| This dissertation examines the role that family and school social capital play in explaining math score differences within Canada, Japan, and the United States, and cross-nationally amongst youths in these three nations. Internationally, Japanese students are known for their high academic achievement. Canadian students tend to follow closely behind their Japanese peers, though U.S. students continually rank at or below the international mean. Such disparities have prompted researchers to examine why these achievement gaps occur. Typically, the cultural hypothesis has been used to explain these differences. Although useful in explaining U.S.-Japanese differences, this hypothesis comes into question when Canadian youths are considered, as Canadian students perform more similarly to Japanese youths, although they share a similar---yet distinct---culture with their U.S. neighbours. In this dissertation social capital theory is used to try to explain national and cross-national math score differences. This theory asserts that social capital in the family and school, and the interaction between these two groups, plays a pivotal role in understanding individual achievement disparities. Traditionally the theory has focused on family-school, nationally based explanations of achievement disparities. This dissertation tests the effect of social capital in the family and school, and the interaction between these two spheres in each country. In addition, differences in the level and effect of social capital on youths' math scores are also examined. The Canadian, Japanese and U.S. 1999 Programme for International Student Assessment data, which sampled 15-year-old youths, were analyzed. After considering a series of family and school variables, social capital was found to influence youths' math scores in each country, though this effect was dependent on the specific measure of social capital that was being assessed. To varying degrees, the interaction of social capital in the family and school influenced youths' math scores in each country. Cross-nationally, the level of social capital in the family and school tended to differ, however the difference that these effects had on youths' math scores was minimal. Although the theory could explain math score disparities within each country, cross-nationally the theory was only able to explain limited variation. |
