .05 .16 −.08 −.08 −.16
Arousal measures
AFARS, physiological hyperarousal scale (n=578) .15** .09 .15** .11* .12* .06
RCMAS, physiological scale (n = 167) .17 .05 .16 .03 .03 −.08

Bold print highlights convergent indices. Italicized print highlights divergent indices
AFARS = Affect and Arousal Scale for Children, RCADS = Revised Children’s Anxiety and Depression Scales, RCMAS = Revised Children’s
Manifest Anxiety Scale
*p<0.01; **p<0.001

predicted, all three of these DSM-Oriented Scales signifi- indices above), the DSM-Oriented Affective Problems Scale
cantly and positively correlated with their respective was found to be significantly more correlated with (a) the
convergent validity criterion measures. Parent Depression Dimensional Rating, z(540) = 7.49,
p<.001, (b) the Child Depression Dimensional Rating, z
Divergent Validity (518)=4.64, p<.001, (c) the RCADS MDD Scale, z(508)=
3.11, p=.002, and (d) the CDI Total Scale, z(326)=3.19,
DSM-Oriented Affective, Anxiety, and Somatic Problems p=.001, than with the Parent Delinquent Dimensional
scales Contrary to expectation, the DSM-Oriented Affec- Rating. Both ANOVA analyses for DD, F(2, 670)=37.9,
tive Problems Scale correlated significantly and positively p<.001, η2 =.10, and MDD, F(2, 670)=42.1, p<.001,
with Parent Oppositional and Delinquent Dimensional η2 =.11, were significant. Follow-up tests evaluating pairwise
Ratings (see Table 4). For comparative purposes, two sets differences supported the fourth hypothesis and pointed to
of pairwise follow-up Fisher’s z-tests (Meng et al. 1992) higher scores for youths with DD and MDD over youths
compared these two unexpected correlations against the with no depressive disorders (see Table 5 for means, standard
correlations between the DSM-Oriented Affective Problems deviations, and all pairwise comparison results).
Scale and its four convergent validity indices (see bolded Results concerning the DSM-Oriented Anxiety Problems
text in Table 3). Concerning the first set of four pairwise Scale were mostly supported (see Table 4). There were no
follow-up tests (i.e., the correlation between the DSM- significant relationships between this scale and one of two
Oriented Affective Problems Scale and the Parent Opposi- divergent validity criteria. Unexpectedly, this scale signif-
tional Dimensional Rating versus the correlation between icantly and positively related to the Parent Oppositional
the DSM-Oriented Affective Problems Scale and (a) the Dimensional Rating (see Table 4). Follow-up Fisher’s z-
Parent Depression Dimensional Rating, (b) Child Depres- tests (Meng et al. 1992) revealed that the DSM-Oriented
sion Dimensional Rating, (c) RCADS MDD Scale, and (d) Anxiety Problems Scale was significantly more correlated
CDI Total Scale), the DSM-Oriented Affective Problems with only one of its four convergent indices (see bolded text
Scale was found to be significantly more correlated with the in Table 3). Specifically, the DSM-Oriented Anxiety
Parent Depression Dimensional Rating than with the Parent Problems Scale was significantly more correlated with the
Oppositional Dimensional Rating, z(579)=4.88, p<.001. Parent Anxiety (Mean) Dimensional Rating, z(536)=8.92,
No other correlations, however, were significantly different. p<.001, than with the Parent Oppositional Dimensional
Regarding the second set of four pairwise follow-up tests Rating. Concerning the fourth hypothesis, all overall DSM-
(i.e., the correlation between the DSM-Oriented Affective Oriented Anxiety Scale ANOVAs were found significant:
Problems Scale and the Parent Delinquent Dimensional (a) SAD, F(2, 670)=59.0, p<.001, η2 =.15, (b) Panic
Rating versus the correlation between the DSM-Oriented Disorder (any type) or Agoraphobia without a History of
Affective Problems Scale and the same four convergent Panic (PDA), F(2, 670)=59.7, p<.001, η2 =.15, (c) SOC,
J Psychopathol Behav Assess (2009) 31:178–189 185

Table 4 Correlation matrix for child behavior checklist DSM-oriented scales and selected externalizing measures

DSM-oriented scales

Affect Anxiety Somatic ADHD ODD CD

Inattention/hyperactivity measures
Parent inattention dimensional rating (n=545) .12* .05 .04 .56** .34** .27**
Child inattention dimensional rating (n=528) .09 .05 .02 .23** .13* .14**
Parent hyperactivity dimensional rating (n=545) .08 .09 .03 .62** .38** .34**
Child hyperactivity dimensional rating (n=529) .05 .05 .00 .31** .20** .20**
Disruptive behavior measures
Parent oppositional dimensional rating (n=547) .31** .16** .08 .43** .67** .64**
Parent delinquent dimensional rating (n=544) .16** −.03 −.09 .21** .37** .65**

Bold print highlights convergent indices. Italicized print highlights divergent indices
*p<0.01; **p<0.001

F(2, 670)=57.7, p<.001, η2 =.15, (d) SPEC, F(2, 670)= Divergent validity correlation results for the DSM-
59.3, p<.001, η2 =.15, (e) OCD, F(2, 670)=62.8, p<.001, Oriented Oppositional Problems Scale are displayed in
η2 =.16, (f) GAD, F(2, 670)=58.8, p<.001, η2 =.15, and (g) Table 3. Predictions were somewhat supported and no
Posttraumatic Stress Disorder (PTSD), F(2, 670=57.7, significant relationship emerged for five of eight examined
p<.001, η2 =.15. As seen in Table 5, all subsequent relationships. For the three unexpected findings, three sets
pairwise test results supported the fourth hypothesis and of follow-up Fisher’s z-tests (Meng et al. 1992) compared
indicated that youths with the specific disorder under the three unexpected correlations against the correlation
examination had significantly higher scores than youths between the DSM-Oriented Oppositional Problems Scale
with no anxiety disorder. and its convergent validity index (i.e., the Parent Opposi-
Predictions on the divergent validity relationships be- tional Dimensional Rating). These follow-up analyses
tween the DSM-Oriented Somatic Problems Scale were revealed that the DSM-Oriented Oppositional Problems
uniformly supported (see Table 4). Regarding the fourth Scale was significantly more correlated with the Parent
hypothesis, all overall ANOVAs were significant: (a) MDD, Oppositional Dimensional rating than with the (a) Parent
F(2, 670)=28.8 p<.001, η2 =.08, (b) PDA, F(2, 670)=28.1, Depression Dimensional Rating, z(541)=10.7, p<.001, (b)
p<.001, η2 =.08, and (c) SAD, F(2, 670)=30.8, p<.001, Parent Anxiety (Mean) Dimensional Rating, z(536)=11.6,
η2 =.08. Follow-up pairwise results partially supported p<.001, and (c) CDI Total Scale score z(327)=9.11,
predictions (see Table 5) in that youths with MDD and p<.001. One-way ANOVA results, F(2, 670)=87.2, p<
PDA scored higher than youths without other somatically .001, η2 =.21, were consistent with the fourth hypothesis
associated diagnoses. However, youths with SAD did not and children with ODD also evidenced higher DSM-
score higher than youths without other somatically associ- Oriented Oppositional Problems Scale scores than youths
ated diagnoses. without a disruptive behavior disorder.
Divergent validity results for the DSM-Oriented Conduct
DSM-Oriented Attention-Deficit/Hyperactivity, Oppositional, Problems scale almost mirrored those demonstrated by the
and Conduct Problems scales As displayed in Table 3, DSM-Oriented Oppositional Problems scale (see Table 3).
predictions for the DSM-Oriented Attention-Deficit/ Specifically, the DSM-Oriented Conduct Problems scale
Hyperactivity Problems Scale were supported, and this scale demonstrated no significant correlations for six of eight
did not correlate significantly with any anxiety or depression examined pairwise relationships. However, follow-up hy-
scale scores. The overall ANOVA examining mean differ- potheses were again supported and Fisher’s z-tests (Meng et
ences between youths with Attention-Deficit/Hyperactivity al. 1992) revealed that this DSM-Oriented Scale was
Disorder, Combined Type (ADHD-C), Attention-Deficit/ significantly more correlated with the Parent Delinquent
Hyperactivity Disorder, Primarily Hyperactive/Impulsive Dimensional rating than with the (a) Parent Depression
Type (ADHD-PH), Attention-Deficit/Hyperactivity Disorder, Dimensional Rating, z(540)=9.74, p<.001 and (b) Parent
Primarily Inattentive Type (ADHD-PI), and youths without Anxiety (Mean) Dimensional Rating, z(535)=10.3, p<.001.
ADHD was found significant, F(3, 669)=35.5, p<.001, One-way ANOVA results were hypothesis-consistent,
η2 =.14. The fourth hypothesis was supported and youths F(2, 670)=172.7, p<.001, η2 =.34, and children with CD
with any type of ADHD scored significantly higher on this evidenced higher DSM-Oriented Conduct Problems scale
scale than youths without this disorder. scores than youths without a disruptive behavior disorder.
186 J Psychopathol Behav Assess (2009) 31:178–189

Table 5 DSM-oriented means DSM-oriented scale by disorder analyses n Mean (SD)


and (standard deviations) for
children with and without se- Affective problems
lected diagnoses anywhere in a
Dysthymic disorder 10 9.1 (5.9)
their diagnostic profile1 a
Major depressive or DEP NOS disorder without dysthymic disorder 37 10.5 (5.2)
b
No dysthymic, major depressive, or DEP NOS disorder 626 4.8 (4.1)
a
Major depressive disorder 34 11.3 (5.4)
b
Dysthymic or DEP NOS disorder without major depressive disorder 13 7.4 (3.8)
b
No dysthymic, major depressive, or DEP NOS disorder 626 4.8 (4.1)
Anxiety problems
a
Separation anxiety disorder 34 5.6 (3.3)
a
Anxiety disorder without separation anxiety disorder 169 4.8 (3.1)
b
No anxiety disorder 470 2.6 (2.5)
a
Panic disorder 9 6.7 (2.5)
a
Anxiety disorder without panic disorder 190 4.9 (3.2)
b
No anxiety disorder 470 2.6 (2.5)
a
Social phobia 96 5.1 (3.1)
a
Anxiety disorder without social phobia 107 4.9 (3.2)
b
No anxiety disorder 470 2.6 (2.5)
a
Specific phobia 40 5.7 (3.3)
a
Anxiety disorder without specific phobia 163 4.8 (3.1)
b
No anxiety disorder 470 2.6 (2.5)
a
Obsessive compulsive disorder 27 6.4 (2.9)
a
Anxiety disorder without obsessive compulsive disorder 176 4.8 (3.1)
b
No anxiety disorder 470 2.6 (2.5)
a
Generalized anxiety disorder 39 5.6 (3.0)
a
Anxiety disorder without generalized anxiety disorder 164 4.8 (3.2)
b
No anxiety disorder 470 2.6 (2.5)
a
Posttraumatic stress disorder 27 4.7 (3.1)
a
Anxiety disorder without posttraumatic stress disorder 176 5.0 (3.2)
DBD NOS = Disruptive b
Behavior Disorder Not Other- No anxiety disorder 470 2.6 (2.5)
wise Specified, DEP NOS = Somatic problems
a
Depressive Disorder Not Major depressive disorder 34 4.8 (3.6)
ab
Otherwise Specified, Panic Panic or separation anxiety disorder without major depressive disorder 36 3.5 (3.2)
b
Disorder = Panic Disorder with No major depressive, panic, or separation anxiety disorder 603 1.9 (2.3)
or without Agoraphobia. Given ab
Separation anxiety disorder 34 3.5 (3.7)
the absence of youths with Panic or major depressive disorder without separation anxiety disorder 36 4.7 (3.2) a

Somatization and Somatoform No major depressive, panic, or separation anxiety disorder 603 1.9 (2.3) b
Disorders in the current clinical a
Panic disorder 9 6.3 (2.5)
sample and Chorpita et al.,’s a
Major depressive or separation anxiety disorder without panic disorder 61 3.8 (3.5)
(2000c) findings on somatic b
arousal in selected internalizing No major depressive, panic, or separation anxiety disorder 603 1.9 (2.3)
disorders, analyses for the Attention-deficit/hyperactivity problems
a
Somatic Problems scale above Attention-deficit/hyperactivity disorder, combined type 84 9.1 (3.3)
ab
examined youths with Major Attention-deficit/hyperactivity disorder, hyperactive type 10 8.7 (2.2)
b
Depressive Disorder, Panic Attention-deficit/hyperactivity disorder, inattentive type 97 6.4 (3.2)
Disorder and/or Separation c
No attention-deficit/hyperactivity disorder 482 5.1 (3.6)
Anxiety Disorder Oppositional problems
1
All diagnoses, whether princi- Oppositional defiant disorder 158 6.3 (2.3) a
pal or non-principal, for which a
Conduct disorder or DBD NOS without oppositional defiant disorder 112 5.7 (2.8)
children met full DSM-criteria b
were included in their diagnostic No oppositional defiant, conduct or DBD NOS disorder 403 3.5 (2.4)
profiles Conduct problems
a
abc
Differing letter superscripts Conduct disorder 77 12.8 (6.7)
b
indicate a significant pairwise Oppositional defiant disorder or DBD NOS without conduct disorder 193 7.0 (4.9)
c
mean differences at 99% No oppositional defiant, conduct or DBD NOS disorder 403 3.4 (3.4)
confidence interval
J Psychopathol Behav Assess (2009) 31:178–189 187

Discussion report) while diagnoses in the present study were based on


the ADIS-IV-C/P as well as overall clinician impressions.
Findings from this study provide support for the reliability Accordingly, the differences between samples and mea-
and validity of the expert judgment-derived DSM-Oriented surement approaches may account in part for these
Scales in a large clinical sample of youth. Internal seemingly competing findings.
consistency was favorable and convergent validity hypotheses Although the results of the present study are promising
were almost uniformly supported across all scales. Divergent with respect to the psychometric properties of the CBCL
validity results were also very favorable, though slightly more DSM Oriented Scales, a few caveats are in order. First, in
mixed than the convergent validity findings. order to maximize the size of the present sample, analyses
Several points warrant discussion. First, contrary to were performed across a pooled sample of participants
prediction, significant relationships were found between the completing either the CBCL/4-18 or CBCL/6-18. Instru-
DSM-Oriented Affective Problems scale and oppositional ment variance may have been decreased had all youths
and conduct indices. Although potentially surprising, these filled out the CBCL/6-18. This was not possible, however,
findings are somewhat consistent with the literature. Namely, given the timing with which the research clinic adopted the
depressive feelings in children can present as irritability CBCL/6-18 over the CBCL/4-18. However, it should again
(American Psychiatric Association 2000) and previous be underscored that DSM-Oriented Anxiety, Somatic, or
research has highlighted a moderate, but significant, rela- Oppositional Problems Scale items and calculations were
tionship between oppositional/rule-breaking behavior and identical across CBCL versions. Issues with missing items,
feelings of depression (Goodman and Scott 1999; Loeber et therefore, only applied to the DSM-Oriented Affective,
al. 2004). In some ways, this relationship can be seen as Oppositional, and Conduct scales, with the authors closely
corroborated within this investigation in that significant and adhering to CBCL/6-18 manual directions (Achenbach,
positive relationships were also found between DSM- 2001) for calculating DSM-Oriented Scales scores for
Oriented Oppositional and Conduct Problems Scales and CBCL/4-18 instruments. Concerning potential differences
depressive indices. in the current study between CBCL/4-18 and CBCL/6-18
Second, concerns have previously been raised regarding youth on demographic data, only one of seventeen
the ability of the DSM-Oriented Scales to adequately measure comparisons pointed to statistically significant differences
various DSM-Oriented constructs, particularly given that the between these groups. Given the substantial size of
CBCL DSM-Oriented Scales contain relatively few items and the CBCL/4-18 sample, noteworthy of mention is that the
do not contain items comprising all DSM symptom criteria authors also performed all study analyses with only the
(e.g., Kendall et al. 2007). Despite these concerns, findings of CBCL/4-18 sample for comparative purposes against those
the present study indicated that all CBCL DSM-Oriented of the combined sample. Presenting these comparative
Scales were able to discriminate youths with and without analyses is outside of the scope of this investigation.
relevant DSM diagnoses. This finding runs counter to prior However, it should be mentioned that all but two of the
research that suggested poor discriminative properties for the results presented above were either replicated or in some
DSM-Oriented Anxiety Problems Scale (Kendall et al. 2007; instances found stronger (i.e., in terms of direction and
van Lang et al. 2005). Specifically concerning this scale’s statistical significance) for all analyses performed with the
performance in this study, it is also worth noting that this CBCL/4-18 sample. Using only CBCL/4-18 youths, youths
scale could discriminate in predicted directions between with PTSD did not score higher than youths without PTSD
children with and without all major types of anxiety on the DSM-Oriented Anxiety Problems Scale and youths
disorders, despite being constructed only around conceptions with ADHD-PI did not score higher than youths without
of GAD, SAD, and SPEC. With constituent items such as, ADHD-PI on the DSM-Oriented Oriented Attention-Deficit/
“Too fearful or anxious” and “Worries”, it seems that this Hyperactivity Problems Scale. Given this study’s mixed
scale may not reflect specific diagnoses but rather anxiety sample, it will be important for future studies examining
disorders in general. psychometric properties of the DSM-Oriented Scales to
Third, these overall findings are somewhat surprising collect data from large clinical samples with only the CBCL/
given that Vreugdenhil et al. (2006) found low concordance 6-18 version.
between the YSR DSM-Oriented Scales and DSM DISC-C Another limitation is that only the CBCL, and not the
diagnoses. However, it should be noted that Vreugdenhil et Teacher Report Form (TRF) or Youth Self-Report (YSR),
al. examined YSR DSM-Oriented Scales (child report) with was examined in this study. Consequently, the reliability
incarcerated youths, whereas the present study examined and validity findings for the DSM-Oriented Scales in the
CBCL DSM-Oriented Scales (parent report) with an present study generalize only to the CBCL. Psychometric
outpatient sample. In addition, diagnoses in Vreugdenhil properties of the TRF and YSR DSM-Oriented scales in
et al.’s investigation were based on the DISC-C (child large clinical samples thus remain open to future inquiry.
188 J Psychopathol Behav Assess (2009) 31:178–189

Research following this study could build upon this attention-deficit hyperactivity disorder: a receiver-operating
characteristic analysis. Journal of Consulting and Clinical
investigation’s limitations for furthering empirical scrutiny
Psychology, 62, 1017–1025.
of the psychometric properties of the CBCL’s DSM- Chorpita, B. F., Daleiden, E. L., Moffitt, C., Yim, L., & Umemoto, L. A.
Oriented Scales. With this first validation study providing (2000a). Assessment of tripartite factors of emotion in children
the necessary foundation for subsequent psychometric and adolescents I: Structural validity and normative data of an
affect and arousal scale. Journal of Psychopathology and
research, forthcoming investigations could perform com-
Behavioral Assessment, 22, 141–160.
parative analyses between DSM-Oriented and Syndrome Chorpita, B. F., Plummer, C. M., & Moffitt, C. E. (2000b). Relations
Scales. For example, receiver operating characteristics of tripartite dimensions of emotion to childhood anxiety and
curve methodology could be used for comparing these mood disorders. Journal of abnormal child psychology, 28(3),
299–310.
scales’ discriminative diagnostic abilities and choosing Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E.
optimal cut-points for making diagnostic decisions. In the (2000c). Assessment of symptoms of DSM-IV anxiety and
meantime, the strong performance of the DSM scales depression in children: a revised child anxiety and depression
suggests warranted supplemental usage with clinical pop- scale. Behavior Research and Therapy, 38, 835–855.
Chorpita, B. F., Moffitt, C., & Gray, J. (2005). Psychometric
ulations of youth.
properties of the revised child anxiety and depression scale
Despite these limitations and indications for future in a clinical sample. Behaviour Research and Therapy, 43,
research, the present study is the first systematic investiga- 309–322.
tion of the CBCL DSM-Oriented Scales’ psychometric Daleiden, E., Chorpita, B. F., & Lu, W. (2000). Assessment of tripartite
factors of emotion in children and adolescents II: concurrent validity
properties in a large clinical sample of youth. Findings lend
of the affect and arousal scales for children. Journal of Psychopa-
empirical support to the basic psychometric properties of thology and Behavioral Assessment, 22, 161–182.
the clinical judgment-derived DSM-Oriented Scales. Drotar, D., Stein, R. E. K., & Perrin, E. C. (1995). Methodological
Results from the present investigation indicate acceptable issues in using the child behavior checklist and its related
instruments in clinical child psychology research. Journal of
internal consistency for these scales as well as excellent Clinical Child Psychology. Special Issue: Methodological Issues
convergent and divergent validity. In conclusion, it appears in Clinical Child Psychology Research, 24(2), 184–192.
that the CBCL DSM-Oriented Scales may provide accurate Doucette, A. (2002). Child and adolescent diagnosis: the need for a
supplementary information that may be considered when model-based approach. In L. E. Beutler, & M. L. Malik (Eds.),
Rethinking the DSM. A psychological perspective (pp. 201–220).
formulating clinical diagnoses.
Washington, DC: American Psychological Association.
Edelbrock, C., & Costello, A. J. (1988). Convergence between
statistically derived behavior problem syndromes and child
psychiatric diagnoses. Journal of Abnormal Child Psychology,
References 16, 219–231.
Eiraldi, R. B., Power, T. J., Karustis, J. L., & Goldstein, S. (2000).
Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/4-18, Assessing ADHD and comorbid disorders in children: the child
YSR and TRF profiles. Burlington: Department of Psychiatry, behavior checklist and the Devereaux scales of mental disorders.
University of Vermont. Journal of Clinical Child Psychology, 29, 3–16.
Achenbach, T. M., & Dumenci, L. (2001). Advances in empirically Galera, C., Fombonne, E., Chastang, J. F., & Bouvard, M. (2005).
based assessment: revised cross-informant syndromes and new Childhood hyperactivity-inattention symptoms and smoking in
DSM-oriented scales for the CBCL, YSR, and TRF: Comment adolescence. Drug and Alcohol Dependence, 78, 101–108.
on Lengua, Sadowski, Friedrich, and Fisher (2001). Garland, A. F., Hough, R. L., McCabe, K. M., Yeh, M., Wood, P. A.,
Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the child & Aarons, G. A. (2001). Prevalence of psychiatric disorders in
behavior checklist and revised behavior profile. Burlington: youths across five sectors of care. Journal of the American
University of Vermont Department of Psychiatry. Academy of Child & Adolescent Psychiatry, 40(4), 409–418.
Achenbach, T. M., & Rescorla, L. A. (2001). The manual for the Goodman, R., & Scott, S. (1999). Comparing the strengths and
ASEBA school-age forms & profiles. Burlington, VT: University difficulties questionnaire and the child behavior checklist: is small
of Vermont, Research Center for Children, Youth, and Families. beautiful? Journal of Abnormal Child Psychology, 27(1), 17–24.
Achenbach, T. M., Dumenci, L., & Rescorla, L. A. (2003). DSM- Gray, J. A., Francis, S. E., & Chorpita, B. F. (2001). Measurement of
oriented and empirically based approaches to constructing scales dimensions of child psychopathology: The reliability and validity
from the same item pools. Journal of Clinical Child and of clinical severity ratings. Poster presented at the annual
Adolescent Psychology, 32, 328–340. meeting of the Association for Advancement of Behavior
American Psychiatric Association (2000). Diagnostic and statistical Therapy, Philadelphia, PA.
manual of mental disorders (4th ed.). Washington, DC: American Green, S. B., & Salkind, N. J. (2005). Using SPSS for windows and
Psychiatric Association. Macintosh: Analyzing and understanding data (4th ed.). New
Barbosa, J., Tannock, R., & Manassis, K. (2002). Measuring anxiety: Jersey: Pearson Education Inc.
parent-child reporting differences in clinical samples. Depression Haslam, N. (2003). Categorical versus dimensional models of mental
and Anxiety, 15, 61–65. disorder: the taxometric evidence. Australian and New Zealand
Brunshaw, J. M., & Szatmari, P. (1988). The agreement between Journal of Psychiatry, 37(6), 696–704.
behavior checklists and structured psychiatric interviews for Jensen, P. S., Saltzberg, A. D., Richters, J., Watanabe, H. K., & Roper,
children. Canadian Journal of Psychiatry, 33, 474–481. M. (1993). Scales, diagnoses, and child psychopathology. I.
Chen, W. J., Faraone, S., Biederman, J., & Tsuang, M. T. (1994). CBCL and DISC relationships. Journal of the American
Diagnostic accuracy of the children behavior checklist scales for Academy of Child and Adolescent Psychiatry, 32, 397–406.
J Psychopathol Behav Assess (2009) 31:178–189 189

Jensen, P. S., Rubio-Stipec, M., Canino, G., Bird, H. R., Dulcan, M. K., et al. Saylor, C. F., Finch, A. J., Spirito, A., & Bennett, B. (1984). The
(1999). Parent and child contributions to diagnosis of mental disorder: children’s depression inventory: A systematic evaluation of
are both informants always necessary? Journal of the American psychometric properties. Journal of Consulting and Clinical
Academy of Child and Adolescent Psychiatry, 38, 1569–1579. Psychology, 52(6), 955–967.
Kasius, M. C., Ferdinand, R. F., van den Berg, H., & Verhulst, F. C. Seligman, L. D., Ollendick, T. H., Langley, A. K., & Baldacci, H. B.
(1997). Associations between different diagnostic approaches for (2004). The utility of measures of child and adolescent anxiety: a
child and adolescent psychopathology. Journal of Child Psychology meta-analytic review of the revised children’s manifest anxiety
& Psychiatry, 38, 625–632. scale, the state-trait anxiety inventory for children, and the child
Kazdin, A. E., & Heidish, I. (1984). Convergence of clinically derived behavior checklist. Journal of Clinical Child and Adolescent
diagnoses and parent checklists among inpatient children. Psychology, 33, 557–565.
Journal of Abnormal Child Psychology, 12(3), 421–435. Silverman, W. K. (1991). The anxiety disorders interview schedule:
Kendall, P., Puliafico, A., Barmish, A., Choudhury, M., Henin, A., et Child and parent versions. Albany: Graywind Publications.
al. (2007). Assessing anxiety with the child behavior checklist Silverman, W. K., & Albano, A. M. (1996). Anxiety disorders
and the teacher report form. Journal of Anxiety Disorders, 21(8), interview schedule for DSM-IV, child and parent versions. San
1004–1015. Antonio: Psychological Corporation.
Kovacs, M. (1981). Rating scales to assess depression in school-aged Silverman, W. K., & Eisen, A. R. (1992). Age differences in the
children. Acta Paedopsychiatrica: International Journal of Child reliability of parent and child reports of child anxious symptom-
& Adolescent Psychiatry, 46(5–6), 305–315. atology using a structured interview. Journal of the American
Loeber, R., Russo, M., Stouthamer-Loeber, M., & Lahey, B. (2004). Academy of Child & Adolescent Psychiatry, 31(1), 117–124.
Internalizing problems and their relation to the development of Silverman, W. K., & Nelles, W. B. (1988). The anxiety disorders
disruptive behaviors in adolescence. Journal of Research on interview schedule for children. Journal of the American
Adolescence, 4(4), 615–637. Academy of Child & Adolescent Psychiatry, 27(6), 772–778.
Manassis, K., Mendlowitz, S., & Menna, R. (1997). Child and parent Spence, S. H. (1998). A measure of anxiety symptoms among
reports of childhood anxiety: differences in coping styles. children. Behaviour Research and Therapy, 36, 545–566.
Depression and Anxiety, 6, 62–69. van Lang, N., Ferdinand, R., Oldehinkel, A., Ormel, J., & Verhulst, F.
Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated (2005). Concurrent validity of the DSM-IV scales affective
correlation coefficients. Psychological Bulletin, 111, 172–175. problems and anxiety problems of the youth self-report. Behaviour
Reynolds, C. R., & Paget, K. D. (1983) National normative and Research and Therapy, 43(11), 1485–1494.
reliability data for the revised children’s manifest anxiety scale. Vreugdenhil, C., van den Brink, W., Ferdinand, R., Wouters, L., &
School Psychology Review, 12, 324–336. Doreleijers, T. (2006). The ability of YSR scales to predict
Reynolds, C. R., & Richmond, B. O. (1978). What I think and feel: a DSM/DISC-C psychiatric disorders among incarcerated male
revised measure of children’s manifest anxiety. Journal of adolescents. European Child & Adolescent Psychiatry, 15(2),
Abnormal Child Psychology, 6, 271–280. 88–96.

Potrebbero piacerti anche