Elsevier

Computers & Education

Volume 156, October 2020, 103938
Computers & Education

Sustaining online academic discussions: Identifying the characteristics of messages that receive responses

https://doi.org/10.1016/j.compedu.2020.103938Get rights and content

Highlights

  • Characteristics of messages matter in eliciting responses in online discussions.

  • Messages that disagree with previous ideas are likely to receive responses.

  • Messages that contain correct or wrong ideas are likely to receive responses.

  • Messages that ask questions are likely to receive responses.

  • Prompt replies and early-stage messages are likely to receive responses.

Abstract

More and more students are learning via online academic discussions, posting messages in an attempt to discuss their learning problems. However, many messages do not receive responses. Posting messages that elicit responses is essential to students' experiences of learning through online discussions, but the characteristics of such messages are seldom studied. To fill this gap, this paper examines the relationship between the characteristics of an online discussion message and its likelihood of receiving a response from others. We conducted the study with a public, online discussion forum about high school-level mathematics—a non-formal learning environment that is not confined to a specific classroom. We randomly sampled 140 topics from the forum and analysed 1,559 reply messages using multilevel logistic regressions at the topic and message level. We found that during an online discussion, a message that either expressed disagreement, included a correct or incorrect idea, or asked a question was more likely to receive a response. Time was another significant predictor; messages posted during the early stage of a discussion or users who responded more promptly were more likely to receive a response. The findings contribute to the understanding of the discourse process and students’ learning behaviour in online academic discussions. We propose several recommendations for future research.

Introduction

Thanks to advances in technology, students can interact with other students in online communities to seek information or discuss questions (Aloni & Harrington, 2018; De Wever, Schellens, Valcke, & Van Keer, 2006; Woo & Reeves, 2007). There are multiple advantages to online discussions, such as open communication, supportive collaboration, information exchange, and the connection of ideas (Garrison, 2007). It has long been recognised that although learners working together may generate cognitive conflicts (Piaget, 1974), this can enable them to solve problems at a more advanced level than if they worked on the same problems alone (Doise, Mugny, & Perret-Clermont, 1975). For this reason, online discussion forums have been used to support student learning in various educational contexts (e.g., Aloni & Harrington, 2018; Wise & Cui, 2018).

Our research focus is on the characteristics of messages that elicit responses in online academic discussions because they are central to sustaining online discussions and developing online communities (Anderson, 2006; Ridings, Gefen, & Arinze, 2002; Tsai & Pai, 2013). Messages that elicit responses increase the lifespan of a discussion, the participants' sense of belonging, and their engagement in online communities (Lee, Reid, & Kim, 2014; Lewallen, Owen, Bantum, & Stanton, 2014; Tsai & Pai, 2013). They can also reinforce the contribution of knowledge to online communities (Jin, Li, Zhong, & Zhai, 2015; Kim & Sundar, 2014; Tausczik & Pennebaker, 2012). In addition, participants can gain a deeper understanding of the learning materials by reviewing and commenting on online messages (Aloni & Harrington, 2018; Cathey, 2007; Tsai & Pai, 2013). It is therefore important to identify those characteristics that affect a message's likelihood of receiving a response in online academic discussion communities.

Past studies have identified various motivational factors that affect online participation and knowledge-sharing behaviours across a range of online communities (e.g., Jin et al., 2015; Joyce & Kraut, 2006; Kim & Sundar, 2014; Wang & Lai, 2006). For example, a survey study by Lee et al. (2014) provided evidence that higher levels of online authors’ sense of belonging would lead to higher levels of knowledge-sharing activities. However, we currently know little about the characteristics of a message per se that drive post-replying behaviour in online academic discussion communities.

The present study seeks to understand how students engage other students to respond during online discussions, and how to enhance discussions in online forums. We create a detailed model of actions and sequences across time that affect others' engagement in online discussions, thereby informing current models of motivation and informing online teaching to enhance students' online discussions. Building on past studies that primarily examine post hoc surveys of motivation, this study examines how sequences of students’ behaviours affect their likelihood of responding in an online discussion.

We study discussions in an online community forum as a form of non-formal learning. Specifically, we analyse high school students' mathematics discussions from an online public forum that is not attached to any class or school. Examining students’ interactions in independent forums not confined to the classroom can improve our understanding of their spontaneous, natural behaviours when responding to one another during online discussions.

Section snippets

Research model and hypotheses

For this study, we adopted the theoretical work of cognitive-social theorists, such as Festinger (1957) and Heider (1946), which affirms that cognitive dissonance and imbalance generate a motivational tendency to resolve contradictory cognitions. We also used the theoretical model of helping behaviour as proposed by social psychologists (e.g., Latané & Darley, 1970; Yalom, 2005) to approach the characteristics of messages that elicit responses during online discussions. We developed a research

Online discussion forum and data

In this study, we collected and analysed data from an online discussion forum for high school mathematics, which is hosted by Art of Problem Solving (AoPS) Online in the US. As one of the largest mathematics communities on the Internet, AoPS Online aims to help students expand and deepen their mathematical thinking (artofproblemsolving.com). The high school mathematics forum is a non-formal learning environment (Schwier & Seaton, 2013) which is moderated but not facilitated by the AoPS team.

Results

Overall, the analysis results indicate that only the characteristics of a current message significantly predicted its likelihood of receiving a response. In other words, the impact of its previous messages along the same thread (e.g., lag 1 and lag 2) on post-replying behaviour was not significant. Also, a variance components model showed that the outcome variable responsiveness did not differ significantly across topics, so single-level modelling (message level) was adequate. The corresponding

Discussion

In this study, we adopted statistical discourse analysis to examine how a message's content-related characteristics might affect the likelihood of receiving a response during online academic discussions. We conducted the study in a non-formal learning environment—an open online high-school-mathematics forum that is not confined to a class or a school. The analysis results support the hypotheses about disagreement (H-1a), correct ideas (H-2a), incorrect ideas (H-2b), and questions (H-3a). Apart

Conclusion and implications

This study aimed to identify the content-related characteristics that affect a message's likelihood of receiving a response in online academic discussions on mathematics topics. The results suggest that how a message looks backward (i.e., disagrees with a previous message), posts knowledge content (i.e., adds a new idea, either correct or incorrect), and looks forward (i.e., asks a question) can affect the likelihood of receiving a response. Slow response messages or those posted in the later

CRediT authorship contribution statement

Gaowei Chen: Conceptualization, Methodology, Formal analysis, Writing - original draft. Chung Kwan Lo: Investigation, Validation, Writing - review & editing. Liru Hu: Data curation, Writing - review & editing.

Acknowledgements

This research was supported by Hong Kong Research Grants Council (RGC) grant No. 17608318 and the Public Policy Research Funding Scheme (grant No. 2017.A8.073.18C) from the Policy Innovation and Co-ordination Office of the Government of the HKSAR.

References (43)

  • R.P. Rooderkerk et al.

    No comment?! the drivers of reactions to online posts in professional groups

    Journal of Interactive Marketing

    (2016)
  • E. Sung et al.

    Five facets of social presence in online distance education

    Computers in Human Behavior

    (2012)
  • H.T. Tsai et al.

    Explaining members' proactive participation in virtual communities

    International Journal of Human-Computer Studies

    (2013)
  • A.F. Wise et al.

    Learning communities in the crowd: Characteristics of content related interactions and social relationships in MOOC discussion forums

    Computers & Education

    (2018)
  • Y. Woo et al.

    Meaningful interaction in web-based learning: A social constructivist interpretation

    The Internet and Higher Education

    (2007)
  • M. Aloni et al.

    Research based practices for improving the effectiveness of asynchronous online discussion boards

    Scholarship of Teaching and Learning in Psychology

    (2018)
  • S.R. Aragon

    Creating social presence in online environments

    New Directions for Adult and Continuing Education

    (2003)
  • Y. Benjamini et al.

    Adaptive linear step-up procedures that control the false discovery rate

    Biometrika

    (2006)
  • C. Cathey

    Power of peer review: An online collaborative learning assignment in social psychology

    Teaching of Psychology

    (2007)
  • M.M. Chiu

    Group problem‐solving processes: Social interactions and individual actions

    Journal for the Theory of Social Behaviour

    (2000)
  • M.M. Chiu

    Flowing toward correct contributions during group problem solving: A statistical discourse analysis

    The Journal of the Learning Sciences

    (2008)
  • Cited by (12)

    • The impact of role assignment on social presence in online discussions: A mixed-method study

      2023, Internet and Higher Education
      Citation Excerpt :

      Throughout the study, participants discussed three different cases (see Table 4) asynchronously on the university's learning management system (Moodle) within their small groups. The cases were selected from the book entitled Cases on Educational Technology Integration in Urban Schools (Chen & McPheeters, 2012). Three different roles were assigned to the participants in the experimental group weekly to provide them with the opportunity of experiencing different roles.

    • Learner behaviors associated with uses of resources and learning pathways in blended learning scenarios

      2022, Computers and Education
      Citation Excerpt :

      But there is no other evidence to support such a conclusion and further investigation is required. Observed behaviors associated with the uses of tools listed in Fig. 6 can be described as consistent with the uses of educational video (Mayer et al., 2020; Pappas et al., 2017; Turan & Cetintas, 2020), apps (Ping et al., 2018) and forums or message boards (Chen et al., 2020; Ruthotto et al., 2020) addressed in literature. The fact that online interaction took place only among students and no messages were posted onto the teachers' wall highlights the potential of peer collaboration in dealing with casually arising questions about the subject content.

    View all citing articles on Scopus
    View full text