Application of WSQ (Watch-Summary-Question) Flipped Teaching in Affective Conversational Robots: Impacts on Learning Emotion, Self-Directed Learning, and Learning Effectiveness of Senior High School Students

Abstract

The study constructed a self-directed learning system with digital art teaching materials by using an affective conversational robot and a learning method guided by watch-summary-question (hereinafter referred to as WSQ) worksheets. The objective of this study, the affective conversational robot was used for interactive purpose and the WSQ learning strategy was integrated as the self-directed learning framework in order to boost the motives for self-directed learning in the students and also to help the teacher know students’ learning and difficulties. The EDRDA system was developed based on the rapid application development (RAD), to make the affective conversational robot enable to detect emotions and give appropriate feedback, we have compiled 4,200 sets of conversations to strengthen the training of Dialogflow intentions, and divided the conversations into two kinds: course knowledge texts (digital art course related) and sentimental texts (sentiment category). Subjects were 190 students from some senior high school in the southern part of Taiwan and lasted for 8 weeks in total. For the sake of consistency, the study did not distinguish between genders. There were 6 groups, with each group consisting of 4–5 randomized students. Then, the odd groups were set to be control groups (CGs, N = 95), the affective self-directed learning system was integrated with the general worksheet and the even groups were experiment groups (EGs, N = 95), the affective self-directed learning system was integrated with the WSQ worksheet while. Students in both groups were compared against one another in their differences in “learning emotions,” “positive learning emotions,” “negative learning emotions,” “learning effectiveness,” “satisfaction with self-directed learning” and “self-directed learning method.” The statistical quantitative analysis methods applied included analysis of covariance (ANCOVA), independent sample t-test, linear regression, and Welch’s-test. Statistical results show that the experiment group scored higher in positive learning emotions such as “hope” (LE = 2.05***, SL = 0.18**), “enjoyment” (LE = 1.91**, SL = 0.20*), and “pride” (LE = 1.59**, SL = 0.18***) in “learning effectiveness (LE)” and in “satisfaction with self-directed learning (SL)” than the control group and lower in negative learning emotions such as “anxiety” (LE = –0.99***, SL = –0.30*) and “anger” (LE = –0.86*, SL = –0.20**). The study data has indicated that it can improve learning sentiments. Learning sentiments will affect learning conditions. A system has been established to detect and identify learners’ affections, and transmit information to conversational robots so that learners can improve learning effectiveness (EG = 82.7368, CG = 70.3158) and motivation and the teaching strategies can be adjusted. It is sufficient to prove that it is beneficial to students’ self-directed learning (d = 0.2562 > 0.2).

Keywords:

• Dialogue robot
• WSQ
• learning emotions
• self-regulated learning
• learning effectiveness

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Additional information

Notes on contributors

Hao-Chiang Koong Lin

Hao-Chiang Koong Lin works as the Chair & Professor of the Department of Information and Learning Technology, National University of Tainan. He has published more than 380 internationally refereed research papers focused on affective computing, learning and education technology, and artificial intelligence.

Meng-Chun Tsai

Meng-Chun Tsai works as the Assistant Professor of the General Competency Center, National Penghu University of Science and Technology. She received her doctoral degree in Digital Learning from Department of Information and Learning Technology, National University of Tainan, 2021. She research focuses on Augmented Reality, Information and Digital learning.

Tao-Hua Wang

Tao-Hua Wang is a curatorial assistant of Science Education Department, National Museum of Natural Science, Taiwan. She received her doctoral degree in Digital Learning from Engineering Science Department, National Cheng Kung University, 2022. She has published several internationally refereed research papers focused on Learning Affects, Digital Learning and Popular Science.

Wen-Yi Lu

Wen-Yi Lu is currently a teacher at Guinan Elementary School in Tainan City. In 2022, she received a master’s degree in the Department Information and Learning Technology National University of Tainan. Her published research papers focus on dialogue robot, self-regulated learning, WSQ, positive emotion, and learning effectiveness.