Which pathways lead to success?

Transformation models and change-point detection for the early identification of students at risk

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contribution

Abstract

In modern Universities, the timely identification of students at risk is valuable to limit failure and withdrawal rates. In this study we define the concept of students’ academic pathways, and use this and other variables to detect when students become at risk. We approach the problem modelling the latent variable “student learning” in the framework of transformation models, and use an indicator function readably available in this framework to detect change-points that depend on the observations already made but do not depend on the future which is not yet observed. We show how change-points can be detected as
quickly as possible after they have occurred, while keeping the false-positive rate at a low predefined level. The methodology is applied to a data base of students graduated with Science Degrees at Macquarie University in the last 10 years.
Original languageEnglish
Title of host publicationProceedings of the 34th International Workshop on Statistical Modelling
Subtitle of host publicationVolume II
EditorsLuis Meira-Machado, Gustavo Soutinho
Place of PublicationGuimaraes, Portugal
PublisherIWSM
Pages212-215
Number of pages4
Volume2
ISBN (Electronic)9789892096308
Publication statusPublished - 2019
EventInternational Workshop on Statistical Modelling (34th : 2019) - Guimaraes, Portugal
Duration: 7 Jul 201912 Jul 2019

Conference

ConferenceInternational Workshop on Statistical Modelling (34th : 2019)
CountryPortugal
CityGuimaraes
Period7/07/1912/07/19

    Fingerprint

Keywords

  • transformation models
  • change-point detection
  • student learning

Cite this

Manuguerra, M., & Sofronov, G. (2019). Which pathways lead to success? Transformation models and change-point detection for the early identification of students at risk. In L. Meira-Machado, & G. Soutinho (Eds.), Proceedings of the 34th International Workshop on Statistical Modelling: Volume II (Vol. 2, pp. 212-215). Guimaraes, Portugal: IWSM.