Title: Dynamic models of student performance: a system identification approach

Authors: Sara I. Khaddaj, Rayyan G. Jaber, Fadi N. Karameh

Addresses: Department of Electrical and Computer Engineering, Faculty of Engineering and Architecture, American University of Beirut, Beirut, 1107-2020, Lebanon. ' Department of Electrical and Computer Engineering, Faculty of Engineering and Architecture, American University of Beirut, Beirut, 1107-2020, Lebanon. ' Department of Electrical and Computer Engineering, Faculty of Engineering and Architecture, American University of Beirut, Beirut, 1107-2020, Lebanon

Abstract: Enhancing the learning experience of college students is strongly tied to the introduction of efficient educational and testing tools which take into account the intellectual, social, and motivational responsiveness of students. It is therefore important to understand the main drives that alter the dynamics of individual student performance. This paper introduces a new paradigm in modelling student performance dynamics based on tools adopted from system identification theory. The presented approach suggests taking into account internal factors such as self-appreciation and previous performance history as well as external factors such as class environment and peer competitiveness. First and second order statistics are used as quantitative measures of academic performance and an existing dataset of student grades is used to identify the developed model parameters. It is shown that the parametric structure of such models naturally admits a classification scheme which qualitatively interprets the students as being class- or internally-driven. The theory of system identification has not been rigorously applied to this area of education. Accordingly, this study supports the effectiveness of such theory in modelling and classifying student performance traits, in providing natural predictors of such performance, and hence, in providing novel tools that could augment the learning enhancement process.

Keywords: higher education; system identification; student performance; ARMAX; mathematical modelling; classification; linear dynamic systems; intellectual responsiveness; social responsiveness; motivation; learning enhancement.

DOI: 10.1504/IJMIC.2011.042339

International Journal of Modelling, Identification and Control, 2011 Vol.14 No.1/2, pp.46 - 59

Published online: 21 Mar 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article