大型資料庫與後設分析結果之交互驗證性研究-以新移民與本土子女學習成就差異為例

A Cross Validation Study of Differentiated Performance between Native and Immigrant Children with Results from Meta-Analysis and Large-Scale Assessment-Taking the Difference in Learning Achievement between New Immigrants and Taiwanese Students as an Example

在本研究中，以新移民子女與本土子女的學習成就差異之議題為分析主軸，來相 互驗證後設分析與大型資料庫研究方法所產生的結果。在後設分析策略中，使用 59 篇文獻進行訊息編碼及後設分析。而大型資料庫則採用 TASA 資料，資料來自 2007 年至 2013 年四年級、六年級、八年級與高中職二年級的實徵數據，研究結果顯示， 兩種數據蒐集方法，經統計分析後所得的結果大致符合。本土子女的整體學習表現較 新移民子女來得佳，其學習表現差異平均效果量是屬小至中等程度，此外，隨年齡增 長，差異有隨之微幅下降的趨勢，兩種數據蒐集方法之優缺點也於文中進行探討。

 

The purpose of this study was to assess the difference in learning achievement between the local children and the new-immigrant children in Taiwan. A large-scale database analysis and meta- analysis were adopted to analyze. TASA data which taken from four, six, eight, and eleven graders from 2007 to 2013 were adopted as the large-scale database. Fifty-nine research papers were used for coding as a meta-analysis. This result was in good consistency with both analyses. The Taiwan local children’s learning performance is better than new-immigrant children’s. However, there was some discrepancy between performances. In addition, with the students’ age increase, the difference in the performance became smaller. The advantages and disadvantages of both study methods were discussed as well.