Affective state and learning environment based analysis of students’ performance in online assessment

Online learning and assessments have become an integral part of education systems around the world. Online education and learning stand out as one of the important applications of cognitive computing where artificial intelligence - based models can be built by taking into account cognitive abilities of the students such as working memory, retention power, reasoning skills etc. The systems used for online assessment can be better designed considering some factors effecting learning environments and cognitive abilities of the learners. These factors such as sleep hours, time of the day, noise level, environment settings and mood etc. are closely linked with his/ her cognitive skills and affective state which in turn may effect performance in online assessment. This paper attempts to identify the effect of mood, sleep hours, time of the day and energy level on the performance of engineering students in an online assessment in India. The students have attempted the test on Moodle (2.9.0) in three sessions in a day viz. morning, afternoon and late afternoon. Sleep hours and energy level have been captured using a set of questions embedded within the online test items. Mood has been captured by the pictorial representation of various moods using robotic characters during the test. One -way ANOVA has been used as statistical measure to test various hypotheses. The study reported that sleep hours and perceived energy level during various times of the day had no significant effect on the performance of students. However, mood and time of the day had an impact on the students’ performance. Higher number of students performed the best in the morning hours of the day. Also, students with positive moods performed better than students who reported negative mood. The findings from this study can be used for designing a framework of an adaptive tutor based on cognitive skills for teaching different subjects to students.