Typology of students graduating from college using dropout probabilities
DOI:
https://doi.org/10.52731/lir.v003.159Keywords:
Typology, Pattern, Dropout Probability, ClusteringAbstract
This study uses dropout probability to identify the degree to which students who graduated came close to dropping out and to examine how students who graduated went from being close to dropping out to graduating. Dropout probability is a number presented by Shiratori that represents how likely a student is to drop out [3]. Although the dropout probability has been used to analyze the patterns of students who dropped out of school, the process by which students graduated has not been analyzed. In this study, the dropout probability was used to quantify the student status of students who graduated, semester by semester and student by student, and the student status up to graduation was summarized as a transition vector. The transition vectors were then typified using k-means to create patterns of graduated student states. Among the students who graduated, we found that about 14% were in a state in which they could have dropped out. By analyzing the above patterns, we were able to determine how many students were in poor condition and at what stage they were likely to graduate.
References
Yomiuri Shimbun Kyoiku Network, [University Competencies 2019] ”Daigaku no Jitsuryoku 2019 (in Japanese)”,Chuo Koron Shinsh, 2018.
Ministry of Education, Culture, Sports, Science and Technology. [Survey on the status of student enrollment (drop-outs and leaves of absence) (as of the end of the 2021 academic year)] “Gakusei no shugakujokyo (Chutaisha, Kyugakusha) ni kansuru Chosa (Reiwa 3nendo matsu jiten) (in Japanese)”. 2022.
Naruhiko Shiratori et al., [Making Dropout Patterns Using Transition of Dropout Probability] ”Chutaikakuritsu no seni wo motiita chutaigakusei no ruikeika (in Japanese)”. Nihon Kyoikukougakkai Ronbunshi (Japan journal of educational technology), vol 44, no.1, pp. 11-22, 2020.
Seidman, Alan. “Taking Action: A Retention Formula and Model for Student Success.” College Student Retention: Formula for Student Success. pp. 267–84. 2012.
Nobuhiko Kondo and Toshiharu Hatanaka., [Modeling of Learning Process based on Bayesian Network] “Bayesian Network niyoru Shugakujoutai Suitei Model no Koutiku (in Japanese)”. Nihon Kyoikukougakkai Ronbunshi (Japan journal of educational technology), vol. 41, no. 3, 2018, pp. 271–81.
Naruhiko Shiratori, et al., [Typification of first-year spring semester study status using predicted GPA trends] “Yosoku GPA no suii wo mochiita 1nenji harugakki gakushu jotai no ruikeika (in Japanese)”. Kyoiku System Johogakkaishi (Transactions of Japansese Society for Information and Systems in Education), vol. 39, no. 4, 2022, pp. 440–51.