Masami Yoshida

As online education continues to gain traction, it is imperative to investigate how interactions within the learning community contribute to individual competencies. There is a growing trend to apply social network analysis to the learning network to study student development. As a package of computational and statistical methods for analyzing social networks, exponential family random graph models (ERGMs) explore intricate network structures based on the relational data within the network. This scoping review examined articles that employed ERGMs and summarized the analysis of emergent communities. Scopus and ScienceDirect were used for the literature search, resulting in 11 articles that met the inclusion and exclusion criteria. Four categories of data extraction schemes were adopted: bibliometrics, analysis design, network profile, and network structure. The underlying framework observed in online discussions was predominantly mutual, with ease of expansion and subsequent formation of a star structure. However, the presence of a triangle structure did not manifest significantly. Conversely, the connections displayed a strong linkage to the broader network, thus forming a small-world structure. The results hold potential for future educational research that elucidates online learning interactions based on students’ relational data.

As the popularity and accessibility of artificial intelligence (AI) in learning continue to rise, online education has increasingly incorporated AI. It is imperative to investigate how AI in the scholarly community supports the learning of autonomous students. This scoping review investigated articles that employed AI in discussion forums. The objective of this study is to develop a summary that denotes the effective integration of AI in discussion forums, thereby characterizing the overall learning experience. In accordance with the search, appraisal, synthesis, and analysis (SALSA) framework, the preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram was employed to design the article screening and selection process. Scopus was utilized for the literature search, resulting in the identification of twelve articles that met the predetermined inclusion and exclusion criteria. The extracted articles covered the following six applications: intelligent tutoring systems, AI-assisted apps, chatbots, automatic formative assessment, AI teaching assistants, and AI-assisted lifelong learning assistants. Three types of man-machine partnerships related to the activities are discussed: student-AI partnerships, student-AI-student partnerships, and student-AI-teacher partnerships, the latter being the most commonly observed. The generated content was frequently processed through teacher adjustments and checks. Components of community of inquiry theory were adapted as indicators. Cognitive presence was observed to be more predominant than social presence, with a particular emphasis on individual learning achievements. Three distinct perspectives regarding the role of AI in learning were identified. The first perspective posited that AI functioned as a virtual teacher, supplementing the teaching functions of human teachers and delivering information to students. In the second perspective, AI facilitated students in collecting information by enabling interactions with messages or peers. In the third perspective, the significance of the automatic evaluation function, which assessed learning activities and provided feedback to enhance the system while simultaneously implementing immediate changes in the learning process, was acknowledged. Regarding the role of communication, teachers clearly intended for students to acquire knowledge through it. Despite the focus on assisting individual learners, each article encompassed references from a wide range of fields and exhibited substantial diversity overall. Consequently, the domain of discussion forums with AI was depicted through the complex interplay between technology, pedagogy, and the learning environment. This context is known as entangled pedagogies, which emphasize understanding the interconnections among these elements and their reciprocal influences. In summary, AI was predominantly utilized in the form of chatbots and GenAI in discussion forums. Emerging AI roles included virtual teachers, peer/message matching, and automated evaluation. Aggregated information was presented in a manner that supported student learning. AI also served as a conduit between students and teachers.

INPLASY202470039 The application of exponential random graph models to online learning networks: a scoping review protocol Keywords: ERGMs, learning, online discussion, SALSA, SIENA, review. Downlo…

Following the COVID-19 pandemic, online technologies have been introduced to facilitate students' engagement in education. Online discussion has been used to provide them a communal learning environment. This investigation examined online scholarly discussions conducted by university students with a view to ascertaining how they demonstrated critical thinking (CT). Course participants were tasked with submitting research reports related to assignments to the Moodle discussion forum. These were subsequently shared with their student peers. After the ensuing discussions were recorded, the statements made by the students were coded using subskill indicators of CT. A code sequence analysis was subsequently conducted, with CT and altruism serving as dependent variables. In total, 121 students contributed 310 messages to the Moodle forum. In cases involving CT, lower-level skills were found to be predominant among the code sequences. In contrast, cases involving altruism displayed sequences that transitioned from higher to lower skills. When both dependent variables were active, the highest count of sequences was recorded for self-regulation skill, indicating higher-order cognitive skill activities. The students recognized their cognitive limitations, actively sought additional information, and re-evaluated the topic. This study establishes a novel analytical methodology for evaluating critical thinking skills. Received: 18 February 2023 / Accepted: 21 April 2023 / Published: 5 May 2023

As a pedagogical method to improve student recognition of Education for Sustainable Development (ESD) through scientific data reading, this study looked at the effects of learning in online academic discussions using data from global indicators. Students’ scholarly messages were coded and introduced into qualitative content analysis, sequential analysis, and social network analysis, which are emphasized, respectively, to investigate code co-occurrence, code sequence, and code distribution. In all, 307 messages appeared from 119 university students in the online community. The ESD competencies and collective intelligence (CI) are used as indicators for analyses. Qualitative content analysis, particularly addressing those sentences, proved that CI enhanced communication among students where they shared individual norms and values. Sequential analysis elucidated characteristics of discussion thread characteristics with CI, which induced further discussion with foresight views and questions. Social network analyses indicated students connected and showed the connection structure was meshed. Key student bridging messages were extracted. Whereas ESD competencies appeared effectively, the expansion of the current online environment must be regarded as including competency in participatory learning. After summarizing the effects of the online learning method in the Moodle forum environment, the method was proved to empower students to represent core competencies of ESD and to lead data-driven concept transformation.

In pedagogical practice, gratitude is recognised not as an emotion, but as an approach to learning. This study introduced gratitude messages into the academic online communication of university students and specifically examined the community in which students shared their messages with gratitude. This study examined the tendency of message connections and how gratitude messages prompted replies. To elucidate their connections, exponential random graph models (ERGMs) were used. A post-event questionnaire to evaluate gratitude experiences was also administered. Results revealed that 77.3% of the 172 connected messages from 123 students involved gratitude. When the post-event questionnaire results were examined using an ERGM, the score effects on increasing message connections were found not to be significant. The most prominent indication was a higher level of significant propensities to make mutual connections. The homophily of the message content was found to have a significant propensity to increase connections. The ERGM results and a review of messages revealed that students expressed gratitude for being both benefactors and beneficiaries of gratitude messages, which confirmed their prosocial behaviour.

Information about the paper titled "ANALYSIS OF COLLECTIVE INTELLIGENCE IN A DISCIPLINARY COMMUNITY" at IATED Digital Library

As part of an applied activity of a university course in Japan, a total of 65 students posted messages on Twitter regarding global issues to expand the scope of scholarly communication with social users on Twitter. The students’ motivations to learn about world issues were evaluated using a standard questionnaire developed under the attention, relevance, confidence, and satisfaction (ARCS) model. Thereafter, students’ connections were categorized into three types: crossing, private, and isolation. A one-way analysis of variance test and multiple comparison test showed a significant result—students with higher levels of motivation could communicate with users in other communities of the crossing type. In this study, we tried for the first time to utilize social network analysis and exponential random graph models to investigate the motivation and structures of scholarly connections on Twitter. Although students selected social users as targets for tweeting, most students who had reciprocal ties with these users did not contribute to the expansion of the community. Since the requirement to be a popular user was explicated by focusing on a remarkable student in the community, the inner-community enhancement of increasing ties among students was concluded for developing social capital in the community.