Gender Differences in Learning Mathematics in Hong Kong: PISA 2003 Study

Many western studies have investigated gender differences in learning Mathematics; however, inconclusive results were obtained. In Hong Kong, research on gender differences is scarce but most scholars have reported that boys are better at learning Mathematics when compared with girls. The present study aims to explore gender differences in learning Mathematics using data from PISA 2003 by adopting self-regulated learning theory. Significant differences in effects of gender on Mathematics achievements were found in the domain of space and shape, change and relationships, and quantity by regression. All the results were in favour of boys significantly except in the domain of uncertainty.;To further investigate the gender effect on Mathematics achievements, three sets of concepts generated from self-regulated learning theory were input into the causal models as mediating variables. The first set of variables was the personal variables, including intrinsic motivation, instrumental motivation, self-efficacy, self-concept, and anxiety. They were added to the causal models to explore the gender effect. The results showed that all the values of direct effects of gender were positive and were statistically significant except in the domain of quantity. All the values of indirect effects of gender were negative and were statistically significant. Three behavioral variables, including control strategies, elaboration, and memorization were added to the causal models. The results showed that all the values of direct effects of gender on the achievements were negative and are statistically significant except in the domain of uncertainty and all the values of indirect effects of gender on the achievements were positive but none of the them was statistically significant. Two environmental variables, namely competitive learning preference and cooperative learning preference; were added to the causal models for exploring the gender effects. The results showed that the values of direct effects of gender were negative in all other three domains except in the domain of uncertainty and the results were statistically significant in the domain of space and shape and the domain of quantity. All the values of indirect effects of gender were negative but only in the domain of space and shape, the result was statistically significant.;Lastly, all the variables were incorporated in one causal model to verify the relative effects among the three sets of variables in the self-regulated learning theory. The results showed that there were no consistent results in direct effect of gender on the four domains. The value of direct effect of gender was negative and was statistically significant in the domain of quantity, which means that the direct effect of gender was significantly in favor of males in the domain of quantity. But, the direct effect of gender was positive and was statistically significant in the domain of uncertainty, it showed that it was significantly in favor of females in the domain of uncertainty; while no significant direct effect of gender in the remaining two domains. However, consistent results were obtained in indirect effect across the four domains. The values of indirect effects of gender on the achievements were all negative and were statistically significant in all four Mathematics domains. Therefore, there is no evidence that Mathematics achievements have been influenced directly by gender but differences existed in the learning styles, preferences and processes. In fact, all these factors are changeable. Effort can be put to make the change and also improve students' learning. Through this study, unidirectional relationships among the personal factors, behavioral factors and environmental factors in theory of self-regulated learning were supported by empirical evidence.