The purpose of this research was to determine whether
student academic achievement in college computer science
programs in the Republic of China (ROC) could be predicted
by factors reported to be effective in US studies. The
relationship between these factors and course performance in
computer science programs was examined. Gender differences
were also interrogated.
Sophomore, junior, and senior students enrolled in five
universities offering computer science programs in the ROC
constituted the population. A researcher-designed questionnaire
was used to collect background information. Validity
and reliability issues were addressed by the conduct of
validity assessment, questionnaire pilot testing, and interviews
with selected pilot test subjects. Scores from the
College Entrance Examination (CEE) and college computer
science courses were accessed through university registrar's
offices. A total of 940 questionnaires were collected,
representing more than 81% of the population.
From data analysis, the predictive powers of CEE test
scores in relation to subsequent college performance appeared
to be limited. The CEE math component was negatively
correlated to performance in college computer science
programs. The positive relation of math ability to academic
achievement in complete computer science programs was
confirmed. High school overall achievement as well as math
course averages were identified as effective performance
predictors for college computer science programs. Prior
computer experience showed no conclusive relationship to
subsequent performance in college computer science courses.
The close relationship between performance in beginning
computer science courses and performance in complete computer
science programs was validated. Significant linear
prediction models with limited predictive powers (R2 ranged
from 0.19 to 0.30) were generated for overall performance,
but not for introductory computer science course performance.
Model predictive powers were significantly improved
(R2 range from 0.59 to 0.63) when performance in introductory
computer science courses was included in the models.
Significant gender differences were not found for CEE performance,
prior computer experience, and prediction models.
However, female subjects outperformed male counterparts in
course performance at both the high school and college
levels. / Graduation date: 1996
| Identifer | oai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/34439 |
| Date | 05 April 1996 |
| Creators | Fan, Tai-Sheng |
| Contributors | Niess, Margaret L. |
| Source Sets | Oregon State University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0012 seconds