Abstract—Learners’ performances in intelligent tutoring
systems or e-learning environments depend on various factors
such as the nature of the task presented, their cognitive and
affective abilities, etc. In this paper, we focus on studying in
detail the variation of these different factors and more
specifically the electroencephalogram (EEG metrics) and how
they differ according to a category and a type of cognitive tasks.
We also studied the possibility of predicting a learner’s
performance using feature selection and multiple regressions.
Primarily, results shows that learners’ scores could be predicted
using in descending order the difficulty level of the task, the type
of a task, the duration of a task and the EEG workload metric by
building a multiple regression model that fit our data.
Index Terms—Cognitive tasks, EEG features, engagement, workload, distraction, multiple regression.
The authors are with the Université de Montréal, Département d’Informatique et de Recherche Opérationnelle, 2900 Chemin de la Tour, Montréal, H3C 3J7 Canada (e-mail: email@example.com, firstname.lastname@example.org, email@example.com).
Cite: Ramla Ghali, Sébastien Ouellet, and Claude Frasson, "Using Electrophysiological Features in Cognitive Tasks: An Empirical Study," International Journal of Information and Education Technology vol. 6, no. 8, pp. 584-590, 2016.