This study explored the differences in academic achievement between Taiwanese school children(reference group) whose parents are both Taiwanese and Taiwanese school children with one non-Taiwanese parent(focal group). We usually called the school children with one-Taiwanese parents as new immigrant student. Many small scale studies have suggested that children of foreign mothers are at an educational disadvantage, but few large scale studies has been conducted to date. In order to cross validated the results of this study, two components were conducted over two years. Empirical data analysis is followed by a meta-analysis. In the first phase, data from the Taiwan Assessment of Student Achievement (TASA), a standardized, national achievement test which measures academic progress in grades four, six, eight and 11 was used to compare the respective academic achievement of the target and focal groups. The TASA scores used in the present study constitute nationwide, randomly sampled data representative of both the urban and rural Taiwanese population over a period of four years (2007-2010). The results show that the reference group consistently outperformed the focal group, showing small to medium effect sizes, which indicates a relationship between the immigrant status of mothers and academic performance of the children. Additionally, the home countries of the mothers is associated with intergroup differences, wherein children of mothers from Mainland China (including HongKong and Macau), will exhibit better academic performance than children of mothers from southeast Asian countries. As the focal group children spent more time in the Taiwanese education system, between-group differences decreased.
The second phase used meta-analysis to explore the academic difference between focal and reference groups. A total of 59 studies were coded and incorporated into the analysis. The results are converged with the results from phase one. The overall performance of reference group members is better than focal group members, and the effect size is small to median. In addition, the meta-analysis corroborated the phase one finding of decrease in differences with the increase of ages.
