Investigating Students' Emotional Dynamics When Working in Small Groups

• In Event: 66.023 - "What Is This Feeling?" The Wickedly Complex Workings of Achievement Emotions in Diverse Learning Situations

Abstract

With cooperative-group learning, students work together in small groups to maximize each student’s learning. Although research has shown that cooperative learning is associated with higher achievement (Kuh, Pace, & Vesper, 1997), college students still value traditional lecture over group-work (Machemer & Crawford, 2007). In this study, we investigated how students’ emotions, experienced in one session of group-work, affected students’ emotions in the next session, suggesting a possible feed-forward effect. Additionally, we wanted to know if this process was different when students were involved in formal, graded group-work and informal non-graded group-work.

A total of 108 undergraduate students were assigned 12 case-study group-projects that were conducted in class. These assignments contributed to 30% of their course grade. Approximately half (n=55) of the students’ case-studies were graded for the first half of the semester, whereas the others’ were not; they received grades on case-studies during the second half. At semester’s start, students’ valuing of group work and their expectations of group work were assessed with seven researcher-developed items. After each case-study, students completed a brief survey asking them to rate their emotions during their group-work (5-point Likert scale, from “not at all” to “extremely”). The emotions were: proud of self, proud of group, enjoyment, hopeful, boredom, anxiety, hopeless, anger, shame of self, shame of group.

An exploratory factor analysis revealed three factors we named optimism, pessimism, and shame; a confirmatory factor analysis with a different sample confirmed these factors. Using these factors, descriptive analysis demonstrated great fluctuation of individual students’ emotions across the semester, suggesting that students’ emotions did not seem to have a feed-forward effect. However, other noticeable patterns emerged. Students who had projects graded during the first half of the semester (but not the second), experienced increased pessimism and shame, and less optimism on the fourth project, a project perceived as more difficult. Students whose projects were graded the second half of the semester (and not the first half) did not experience increased pessimism and decreased optimism at that time.

Students whose projects were graded the second half of the semester tended to experience more pessimism and less optimism the week before their first graded-project, perhaps judging their group performance as “lacking” as they anticipated impending grades. Students who received grades during the second half of the semester did not have much emotional fluctuation. However, students who did not receive grades the second half of the semester tended not to complete surveys as the semester came to an end. We suspect these students may have become less engaged because they were not receiving grades.

Further analysis will investigate the extent to which a hidden Markov model demonstrates students’ change of emotions at a particular point for most students, between the two groups. A latent state-trait model is expected to represent a special subgroup of individuals. This analysis is intended to improve our understanding of non-linear changes in students’ emotions and the value of this application to longitudinal investigations.

Authors

• Jeannine E. Turner, Florida State University

• Jessica J. Summers, The University of Arizona

• Sicong Liu, Florida State University

• Russell Almond, Florida State University