The Impact of the Affective Features of Feedback Providers on Multimedia Learning: An Eye-Tracking Study

• In Event: Beyond Self-Report Measures: Unlocking Insights Into Cognitive and Affective Feedback Processes With Multimodal Data Sources

Abstract

Objectives
The primary objective of this study is to improve learning from online lessons involving practice questions, by examining how the emotional tone of feedback providers displayed in video clips affects students’ learning experiences and outcomes. Despite extensive research on text-based feedback, the role of affective cues in video-based feedback remains underexplored. This study seeks to fill this gap by using eye-tracking technology to understand the cognitive and affective processes involved in learning.

Theoretical Framework
The feedback principle posits that feedback enhances learning, with explanatory feedback—clarifying why answers are correct or incorrect—being particularly beneficial for learners with less prior knowledge (Mayer, 2014; Padalkar & Hegarty, 2015). Additionally, the Cognitive-Affective Model of E-Learning suggests that effective learning involves a combination of cognitive, affective, and social processing (Mayer, 2020). Positive emotional elements in lessons can improve students’ felt emotions, enhance their social connection with instructors, boost motivation, and lead to better learning outcomes (Lawson et al., 2021; Zhao & Mayer, 2023). Lastly, eye-tracking data can provide objective measures of cognitive and affective processes during learning (Holmqvist & Andersson, 2017).

Methods and Data Source
This between-subjects study involved 117 college students who viewed a multimedia lesson on human color vision, consisting of five sections with two practice questions and immediate feedback after each section. Students were randomly assigned to one of three feedback conditions: the positive feedback provider condition featured a video of an instructor displaying positive affect while explaining the correct answer (n=38); the neutral feedback provider condition featured the same instructor displaying neutral affect (n=39); and the corrective text feedback condition provided on-screen text indicating the correct answer without detailed explanations (n=40). Following the lesson, students rated the instructor’s emotion, their own emotion, their social connection with the instructor, their motivation and their effort level in several surveys, and took a posttest that assessed their learning outcome. Eye-tracking data, including fixation duration, saccades, pupil dilation, and eye openness, were collected during the lesson and feedback presentations.

Results and Scientific Significance
Using one-way ANOVA, we found that the positive feedback provider group rated the instructor as more positive (p < .001), facilitative (p < .001), credible (p = .004), human-like (p < .001), and engaging (p < .001). Additionally, the positive feedback provider group reported the enhanced positive feelings (p < .001), motivation (p = .04), effort (p = .04), and transfer test scores (p < .001). Lastly, the eye-tracking analysis also revealed a significantly longer fixation duration during learning with positive feedback provider than with neutral feedback provider (p = .003).

Scientific Significance
This study expands the understanding of the role of affective features in feedback-based learning, an area that has predominantly focused on text-based feedback and cognitive aspects of learning. By incorporating eye-tracking technology, the research provides objective data on how affective and attentional processes contribute to learning outcomes. Overall, the findings highlight the importance of positive emotional cues in feedback, suggesting that incorporating positive affect in feedback can enhance student engagement and learning outcomes.

Authors

• Fangzheng Zhao, University of California - Santa Barbara

• Richard E. Mayer, University of California - Santa Barbara