Test Reviews 399

Bracken, B. A., & Keith, L. K. (2004). Clinical Assessment of Behavior. Lutz, FL:
Psychological Assessment Resources.
DOI: 10.1177/0734282906291449

The Clinical Assessment of Behavior (CAB) is designed to assess both adaptive and
problematic behaviors of children and adolescents from age 2 to 18 years. It can be indi-
vidually or group administered, measures behaviors in different contexts, and includes both
parent and teacher forms. The test was developed to be consistent with current diagnostic
criteria including the Diagnostic and Statistical Manual of Mental Disorders–Fourth
Edition–Text Revision (DSM-IV-TR; American Psychiatric Association, 2001) and Indivi-
duals With Disabilities Education Act Amendments of 1997, as well as including current
sociopolitical concerns such as school bullying. To increase utility, the authors focused on
developing a “user-friendly,” culturally sensitive, and concise measure that requires mini-
mal reading ability. It is intended for use in clinical, educational, and research settings. Its
clinical uses include assisting with assessment, diagnosis, screening, and treatment. Its edu-
cational use is unique in that it can help to identify educationally relevant abilities such as
gifted and talented. In research, it can be used to substantiate diagnostic information, to pre-
dict outcomes, to evaluate programs, and to conduct intervention follow-up.

General Description
The CAB was normed on children from a wide range of geographic and racial/ethnic
backgrounds. Test authors recommend that individuals without formal graduate training in
assessment and psychology administer the test under supervision. Interpretation of test
scores, however, should be conducted by professionals with graduate training in psychol-
ogy and assessment.
The CAB materials consist of the Professional Manual, computer scoring system
(required for scoring as items cannot be hand-scored), parent rating form (70 items),
extended parent rating form (170 items), and teacher rating form (70 items). All items are
rated on a Likert-type scale from 1 (always or very frequently) to 5 (never), and some are
reverse-scored. Items are scored to derive a total Behavioral Index score, scale scores, clus-
ter scores, raw scores, standardized T scores, percentile ranks, 90% confidence intervals,
Clinical Risk or Adaptive Weakness classifications, and graphed profiles. The parent and
teacher rating forms measure the same student behaviors, and the extended parent rating
forms measure more serious and less frequent behaviors (e.g., hallucinations). The authors
recommend that 10 to 15 minutes are required to complete the parent and teacher forms.
About 40 minutes are required to complete the extended parent rating form. It is preferable
that the examiner be present to answer questions the respondent may have, look over items
missed, and help record demographic information requested on the first page of the rating
forms. Once the responses are recorded on the rating forms, they are entered manually into
the computer scoring system. A psychology assistant, teacher, clerical staff, or anyone with
400 Journal of Psychoeducational Assessment

data entry skills can input these data. Help screens on the computer scoring system are
available for entering data, generating reports, and managing client files. The score report
does not include narrative descriptions of scores, nor is there an option to produce a parent-
friendly report. The score report does provide graphic representations of T scores and per-
centile scores. Although the graphs are more than adequate for a trained evaluator, the
graphs could be more effective if they included clear demarcations between scale and clus-
ter scores as well as between clinical and adaptive scores. Scores on the following scales
are then automatically computed:

Clinical Scale
Internalizing
Externalizing
Critical Behaviors (Parent Extended only)
Adaptive Scale
Social Skills
Competence
Adaptive Behaviors (Parent Extended only)
CAB Behavioral Index
Clinical Cluster
Anxiety
Depression
Anger
Aggression
Bullying
Conduct Problems
Attention Deficit/Hyperactivity
Autistic Spectrum Behaviors
Learning Disability
Mental Retardation
Adaptive Cluster
Executive Function
Gifted and Talented
Veracity Scale

On the Social Skills, Competence, Adaptive Behaviors (Parent Extended version only),
Executive Function, and Gifted and Talented clusters, high scores indicate adaptive weak-
ness, which can be confusing for the test user. A score on the Veracity Scale is also calcu-
lated to measure accuracy in ratings. Test users should use this score to consider the rater
validity when interpreting the other scale scores. The validity score is determined by the
number of T scores greater than 70 or less than 30. For students who exhibit extreme behav-
iors, their high or low test scores may indeed be valid, so the validity score should be care-
fully considered when determining if test results are valid. In addition to standardized
scores, the scoring system presents qualitative classifications of T scores. Classifications
range from normal to very significant.
 Test Reviews 401

Technical Adequacy
Item development. Items were developed from published research on child and adolescent
development, existing instruments, diagnostic criteria used in the DSM-IV-TR, current politi-
cal interests, and colleague input. These items were then conceptually organized into the six
main scales and reduced based on parent consistency ratings obtained from a pilot study.
Items were then assigned to scales through trial and error to obtain high internal consistency
coefficients for the six main scales, and confirmatory factor analysis was then conducted to
obtain the structure of the scale. The remaining items were then judged according to whether
teachers would be aware and able to rate those behaviors. Items that did not meet these crite-
ria were dropped from the teacher form, and a matching parent form was created. Those addi-
tional items were retained in the parent extended rating form.

Standardization. The CAB was standardized on 2,114 parents of children ages 2 to 18

years and 1,689 teachers of students ages 5 to 18 years. These samples were recruited from
preschools, private and public schools, and youth and community organizations from 17
states across the United States. Also, education and school psychology professionals who
had purchased products from the publisher in the previous 2 years were contacted and
asked to administer the CAB. Part of the normative sample was drawn from an Internet-
based survey research company to obtain parent and teacher participants. These parents
were selected according to the child’s age and gender, race/ethnicity, parental education,
and geographic region to represent the population of U.S. children (U.S. Bureau of the
Census, 2001). Although the four major demographic regions of the United States (i.e.,
Midwest, Northeast, South, and West) are represented, the norm group may consist of
adults more educated than is true in the U.S. population. The manual provides age, gender,
race/ethnic, education, and region characteristics in several tables. Also, the percentile con-
version tables in the appendix of the manual are specified by gender, age, and reporting
form (parent extended, parent, and teacher). These tables show the distribution of stan-
dardized T scores, and the distributions of scores for each scale appear consistent with
research on those behaviors.

Reliability. Three forms of reliability were considered in test development. Internal con-
sistency measured by Cronbach’s alpha is high on all report forms. It ranges from .91 to .98
across scales on the extended parent rating form. On the parent rating form, the alpha coef-
ficients range from .88 to .97, and on the teacher rating form they range from .92 to .99.
Several tables show the reliability estimates according to demographic characteristics of the
sample. In general, they are high but are somewhat lower for the youngest children (age
2-6 years), which is consistent with other behavioral measures of children this age. Given
the high reliability coefficients, the standard error of measurement is low (ranging from 1.4
to 3.46 T-score points). A second form of reliability, test-retest reliability, is reported in the
manual for both scale and cluster scores. Parent ratings during a period of 11 to 33 days
range from .77 to .94 (highest reliability for the total scale score) on the parent extended
scale and from .80 to .93 (highest reliability for the scale score and Gifted and Talented clus-
ter score) on the parent scale. During a period of 7 to 36 days, teacher test-retest reliability
402 Journal of Psychoeducational Assessment

across scales ranged from .89 to .95 (highest reliability for the total scale score). Interrater
reliability is the third form of reliability reported in the manual. Consistency between teacher
and parent ratings was moderate, ranging from .44 to .58 (highest reliability for the
Competence cluster score). The authors suggest that children are rated across different con-
texts and thus are likely to exhibit different behaviors. Interrater reliability between parents
was higher, however, ranging from .41 to .90 (highest reliability for the Learning Disability
cluster score) on the extended parent scale and from .64 to .87 (highest reliability for the
Learning Disability cluster score) on the parent scale.

Validity. As with reliability, various forms of validity are reported. Content validity is indi-
cated by obtaining items representing the universe of content of the construct of interest.
According to the manual, an extensive review of journal articles, textbooks, reference books,
the DSM-IV-TR, and existing behavior scales was conducted to derive a universe of content,
from which items were sampled for the CAB. The structure of the scale is supported by the
high correlation of cluster-to-scale scores (e.g., anxiety and depression correlate more highly
with the internalizing, rather than the externalizing, scale score). Factor analysis results also
provide evidence in support of the structure of the CAB. Principal components analyses with
equimax rotation yielded a six-factor solution that corresponds with the six scales of the CAB
extended parent rating form. Also, four factor solutions were derived for both the parent and
teacher rating forms that match the four main scales of those forms. These results provide evi-
dence of validity; however, additional statistics from the factor analysis such as number of
iterations, percentage of explained variance, and eigenvalues were not reported. In terms of
criterion-related validity, the CAB parent and teacher scores were compared with the parent
and teacher scores on the Behavior Assessment System for Children (Reynolds & Kamphaus,
2004) and Devereux scales of mental disorders (Naglieri, LeBuffe, & Pfeiffer, 1994).
Corresponding scales are highly correlated, and all correlations are displayed in tables.
Several clinical studies were also conducted, and there is evidence that children with specific
disorders receive clinically significant cutoff scores on the corresponding scales. However,
diagnostic criteria used to categorize the disorders of children were not specified, and the spe-
cific ages of children were not reported.

Conclusion
Overall, the manual is user-friendly and detailed. The test materials are easily adminis-
tered, and scoring is simplified by the use of computer software. The items are crisp and
concise. Although there are gender differences in some scales, these results are consistent
with research. The standardized sample appears representative of the U.S. population but
may consist of more highly educated people than in the general population. Reliability and
validity evidence are strong and support the use of the CAB as a measure of adaptive and
problem behaviors. Although the test developers suggest that the CAB can be used to pre-
dict behavioral outcomes, there is no empirical evidence of such studies using the CAB
reported in the manual. Thus, it is recommended that the CAB be used to measure current
behavioral functioning, rather than predict future functioning. Also, the clinical studies
 Test Reviews 403

reported in the manual do not specify the exact age of the children or how they were diag-
nosed. As with all rating scales and as indicated in the CAB score report, no decisions
should be made based solely on the CAB results, and results should be integrated with other
sources of information for decision making. Nevertheless, this scale can be used as part of
a psychological assessment, and it includes many of the same and some additional scales
from the Behavioral Assessment for Children, Second Edition (Reynolds & Kamphaus,
2004) and the Child Behavior Checklist (Achenbach & Rescorla, 2001). Thus, the CAB can
be used as an alternative to, or in addition to, these other behavioral scales to substantiate
clinical decision making and provide additional information (e.g., autism and gifted func-
tioning) not provided in these two other scales. Another advantage over these other tests is
that the CAB offers both a shorter and an extended version to parents. Its main strength is
that it is a short, easily administered tool that can provide information in many behavioral
domains that, as part of an assessment battery, can inform treatment and educational deci-
sions, as well as provide data for research purposes.

Tanya N. Beran
University of Calgary, Division of Applied Psychology

References
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms & profiles. Burlington:
University of Vermont, Research Center for Children, Youth, & Families.
American Psychiatric Association. (2001). Diagnostic and statistical manual of mental disorders (4th ed., text
revision). Washington, DC: Author.
Individuals With Disabilities Education Act Amendments of 1997, Pub. L. No. 103-218 (GPO 1997).
Naglieri, J. A., LeBuffe, P. A., & Pfeiffer, S. I. (1994). Devereux scales of mental disorder. San Antonio, TX:
Psychological Corporation.
Reynolds, C. R., & Kamphaus, R. W. (2004). BASC-2 Behavior Assessment System for Children, second edi-
tion manual. Circle Pines, MN: American Guidance Service.
U.S. Bureau of the Census. (2001). Current population survey, March 2001 [Data file]. Washington, DC: U.S.
Department of Commerce.