ISSN 1671-3710
CN 11-4766/R

Advances in Psychological Science ›› 2022, Vol. 30 ›› Issue (7): 1524-1535.doi: 10.3724/SP.J.1042.2022.01524

• Regular Articles • Previous Articles     Next Articles

Can affective pedagogical agent facilitate multimedia learning?

WANG Yanqing, GONG Shaoying(), JIANG Tiantian, Wu Yanan   

  1. Key Laboratory of Adolescent Cyberpsychology and Behavior, Ministry of Education;School of Psychology, Central China Normal University, Wuhan 430079, China
  • Received:2021-08-24 Online:2022-07-15 Published:2022-05-17
  • Contact: GONG Shaoying


In recent years, with the development of computer technology and intelligent tutoring systems, educational psychologists have paid much attention to the effect of adding an affective pedagogical agent (PA) to multimedia learning. Affective PA is a type of agent that is designed to elicit certain affective experiences in learners through multiple modalities such as facial expressions, voices, and gestures. Previous studies mainly focused on two types of affective PA: expressive affective PA and empathic affective PA. Expressive affective PAs are agents that affect learners' emotional experiences only through emotional expression (e.g., using smiling facial expressions and enthusiastic voices). Empathic emotional agents are agents with the goal to encourage and persuade learners to make greater effort by giving emotional feedback (such as nodding, encouragement, and additional remarks) according to the learner's emotional states. Although previous studies differed in the specific manipulation of affective PAs, no matter what type of affective PA is designed to increase the positive emotions in learners, improve internal motivation, and ultimately promote learning. Five theories were used to explain the effectiveness of affective PAs in multimedia learning environments. Emotional contagion theory holds that the emotional state of one person is automatically affected by another person's emotional expression. When learning with a smiling pedagogical agent, students are more likely to experience more positive emotions and higher motivation. Emotional response theory proposes that pedagogical agents with enthusiastic voices, smiling faces, and expressive gestures could elicit positive emotional responses in learners and promote them to engage in learning-related activities. The cognitive affective theory of learning with media (CATLM) emphasizes the importance of emotions and motivation in the learning process. Based on the CATLM, the cognitive-affective model of learning with instructional video states that when pedagogical agents display positive emotions during online learning, learners may experience four key steps: (1) the learners first need to recognize the PA’s positive emotions; (2) the learners respond to the PA’s affective stances (such as feeling the same emotions as the affective PAs); (3) the learners’ positive emotions improve the level of motivation to engage in deep cognitive processing; (4) the motivational states lead to better learning outcomes. However, cognitive load theory and interference theory hold the opposite opinion that affective agents may increase external cognitive load and distract learners from relevant information. Accordingly, learners no longer focus their full attention on the learning content and become distracted by the presence of the affective PAs, which may interfere learning. Guided by previous theories, a large number of studies have examined the role of affective pedagogical agents in multimedia learning. Most empirical studies found that affective pedagogical agents positively affect the learners’ emotions (d positive emotions = 0.45) and motivations (d intrinsic motivation = 0.52). Nevertheless, affective pedagogical agents have weak effects on cognitive load(d intrinsic cognitive load = -0.01;d external cognitive load = 0.09;d germane cognitive load = 0.08)and learning performance (d retention = 0.18, d comprehension = 0.32, d transfer = 0.14,d unite = 0.32). Previous studies showed that adding affective PA to multimedia lessons does not lead to stable improvement in learning performance, which may be due to the potential moderating variables. In the current study, we identified several potential factors that may moderate the effects of affective PA, such as learners’ individual characteristics (e.g., learner’s working memory and grade levels), the types of affective PAs, learning domain, and the way of tests. In conclusion, adding affective PA is regarded as a promising approach for education because of its potential to help students keep motivated during learning. Therefore, in educational practice, instructional designers can consider presenting a positive affective PA for learners to help them learn more happily. Further research should focus on the manipulation and evaluation methods, boundary conditions, underlying mechanisms, and ecological validity, etc.

Key words: affective pedagogical agent, positive emotions, motivation, learning performance, multimedia learning

CLC Number: