學刊論文
The Investigation of a Lisrel Achievement Model

中華心理學刊 民 80,33 卷,59-75
Chinese Journal of Psychology 1991, Vol.33, 59-75


Huey-Wen Chou(College of Education University of Illinois at Urbana-Champaign)

 

Abstract

The purpose of the present study was to examine the causal effects of selected variables on learning performance. Specifically, this study investigated the relative importance of two types of time variables, that is, homework time and the ratio of time spent in learning (TSL) to time needed for learning (TTL). A total of 180 seventh graders from a junior high school in Taiwan, were used as subjects. A teaching unit involving a mathematics learning task was conducted to estimate TSL and TTL. Students' personal information was collected from their mathematics teachers, the school counselor, and a questionnaire.
The linear structural relations (LISREL) technique was employed as the main analytical tool to evaluate a proposed achievement model with gender, numerical ability, prior knowledge, time ratio, and homework time as antecedent variables.
A significant difference in achievement for boys and girls was found; that is, girls consistently performed better than boys in all achievement variables. Since there was no gender difference in numerical ability, what caused the achievement differences brought to our attention.
Regarding the relative importance of two types of time variables, it was found that the time ratio was a much better predictor of achievement in overall and boys groups whereas tutorial homework time was a better predictor of girls' achievement. This finding may be explained by differences in study patterns for boys and girls.
After covariance structure analyses were performed for separate gender groups, it was found that there were large gender differences in structural relationships. That is, girls benefited more from tutorial homework time whereas boys benefited more from total homework time. In other words, tutorial and total homework time had differential effects on boys' and girls' achievement and boys and girls did have different
study patterns. In addition to this, tutorial homework time also served as a powerful mediating variable to compensate for the lower direct effects of numerical ability on achievement for girls.
The largest gender differences came from the effect of numerical ability on achievement, 0.517 for girls and 0.192 for boys. The above results suggested that part of the gender differences in achievement may be explained by the large direct effects of tutorial homework time on girls' achievement.

 

登入
會員登入
更新驗證碼