| Summary: | 碩士 === 國立臺灣大學 === 流行病學研究所 === 85 === This study investigated the behavior problems of junior high
school students in Taipei using the Chinese version of the Child
Behavior Checklist (CBCL). Three schools participating in
this study were each randomly selected from three strata on
the basis of the education level of the city. Classes were then
systematically sampled from each school and the students of the
whole class were invited to participate in the study. Parent
form of CBCL for all subjects (N=854) were completed by
students'' parents or parent surrogates on the basis of the
child''s behavior in the past six months. One seventh of the
subjects (N=162) had their class teacher completing the
Teacher''s Report Form (TRF) regarding the student''s
behavior in the past two months. The response rate was 0.88
for the CBCL and 1.0 for the TRF. Test-retest of the CBCL
(N=164) and the TRF (N=53) were performed on the students of two
schools, respectively, one month later.
Cronbach''s alpha values of CBCL internalizing, externalizing and
total problem were 0.80, 0.83 and 0.90, respectively. Such
values on TRF were 0.70, 0.91 and 0.88. Test-retest
reliability using ICCR (Intraclass Correlation
Coefficient Reliability) for one month period on CBCL
internalizing, externalizing and total problem were 0.68, 0.70
and 0.72; while on TRF were 0.60, 0.88 and 0.78, respectively.
Agreements between CBCL and TRFs'' internalizing, externalizing
and total problem using ICCR were 0.08, 0.45 and 0.24,
respectively. For the analysis of CBCL, we found that
the mean of girls'' internalizing scores was higher than
that of boys'', and the means of externalizing scores and
total problem scores were higher if parents'' education was
lower. No differences were found between genders on
competence scales. Low grade and high parents'' education seemed
to have higher competence scores. Comparing with American
norm aged 12- 18, we found that there were no significant
differences between our and American samples by t-test. On
clinical borderline area, the mean of girls''
internalizing scores was higher than that of boys'' and the
means of boys'' externalizing and total problem scores were
higher than that of girls''. Factor analysis of original items
extracted 8 syndromes and 9 syndromes extracted when adding in
native items. Native syndromes were simular to those extracted
by Achenbach, especially the first 7 factors with greater
variance. For TRF anaysis, the mean scores of
internalizing and total problem were higher and the mean
scores of competence scales were lower if parents''
education was lower. No significant differences between
American norm sample and this sample on competence and
problem scores by t-test. Clinical borderline area indicated
that the mean scores of internalizing and competence
scales were higher among girls than boys, and the mean
score of externalizing was higher among boys than
girls.
|
|---|
