There is a broad interest in narrowing achievement gaps among all groups of students and improving education by scientifically sound methods. On October 25, 2006, the United States Department of Education published new regulations allowing single-sex education in public schools whenever schools think it will improve student achievement. Thus far, studies comparing single-sex with coeducational schools have been carried out at the national level mostly in England, Australia and Jamaica, while US studies were limited to Catholic schools. Few studies reported descriptive statistics or effect sizes and most studies differ in the criteria and statistical controls they use to compare single-sex and coeducation. This dissertation presents models for science achievement and attitudes towards science for 8th-grade students attending either single-sex or coeducation schools in Hong Kong and New Zealand, using the Trends in Mathematics and Science Study (TIMSS) datasets from 1995, 1999, and 2003. To properly account for the nested structure of data, an HLM model was estimated for each sex, for each of the two jurisdictions at three time points, corresponding to the three TIMSS administrations. The within-country results were compared to see if differences between single-sex and coed schools were consistent over time. In addition, this dissertation proposed an approach to examine the sensitivity of the estimated effects of school-type on student outcomes to the presence of unmeasured variables which may introduce hidden selection bias, using a modification of the method proposed by Rosenbaum and Rubin (1983). Based on its conditional distribution with an instrumental variable, chosen based on the review of single-sex literature, the Monte Carlo simulated values of the unobserved variable were used as level-1 predictors in a one-way "ANCOVA" with random effects. The sensitivity analysis was limited to science achievement of Hong-Kong's girls in TIMSS 2003. Findings show that single-sex education contributed to girls' science performance and attitudes in NZL 1999 and HKG 1999 and 2003, and low sensitivity for school-type contrast. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]