Scholarlink

<p>Revealing the dynamics of collaborative engagement and its association with collaborative problem-solving (CPS) outcomes is beneficial for guiding teachers to provide adaptive support during CPS processes. Although existing research has indicated that the various dimensions of collaborative engagement influence each other over time, it remains unclear how the internal dynamics of collaborative engagement are associated with CPS outcomes. Additionally, previous studies have&#xa0;primarily employed manual coding to analyze collaborative engagement, which limits the exploration of the dynamic characteristics of collaborative engagement in large-scale datasets. This study aimed to investigate the internal dynamics of collaborative engagement, encompassing behavioral, cognitive, socio-emotional engagement, and their relationship with CPS outcomes. Specifically, a learning analytics method based on deep learning models was employed for the automatic detection of collaborative engagement. Subsequently, according to the results of automated detection, hidden Markov modeling was utilized to investigate the difference in the internal dynamics of collaborative engagement between high- and low-performing groups. These investigations were grounded in a dataset comprising 57,400 utterances collected from 20 groups participating in a CPS activity within a university setting. The findings showed that compared with high-performing groups, low-performing groups were more likely to become stuck in the states of "Limited Cognitive Engagement" and "Lone Neutral Participation" during CPS process. Furthermore, high-performing groups were found to transition away from the collaborative engagement state of "Cognitive Conflict with Confusion" by enhancing behavioral engagement and deepening cognitive engagement. While the low-performing groups remained trapped in the cycle between the states of "Cognitive Conflict with Confusion" and "Limited Cognitive Engagement." According to these findings, pedagogical insights and analytical implications were addressed.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exploring the internal dynamics of collaborative engagement in high- and low-performing collaborative problem-solving groups

  • 期刊论文

  • 发表日期：

    2025年8月2日

查看全文
  • Wanqing Hu,
  • Yanyan Li,
  • Ruiyan Huang,
  • Rushi Gong,
  • Xin Li

摘要

Revealing the dynamics of collaborative engagement and its association with collaborative problem-solving (CPS) outcomes is beneficial for guiding teachers to provide adaptive support during CPS processes. Although existing research has indicated that the various dimensions of collaborative engagement influence each other over time, it remains unclear how the internal dynamics of collaborative engagement are associated with CPS outcomes. Additionally, previous studies have primarily employed manual coding to analyze collaborative engagement, which limits the exploration of the dynamic characteristics of collaborative engagement in large-scale datasets. This study aimed to investigate the internal dynamics of collaborative engagement, encompassing behavioral, cognitive, socio-emotional engagement, and their relationship with CPS outcomes. Specifically, a learning analytics method based on deep learning models was employed for the automatic detection of collaborative engagement. Subsequently, according to the results of automated detection, hidden Markov modeling was utilized to investigate the difference in the internal dynamics of collaborative engagement between high- and low-performing groups. These investigations were grounded in a dataset comprising 57,400 utterances collected from 20 groups participating in a CPS activity within a university setting. The findings showed that compared with high-performing groups, low-performing groups were more likely to become stuck in the states of "Limited Cognitive Engagement" and "Lone Neutral Participation" during CPS process. Furthermore, high-performing groups were found to transition away from the collaborative engagement state of "Cognitive Conflict with Confusion" by enhancing behavioral engagement and deepening cognitive engagement. While the low-performing groups remained trapped in the cycle between the states of "Cognitive Conflict with Confusion" and "Limited Cognitive Engagement." According to these findings, pedagogical insights and analytical implications were addressed.