t‘sycholoxy
in
/he
Schools
Volume
32,
Ju/y
1995
WRITTEN VERSUS ORAL ADMINISTRATION OF A SOCIAL-EMOTIONAL
SELF-REPORT TEST FOR CHILDREN:
DOES METHOD OF EXECUTION MAKE A DIFFERENCE?
AMY
S. WALTERS
AND
KENNETH W. MERRELL
Utah State Universily
This study examined administration method (standard written administration vs. oral
administration by an examiner) as a variable in influencing children’s self-report test
scores. Subjects included
139
students in grades
3-6,
randomly assigned to one
or
the other administration condition. Subjects completed the Internalizing Disorders
Evaluation Scale for Children (IDESC) according to the assigned administration
method. Internal consistency estimates of each group were essentially similar. Mean
IDESC scores
of
the two groups did not differ significantly from either a statistical
or
practical standpoint, based on 1-test and effect size calculations. Results suggest
that method
of
administration did not affect test performance. Implications for child
assessment and future research are discussed.
During the past several years, there has been a marked increase in clinical use,
research, and professional acceptance of self-report assessment instruments for children
(Flanery, 1990). Prior to the mid-l970s, there was a substantial negative bias among
behaviorally oriented psychologists against the use
of
self-report measures with children.
This view seemed to exist as an outgrowth of research indicating that self-report data
often did not coincide with behavioral observations (Finch
&
Rogers, 1984). More re-
cently, the importance of assessing children’s own perceptions has been emphasized in
the school and child clinical psychology literature (e.g., LaGreca, 1990). Specifically,
most current child assessment paradigms consider a multimethod assessment process
(which includes child self-report data along with other data sources) to be a best prac-
tice in assessment design (LaGreca, 1990). A recent national survey of the test usage
among practicing school psychologists indicated that child self-report measures were
among the most widely used social-emotional-behavioral assessment methods (Stinnett,
Havey,
&
Oehler-Stinnet, 1994).
The majority of self-report measures for children are oriented toward social-
emotional assessment. Based on our review of various book chapters and journal articles
on self-report assessment with children, the most common constructs appear to be in-
ternalizing, self-related problems such as depression and anxiety, locus of control, and
self-concept. Although the availability of child self-report assessment instruments does
not approach the availability of adult measures, several widely known instruments are
available in each
of
the above named construct domains.
As the use and acceptability of child self-report assessment have increased,
a
number
of measurement issues have emerged. Although many of these measurement issues have
been investigated empirically in detail, many others have not. One of the measurement
issues that has not been sufficiently investigated involves method of test administration.
Most children’s self-report measures use
a
standard administration method in which the
child is given a list of items to read and is then asked to respond using a rating or checklist
format. However, the test manuals for many self-report measures recommend deviating
from the standard administration method when subjects have reading problems or are
Requests for reprints should be sent to Kenneth
W.
Merrell, Department of Psychology, Utah State Univer-
sity, Logan, UT 84322-2810.
186