| dc.description.abstract | Emotions have been shown to have a significant impact on the learning process, being capable of either improving or hindering it. In particular, the academic emotion of confusion is attributed to having both a positive and a negative role in educational activities. Whereas confusion can foment willful deliberation regarding a task, a student that is confused for an extended period may transition to purely negative academic emotions, such as frustration or boredom. Human tutors are skilled in dealing with students with distinct traits of personality, as well as in detecting emotional cues and identifying learning difficulties, adapting their teaching strategies accordingly. Thus, to develop intelligent tutoring systems that perform as well as human tutors, it is important to consider these factors. Through the analysis of data obtained from 25 students
that made use of the PAT2Math system and completed a personality test, this work aims
to analyze how different traits of personality and past algebra performance on the PAT2Math system influence student’s permanence time in the academic emotion of confusion. To that end, statistical analysis and modeling were conducted through the lens of survival analysis, which is a powerful tool when the interest resides in analyzing time-to-event occurrences. The results obtained from the analysis indicated that students with higher levels of previous knowledge tolerated confusion for longer than students with lesser past performances, and personality-wise, students with dominant traits of neuroticism and agreeableness had similar threshold times of permanence in confusion, whereas extraverted students transitioned away from confusion substantially
quicker. Lastly, a model of student’s permanence time in confusion is proposed and
implemented, with the objective of future incorporation on the PAT2Math ITS, so that the student’s median and threshold permanence times in confusion may be used to regulate students’ emotions during learning, considering their respective personality traits and previous performance on the system. | |
