Indonesia National Examination Analysis

Transcript

1. The Role of Education System Decentralization on National Examination Outcome A Case Study of West Java, Indonesia Ramda Yanurzha
2. Newly released school-level data on national examination performance and geographic location. Why?
3. Inspirations S. Kristiansen and Pratikno, “Decentralising education in Indonesia,” Int. J. Educ. Dev., vol. 26, no. 5, pp. 513–531, 2006. A. Toi, “An empirical study of the effects of decentralization in Indonesian junior secondary education,” Educ Res Policy Pr., vol. 9, pp. 107–125, 2010. T. Muttaqin, M. Van Duijn, L. Heyse, and R. Wittek, “The impact of decentralization on educational attainment in Indonesia,” in Decentralization and Governance in Indonesia, Cham: Springer International Publishing, 2015, pp. 79–103. X. Qian and R. Smyth, “Measuring regional inequality of education in China: widening coast–inland gap or widening rural–urban gap?,” J. Int. Dev., vol. 20, no. 2, pp. 132–144, Mar. 2008.
4. Core Ideas • Education performance is both a byproduct and determinant of local development • Governance & fiscal decentralization puts greater responsibility on local government → magnified by “big bang decentralization” • Prior to decentralization, schools performance were largely dictated by distance to capital & major cities • Decentralization diffuses education quality within region but increase disparities between region
5. For schools in well-developed regions, boundaries matter more than distance. On Education Outcome
6. Reframing The Question Then Now Research Question What factors lead to education performance difference between districts? Do disparities in school national exam result reflect the regional decentralization in education system? Scope & Unit National, school-level 27 districts within a province, school-level Technique K-means Anselin Local Moran I (spatial clustering)
7. Approach • Prior studies on decentralization were always on macro level ◦ High level of aggregation on available data ◦ Shifting policies on administering national standardized exam • Through newly released school-level data, spatial clustering can reveal education disparities with more detail ◦ Is it caused by small number of very good schools instead of regional trend?
8. • 1 high-scoring + 2 low-scoring clusters, follows the district boundary • No correlation between exam score and distance to major cities and district capitals
9. What’s Next • Open up data for future researchers ◦ In talks with Ministry of Education • Link to governance & fiscal aspects ◦ Local education policy, education budget allocation mix • Comparison pre & post decentralization ◦ Requires additional data • Role of school-administered exam as a “shortcut” ◦ Evidence pointing towards score padding