Discovering Gender-Specific Knowledge from Finnish Basic Education using PISA Scale Indices
Saarela, M., & Kärkkäinen, T. (2014). Discovering Gender-Specific Knowledge from Finnish Basic Education using PISA Scale Indices. In J. Stamper, Z. Pardos, M. Mavrikis, & B. McLaren (Eds.), EDM 2014 : Proceedings of the 7th International Conference on Educational Data Mining (pp. 60-67). International Educational Data Mining Society (IEDMS). Retrieved from http://educationaldatamining.org/EDM2014/uploads/procs2014/long%20pape...
© the Authors & IEDMS, 2014.
The Programme for International Student Assessment, PISA, is a worldwide study to assess knowledge and skills of 15- year-old students. Results of the latest PISA survey conducted in 2012 were published in December 2013. According to the results, Finland is one of the few countries where girls performed better in mathematics than boys. The purpose of this work is to refine the analysis of this observation by using education data mining techniques. More precisely, as part of standard PISA preprocessing phase certain scale indices are constructed based on information gathered from the background questionnaire of each participating student. The indices describe, e.g., students’ engagement, drive and self-beliefs, especially related to mathematics, the main assessment area in PISA 2012. However, around 30% of the scale indices are missing so that a nonstructured sparsity pattern must be dealt with. We handle this using a special, robust clustering technique, which is then applied to Finnish subset of PISA data. Already direct interpretation of the created clusters reveals interesting patterns. Clusterwise analysis through relationship mining refines the confidence on our final conclusion that attitudes towards mathematics which are often gender-specific are the most important factors to explain the performance in mathematics. ...
PublisherInternational Educational Data Mining Society (IEDMS)
Parent publication ISBN978-0-9839525-4-1
Is part of publicationEDM 2014 : Proceedings of the 7th International Conference on Educational Data Mining, ISBN 978-0-9839525-4-1