2021, Volume 24, Issue 3
Special Issue on "From Conventional AI to Modern AI in education- Re-examining AI and Analytics Techniques for Teaching and Learning"
Guest Editor(s): Haoran Xie, Gwo-Jen Hwang and Tak-Lam Wong
Full Length Articles
Ahmed Ramadan Khtere and Ahmed Mohamed Fahmy Yousef
Ahmed Ramadan Khtere
Foundations of Education Department, Faculty of Education, Fayoum University, Egypt // email@example.com
Ahmed Mohamed Fahmy Yousef
Educational Technology Department, Faculty of Specific Education, Fayoum University, Egypt // firstname.lastname@example.org
The study aimed to examine the readiness of faculty members in Arab universities for blended learning environments through an investigation of the attributes, skills, and knowledge in three roles of professional online teachers. Online teaching professionalism has been described as a set of required competencies, and behaviours for the effectiveness of educational online sessions. The authors have argued some requirements of teachers’ roles as an instructional planner, an assessor, and as a mentor. A purposive sample of 24 experts from diverse disciplines contributed to the reference panel in a Delphi study through three rounds. Qualitative content analysis and some descriptive statistics e.g., the median and frequency distribution, have been used to reach a consensus among the panel of experts. A matrix of 30 requirements was shortlisted by experts in different roles. The panelists provided insight into the top 10 requirements for each role to measure the professionalism of faculty before, during, and after the online sessions. The readiness for online teaching was concluded by six main domains namely evaluating students’ achievements and limitations, problem-solving skills, information technology and computer skills, monitoring and motivating techniques, communication, and class management skills. The study results can be used to plan faculty development programs based on performance gaps of faculty members at three levels: individual, departmental, and program or college. Moreover, the listed faculty attributes help higher education institutions to evaluate the perceptible skills and personal characteristics of faculty in enhancing the efficacy of online teaching in different academic disciplines.
Teaching professionalism, Online learning, Faculty readiness, Faculty competences
Wu-Yuin Hwang, Uun Hariyanti, Yan Amal Abdillah and Holly S. L. Chen
Graduate Institute of Network Learning Technology, National Central University, Jhongli City, Taiwan // email@example.com
Graduate Institute of Network Learning Technology, National Central University, Jhongli City, Taiwan // firstname.lastname@example.org
Yan Amal Abdillah
Graduate Institute of Network Learning Technology, National Central University, Jhongli City, Taiwan// email@example.com
Holly S. L. Chen
Luzhou Elementary School, Taipei City, Taiwan // firstname.lastname@example.org
Geometry is essential for mathematics learning given that it is strongly related to our surroundings; however, few studies concentrated on using geometry in our daily life, especially using mobile devices with their sensors. Thus, this study proposed one app, Ubiquitous Geometry (UG), and explored its effects on learning angles and polygons in authentic contexts. The experiment was conducted for grade four learners of an elementary school. The control group used protractors and pencil/paper in measuring angles and polygons, whereas the experimental group did measurements with UG. The results showed that in terms of learning achievement, the experimental group outperformed the control group. Further investigation of the relationship between learning behaviors and learning achievement in the experimental group found that both learning effectiveness and quantity of learning, including measuring angles of elevation and depression (MED), note drawing, and comment drawing, have significantly positive correlations with learning achievement. These three behaviors also become significant predictors of learning achievement after multiple regression analysis. Moreover, MED was found to be the most critical factor to affect learning achievement. Additionally, in perception evaluation, participants felt satisfied with UG and authentic measurement activities by which their learning motivation and interests in authentic contexts were indeed stimulated. Hence, we suggested that UG was worth promoted and further investigated its effects on authentic geometry learning.
Measurement in authentic contexts, Learning behaviors, Cognitive abilities, Ubiquitous Geometry
Tsung-Yen Chuang, Martin K.-C. Yeh and Yu-Lun Lin
National University of Tainan, Taiwan // email@example.com
Martin K.-C. Yeh
Penn State University – Brandywine, Pennsylvania, USA // firstname.lastname@example.org
National University of Tainan, Taiwan // email@example.com
Students with different cognitive styles benefit from different instructional strategies, including learning through playing video games. Although playing video games can be an effective learning method, we do not know its impact on the reasoning ability of students with different cognitive styles. The purposes of this study are to investigate whether students with different cognitive styles improve their reasoning ability after playing video games and whether the effect is the same for all students. We used a pretest-posttest experimental design with multivariant analyses and found that elementary school students’ reasoning ability improved reliably after playing a puzzle adventure game for four weeks, twice a week. In addition, field-independent students’ reasoning ability improved reliably more than field-dependent students did. Students with different cognitive styles also demonstrated noticeably different information search strategies during game playing. Our work answers the questions regarding the impact of playing video games in students’ reasoning ability and in students with different cognitive styles. We also suggested guidelines of designing educational video games for field-dependent and field-independent students. Future studies are needed to expand our understanding to the relationships between other types of video game, cognitive ability, and cognitive styles.
Cognitive style, Digital game, Reasoning ability, Game-based learning
Facultad de Educación y Psicología, Universidad Francisco de Vitoria, Spain // firstname.lastname@example.org
Flipped Classroom methodology is gaining relative importance as time goes by, in part due to the spreading and accessibility of technological resources in the educational field. Nonetheless, the effectiveness of this methodology is still being discussed. In this sense, the aim of this study is to analyse whether flipped classroom methodology is a more effective methodology than other methodologies. For this purpose, a systematic review was carried out, considering as valid studies those that had a pre-post and a control group. Based on a total of 61 studies (n = 5541 students) from 18 databases, results revealed that Flipped Classroom methodology is more effective than other methodologies in terms of learning achievement, in secondary and higher education, and it could be more beneficial than other methodologies in other constructs as motivation, self-efficacy, cooperativeness and engagement, among others. In primary education, findings revealed that Flipped Classroom could be as effective as other methodologies with regard to learning achievement, and other construct, such as self-concept and social climate. Depending on the educational stage, the effect size of differences was between 1.36 to 1.80 times larger in the case of Flipped Classroom group in comparison with control group. Based on these results, the Flipped Classroom could be more beneficial in comparison with traditional methodologies that are mainly used in higher education. However, it would not more beneficial in other educational stages where traditional approaches are not commonly used, such as in primary education.
Flipped classroom, Primary education, Secondary education, Higher education, Effectiveness
Mohamad Ridhuan Mat Dangi and Maisarah Mohamed Saat
Mohamad Ridhuan Mat Dangi
Faculty of Accountancy, Universiti Teknologi MARA, Selangor Branch, Puncak Alam Campus, Malaysia // email@example.com
Maisarah Mohamed Saat
Azman Hashim International Business School, Universiti Teknologi Malaysia, Johor, Malaysia // firstname.lastname@example.org
The findings of this study reveal that it is unlikely for the interaction effects of situational context, namely educational technology experience (EXP), training frequency (TF), voluntariness (VOL), and class size (CSIZE), to influence accounting educators’ intention to adopt educational technology. The original Technology Acceptance Model (TAM), which has been modified numerous times, is still relevant, especially for developing countries since their educational technology penetration is still very low. Conscientiousness trait from the Big Five Personality Model was applied in this study to measure intention as a powerful factor associated with the nature of individuals involved in the accounting profession. Measuring the factors from the individual perspective adds insight into the extant literature since past studies focused on organisational factors and student as the subject. The current study also overcomes the issue of stagnation in the accounting literature, specifically in the field of educational technology. Furthermore, this paper contributes by offering a good indication of using Structural Equation Modelling in the study, especially in the area of accounting and education, and using the most current reporting requirement for information system research.
Accounting Education, Acceptance Behaviour, Conscientiousness, Educational Technology
Haoran Xie, Gwo-Jen Hwang and Tak-Lam Wong
Department of Computing and Decision Sciences, Lingnan University, Hong Kong SAR // email@example.com
Graduate Institute of Digital Learning and Education, National Taiwan University of Science and Technology, Taiwan // firstname.lastname@example.org
Department of Computing Studies and Information Systems, Douglas College, Canada // email@example.com
With the rapid development and significant successfulness of various deep learning techniques in artificial intelligence (AI) in recent years, the connotation of AI has been transformed from traditional rule-based or statistical learning models to deep learning models. Such a transformation of AI has led to a significant evolution in both academic and industrial fields. To understand the potential impact of AI evolution for future teaching and learning, it is necessary to re-examine the opportunities, research issues, and roles of AI in education as modern AI enables the possibility of playing vital roles in education, which are not only limited to intelligent tutors/tutees but also intelligent learning partners or policy making advisors. Motivated by the recent transformation and trends in AI in education, this special issue, including 13 research articles, aims to launch an in-depth discussion on re-examining AI and analytics techniques in teaching and learning applications.
Modern AI, AI transformation, Deep neural networks, Analytic techniques, AI in education
Special Issue Articles
Ching Sing Chai
Department of Curriculum and Instruction & Centre for Learning Sciences and Technologies, The Chinese University of Hong Kong, Hong Kong, China // firstname.lastname@example.org
Department of Education, National Kaohsiung Normal University, Taiwan // email@example.com
Morris Siu-Yung Jong
Department of Curriculum and Instruction & Centre for Learning Sciences and Technologies, The Chinese University of Hong Kong, Hong Kong, China // firstname.lastname@example.org
Department of Curriculum and Instruction & Centre for Learning Sciences and Technologies, The Chinese University of Hong Kong, Hong Kong, China // email@example.com
Thomas K. F. Chiu
Department of Curriculum and Instruction & Centre for Learning Sciences and Technologies, The Chinese University of Hong Kong, Hong Kong, China // firstname.lastname@example.org
School of Mechanical-Electronic and Vehicle Engineering, Beijing University of Civil Engineering and Architecture, Beijing, China // email@example.com
Artificial Intelligence (AI) is increasingly popular, and educators are paying increasing attention to it. For students, learning AI helps them better cope with emerging societal, technological, and environmental challenges. This theory of planned behavior (TPB)-based study developed a survey questionnaire to measure behavioral intention to learn AI (n = 682) among primary school students. The questionnaire was administered online, and it measured responses to five TPB factors. The five factors were (1) self-efficacy in learning AI, (2) AI readiness, (3) perceptions of the use of AI for social good, (4) AI literacy, and (5) behavioral intention. Exploratory factor analysis and a subsequent confirmatory factor analysis were used to validate this five-factor survey. Both analyses indicated satisfactory construct validity. A structural equation model (SEM) was constructed to elucidate the factors’ influence on intention to learn AI. According to the SEM, all factors could predict intention to learn AI, whether directly or indirectly. This study provides new insights for researchers and instructors who are promoting AI education in schools.
Artificial intelligence, Self-efficacy, Readiness, Social good, Literacy, Behavioral intention
Beyin Chen, Gwo-Haur Hwang and Shen-Hua Wang
Department of Information Technology, Ling Tung University, Taiwan // firstname.lastname@example.org
Bachelor Program in Industrial Technology, National Yunlin University of Science and Technology, Taiwan // email@example.com
Department of Information Management, Ling Tung University, Taiwan // firstname.lastname@example.org
The application of artificial intelligence (AI) in education is now widespread, and the use of robots in education has demonstrated a positive influence on students’ behavior and development. However, the use of emerging technologies usually results in cognitive load, especially for elementary school students whose learning capacity has not yet been established. In addition, students of different genders have different physical, psychological and learning characteristics, so gender differences affect cognitive load. Cognitive load can be divided into two types: positive cognitive load and negative cognitive load. Usually, positive cognitive load results in good learning performance while negative cognitive load results in bad learning performance. Therefore, we use the cognitive load theory to define learning efficiency as the co-impact of learning performance and cognitive load. We take game-based intelligent robots for Chinese idiom learning as an example, and explore the impacts of gender differences on elementary school students. To achieve these aims, this study combined games and Zenbo robots, and applied them to educate elementary school students in the use of Chinese idioms. Secondly, this study conducted an experiment and analyzed the experimental results. The participants were 24 fourth-grade elementary school students from the central region of Taiwan. Results showed that this system is more beneficial for boys as their cognitive load was significantly lower. Boys’ learning performance was also better, although the difference did not reach significance. Furthermore, learning efficiency for boys was significantly higher. Reasons for these results are explained.
Artificial intelligence, Cognitive load theory, Game-based learning, Gender differences, Robots
Youmei Wang, Chenchen Liu and Yun-Fang Tu
Department of Educational technology, University of Wenzhou, 325035, China // email@example.com
Department of Educational technology, University of Wenzhou, 325035, China // firstname.lastname@example.org
Department of Educational technology, University of Wenzhou, 325035, China // Department of Library and Information Science, Research and Development Center for Physical Education, Health, and Information Technology, Fu Jen Catholic University, Taiwan // email@example.com
Owing to the rapid advancements in artificial intelligence (AI) technologies, there has been increasing concern about how to promote the use of AI technologies in school settings to enhance students’ learning performance. Teachers’ intention to adopt AI tools in their classes plays a crucial role in this regard. Therefore, it is important to explore factors affecting teachers’ intention to incorporate AI technologies or applications into course designs in higher education. In this study, a structural equation modeling approach was employed to investigate teachers’ continuance intention to teach with AI. In the proposed model, 10 hypotheses regarding anxiety (AN), self-efficacy (SE), attitude towards AI (ATU), perceived ease of use (PEU) and perceived usefulness (PU) were tested, and this study explored how these factors worked together to influence teachers’ continuance intention. A total of 311 teachers in higher education participated in the study. Based on the SEM analytical results and the research model, the five endogenous constructs of PU, PEU, SN, and ATU explained 70.4% of the changes in BI. In this model, SN and PEU were the determining factors of BI. The total effect of ATU was 0.793, followed by SE, with a total effect of 0.554. As a result, the intentions of teachers to learn to use AI-based applications in their teaching can be predicted by ATU, SE, PEU, PU and AN. Among them, teachers’ SE positively influenced teachers’ PEU and ATU towards adopting AI-based applications, and also influenced PU through PEU. In addition, the relationship between teachers’ SE and AN was negatively correlated, which indicated that enhancing teachers’ SE could reduce their AN towards using AI-based applications in their teaching. Accordingly, implications and suggestions for researchers and school teachers are provided.
Artificial intelligence, Higher education, Anxiety, Self-efficacy, Technology acceptance model
Chia-An Lee, Jian-Wei Tzeng, Nen-Fu Huang and Yu-Sheng Su
Department of Computer Science, National Tsing Hua University, Taiwan // firstname.lastname@example.org
Center for Teaching and Learning Development, National Tsing Hua University, Taiwan // email@example.com
Department of Computer Science, National Tsing Hua University, Taiwan // firstname.lastname@example.org
Department of Computer Science and Engineering, National Taiwan Ocean University, Taiwan // email@example.com
Massive open online courses (MOOCs) provide numerous open-access learning resources and allow for self-directed learning. The application of big data and artificial intelligence (AI) in MOOCs help comprehend raw educational data and enrich the learning process for students and instructors. Thus, we created two deep neural network models. The first model predicts learning outcomes on the basis of learning behaviors observed when students watch videos. The second is a novel exercise-based model that predicts if a student will correctly answer examination questions on relevant concepts. The study data were collected from two courses conducted on the National Tsing Hua University’s MOOCs platform. The first model accurately evaluated student performance on the basis of their learning behaviors, and the second model efficiently predicted student performance according to how they answered the exercise questions. In conclusion, our AI system remedies the present-day inability of MOOCs to evaluate student performance. Instructors can use the systems to identify poor-performing students and offer them more assistance on a timely basis.
Learning analytics, Educational big data, Massive open online courses, Artificial intelligence
Albert C. M. Yang, Irene Y. L. Chen, Brendan Flanagan and Hiroaki Ogata
Albert C. M. Yang
Graduate School of Informatics, Kyoto University, Japan // firstname.lastname@example.org
Irene Y. L. Chen
Department of Accounting, National Changhua University of Education, Taiwan // email@example.com
Academic Center for Computing and Media Studies, Kyoto University, Japan // firstname.lastname@example.org
Academic Center for Computing and Media Studies, Kyoto University, Japan // email@example.com
Reviewing learned knowledge is critical in the learning process. Testing the learning content instead of restudying, which is known as the testing effect, has been demonstrated to be an effective review strategy. However, education research recommends that instructors generate practice tests, but this burdens teachers and may also hinder teaching quality. To resolve this issue, the current study applied a modern artificial intelligence technique (BERT) to automate the generation of tests and evaluate the testing effect through e-books in a university lecture (N = 74). The last 5 minutes of each course session were utilized to review the taught content by having students either answer cloze item questions or restudy the summary of the core concepts covered in the lecture. A reading comprehension pretest was conducted before the experiment to ensure that the differences in prior knowledge were nonsignificant between groups, and a posttest was performed to examine the effectiveness of testing. In addition, we evaluated students’ reading skills and reading engagement through their ability to identify key concepts and their interaction with e-books, respectively. A positive effect was observed for students who engaged in cloze item practice before the end of each class. The results indicated that the repeated testing group exhibited significantly better reading skills and engaged more with e-books than the restudying group did. More importantly, compared with only restudying the key concepts, answering the cloze items questions significantly improved students’ reading comprehension. Our results suggest that machine-generated cloze testing may benefit learning in higher education.
Modern AI, Repeated testing, Testing effect, Test-enhanced learning
Owen H. T. Lu, Anna Y. Q. Huang, Danny C. L. Tsai and Stephen J. H. Yang
Owen H. T. Lu
College of Computer Science, National Pingtung University, Taiwan // firstname.lastname@example.org
Anna Y. Q. Huang
Department of Computer Science and Information Engineering, National Central University, Taiwan // email@example.com
Danny C. L. Tsai
Department of Computer Science and Information Engineering, National Central University, Taiwan // firstname.lastname@example.org
Stephen J. H. Yang
Department of Computer Science and Information Engineering, National Central University, Taiwan // email@example.com
Human-guided machine learning can improve computing intelligence, and it can accurately assist humans in various tasks. In education research, artificial intelligence (AI) is applicable in many situations, such as predicting students’ learning paths and strategies. In this study, we explore the benefits of repetitive practice of short-answer questions could enhance students’ long-term memory for subsequent improvements in learning performance. However, frequent authoring questions and grading requires teachers’ professionalism, experience, and considerable efforts. Therefore, this study using modern AI technologies, specifically natural language processing, to provide Automatic question generation (AQG) solution, a combined semantics-based and syntax-based question generation system: Hybrid automatic question generation (Hybrid-AQG) was proposed in this study. We assessed its functionality and student learning performance by asking 91 students to complete short-answer questions and then applied a process similar to the Turing test to evaluate the question and grading quality. The results demonstrated that modern AI technologies can generate highly realistic short-answer questions because: (1) Compared with the control group, the experimental group exhibited significantly better learning performance, implying that students acquired long-term memory of course knowledge through repetitive practice with machine-generated questioning. (2) The experimental group could better distinguish machine-generated and expert-authored questions. Nevertheless, both groups in distinguishing questions presented like guessing. (3) Machine grading was deficient in some respects; but the way students answer questions can be adapted for machine understanding through repetitive practice.
Automatic question generation, Learning performance, Artificial intelligence, Practice testing, Turing test
Lanqin Zheng, Lu Zhong, Jiayu Niu, Miaolang Long and Jiayi Zhao
School of Educational Technology, Faculty of Education, Beijing Normal University, Beijing, China // firstname.lastname@example.org
School of Educational Technology, Faculty of Education, Beijing Normal University, Beijing, China // email@example.com
School of Educational Technology, Faculty of Education, Beijing Normal University, Beijing, China // firstname.lastname@example.org
School of Educational Technology, Faculty of Education, Beijing Normal University, Beijing, China // email@example.com
School of Educational Technology, Faculty of Education, Beijing Normal University, Beijing, China // firstname.lastname@example.org
In recent years, the rapid development of artificial intelligence has increased the power of personalized learning. This study aimed to provide personalized intervention for each group participating in computer-supported collaborative learning. The personalized intervention adopted a deep neural network model, Bidirectional Encoder Representations from Transformers (BERT), to automatically classify online discussion transcripts and provide classification results in real time. Personalized feedback and recommendations were automatically generated from the classification results. A quasi-experimental research design was adopted to examine the effects of the proposed personalized intervention approach on collaborative knowledge building, group performance, socially shared metacognitive regulation, and cognitive load. Sixty-six college students participated in this study and were randomly assigned to the experimental and control groups. For online collaborative learning, students in the experimental group adopted the personalized intervention approach, whereas those in the control group used the conventional approach. Both quantitative and qualitative research methods were adopted to analyze data. The results indicated significant differences in the level of collaborative knowledge building and group performance between the experimental and control groups. Furthermore, the experimental group demonstrated more socially shared metacognitive regulation than the control group. There was no significant difference in cognitive load between the experimental and control groups. The results obtained from interviews were consistent with the quantitative data. The main findings together with the implications for practitioners are discussed in depth.
Personalized intervention, Deep neural network, Collaborative learning, Knowledge building, Socially shared metacognitive regulation
Jian-Hua Han, Keith Shubeck, Geng-Hu Shi, Xiang-En Hu, Lei Yang, Li-Jia Wang, Wei Zhao, Qiang Jiang and Gautum Biswas
School of Information Science and Technology, Northeast Normal University, China // email@example.com
Department of Psychology, The University of Memphis, USA // Institute for Intelligent Systems, The University of Memphis, USA // firstname.lastname@example.org
Department of Psychology, The University of Memphis, USA // Institute for Intelligent Systems, The University of Memphis, USA // email@example.com
Department of Psychology, The University of Memphis, USA // Institute for Intelligent Systems, The University of Memphis, USA // School of Psychology, Central China Normal University, China // Electrical and Computer Engineering, The University of Memphis, USA // firstname.lastname@example.org
School of Psychology, Central China Normal University, China // email@example.com
Institute for Intelligent Systems, The University of Memphis, USA // Electrical and Computer Engineering, The University of Memphis, USA // firstname.lastname@example.org
School of Information Science and Technology, Northeast Normal University, China // email@example.com
School of Information Science and Technology, Northeast Normal University, China // firstname.lastname@example.org
Institute for Software Integrated Systems, Vanderbilt University, USA // email@example.com
Intelligent learning technologies are often applied within the educational industries. While these technologies can be used to create learning experiences tailored to an individual student, they cannot address students’ affect accurately and quickly during the learning process. This paper focuses on two core research questions. How do students regulate affect and what are the processes that affect regulation? First, this paper reviews the affect regulation methods and processes in an intelligent learning environment based on affective transition and affect compensation. This process, along with affect analysis, affect regulation, intelligent agents, and an intervention strategy can be used to analyze specific affect regulation methods and improve the affective regulation system. Seventy-two 7th grade students were randomly placed into an experimental condition that used Betty’s Brain, an intelligent tutoring system (ITS), or a classroom control. A lag sequence analysis and a multinomial processing tree analysis of video data captured at 25-minute intervals revealed significant differences in affect transitions frequencies between the two groups. Based on the results of the above analyses and after-class interviews, we found that Betty’s Brain was able to promote effective affect-regulation strategies to students in the domain of forest ecosystems.
Teachable agent, Affect, Regulation, Tutoring, Betty’s Brain
Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province, Zhejiang Normal University, Jinhua, China // School of Information Technology in Education, South China Normal University, Guangzhou, China // firstname.lastname@example.org
chool of Information Technology in Education, South China Normal University, Guangzhou, China // email@example.com
Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province, Zhejiang Normal University, Jinhua, China // firstname.lastname@example.org
chool of Information Technology in Education, South China Normal University, Guangzhou, China // email@example.com
Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province, Zhejiang Normal University, Jinhua, China // firstname.lastname@example.org
Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province, Zhejiang Normal University, Jinhua, China // email@example.com
Collaborative reflection (co-reflection) plays a vital role in collaborative knowledge construction and behavior shared regulation. Although the mixed effect of online co-reflection was reported in the literature, few studies have comprehensively examined both individual and group factors and their relationships that affect the co-reflection level. Therefore, this study explored the structural relationships between achievement goals (task-based, self-based, and other-based goals), online community identification, and co-reflection, which can consequently assist instructors in improving the related pedagogical strategies. To this end, 26813 posts on MOOC and college online learning platforms were gathered. Specifically, deep learning techniques were first used to train a classifier that classifies the large-scale co-reflection text automatically. The Bayesian method was then applied to disclose the structural relationships among achievement goals, community identification, and co-reflection. The results showed that the proposed classification algorithm achieved the best performance. Two best-fit models for characterizing the respective relationships between co-reflection and community identification as well as achievement goals were obtained using the Bayesian method. The results of the experiments on these two models demonstrated that both task-avoidance and other-avoidance goals were related directly to co-reflection, all task-approach, self-approach and other-approach goals were related indirectly to co-reflection, but self-avoidance goals had both a direct and an indirect relationship with co-reflection. The relationship between community identification and co-reflection was mediated by other-based goals. Some theoretical and practical implications were discussed for instructors and practitioners to build an online community.
Deep learning, Bayesian network, Achievement goals, Co-reflection, Community identification
Chun-Hung Lin, Chih-Chang Yu, Po-Kang Shih and Leon Yufeng Wu
Center for Teacher Education, Chung Yuan Christian University, Taiwan // firstname.lastname@example.org
Department of Information and Computer Engineering, Chung Yuan Christian University, Taiwan // email@example.com
Center for General Education, Chung Yuan Christian University, Taiwan // firstname.lastname@example.org
Leon Yufeng Wu
Graduate School of Education, Chung Yuan Christian University, Taiwan // email@example.com
This article describes STEM education with artificial intelligence (AI) learning, particularly for non-engineering undergraduate students. In the designed three-week learning activities, students were encouraged to put their ideas about AI into practice through two hands-on activities, utilizing a provided deep learning-based web service. This study designed pre-test and post-test surveys to investigate the performance of students in different aspects of AI. With 328 students involved in these learning activities, we discovered from the surveys that the proposed learning method can effectively improve AI literacy among non-engineering students. This study also found that students’ AI literacy correlated significantly with their awareness of AI ethical issues and that the STEM-based AI curriculum increased the awareness of AI ethical issues among low-AI-literate learners. This article discusses the association between learning activities and different aspects of AI learning. The proposed method can be used by teachers who want to introduce AI knowledge into general education courses.
Artificial intelligence, STEM education, General education, Non-engineering students, Artificial intelligence literacy
Fengying Li ,Yifeng He and Qingshui Xue
School of Continuing Education, Shanghai Jiaotong University, Shanghai, China // firstname.lastname@example.org
Division for Development of Liberal Arts, Shanghai Jiaotong University, Shanghai, China // email@example.com
School of Computer Science and Information Engineering, Shanghai Institute of Technology, Shanghai, China // firstname.lastname@example.org
With the deep application of artificial intelligence and big data in education, adaptive learning has become a new research hotspot in online education. Based on the systematic review of the connotation and research progress of adaptive learning, a new definition of adaptive learning is given. By literature analysis, this paper points out the challenges faced by adaptive learning research, such as the lack of cognition of brain and technology, the bottleneck of the model of emotion domain, the separation of education and technology, the security of data management and the risk of privacy leakage. These challenges can be summarized into two aspects: one is mechanical issues, the other is safety issues. Different from traditional research perspectives, the paper opens a new research window, and puts forward countermeasures from the perspectives of cognitive principles, zone of proximal development theory in technology, breakthrough in the emotional domain model, learning data management and privacy security. In view of the centralization of learning data management nowadays, the concept of code chain and the decentralized management mode based on code chain are proposed. Different from the traditional adaptive learning recommendation technology, a new adaptive learning pulling model is proposed.
Adaptive learning, Learning recommendation, Learning pulling, Code chain technology, Data security
Chuang-Kai Chiu and Judy C. R. Tseng
College of Teacher Education, Wenzhou University, China // email@example.com
Judy C. R. Tseng
Department of Computer Science and Information Engineering, Chung Hua University, Taiwan // firstname.lastname@example.org
Awareness of students’ learning status, and maintaining students’ focus and attention during class are important issues in classroom management. Several observation instruments have been designed for human observers to document students’ engagement in class, but the processes are time-consuming and laborious. Recently, with the development of artificial intelligent technologies, artificial intelligence in education (AIED) has become an important research topic. Several studies have applied image recognition technologies to determine students’ learning status. However, little research has employed both sensor technology and image recognition technology in learning status analysis. Moreover, it remains unknown if learning status analysis is accurate enough to substitute for human observers. Furthermore, no feedback has been provided individually to students to manage their learning status by maintaining their attention in class. In this paper, a learning status management system in an intelligent classroom is proposed. Several types of information about students were detected and collected by both sensor technology and image recognition technology, and a Bayesian classification network was employed to inference the students’ learning status. Moreover, the system includes a feedback mechanism, which not only provides the results of the just-in-time learning status analysis to teachers, but also notifies students who are detected as being unfocused in class. Two experiments were conducted to verify the accuracy and effectiveness of the proposed system. Results showed that the learning status analysis highly corresponded to the observation of human beings, and the students were more attentive in class.
Classroom management, Intelligent classroom, Learning status analysis, Bayesian classification network