運用開放軟體R 處理大型教育資料庫

Abstract

目前教育工作者處理大型教育資料庫常用的工具多為商用的統計軟體。資料管理與分析的流程因此不易與其他研究者分享，結果也較難被覆驗。信度不能累積的研究成效也不容易幫助指引教育政策的改進方向。本研究的主旨在示範如何運用開放軟體R 管理與分析大型教育資料庫的檔案。案例為，包括本國在內共十五個國家，十五歲學生數學成就表現與多個影響該表現的相關因素（PISA 2012）。範例分析，除了使用R 的基本功能外，特別嵌入Caro 與Biecek（2014）針對處理大型教育資料庫發展的intsvy 套件並結合maptools 等相關套件將分析結果呈現在地圖上以利比較。文中提及的檔案管理及資料分析過程皆可使用提供在附錄的R 程式碼複製。

Currently, researchers use mostly commercial software to analyze educational data. Researchers who do not have access to these commercial software can not easily replicate results others have reported in publications. Consequently, computational findings often do not cumulate reliably to offer valid guidelines for educational policy making. This paper demonstrates how to use the free, open-source R computing environment (R Core Team, 2015) to manage and analyze international large-scale educational assessment data. The example is comprised of mathematics achievement and covariates for 15-yearold students from 15 countries, including Taiwan (PISA 2012). In particular, the “intsvy” package developed by Caro and Biecek (2014) is applied to manage the data and the “maptools” package (Bivand & Lewin-Koh, 2015) is used to link numerical summaries to geographical boundaries of countries examined in the illustration. R code to perform analysis of data from PISA detailed in the paper is provided to facilitate reproducibility.