Investigating Sequence Patterns of Collaborative Problem-Solving Behavior in Online Collaborative Discussion Activity

Yafeng Zheng, Haogang Bao, Jun Shen, Xuesong ZhaiSustainability 2020 · DOI

TL;DR

This paper introduces a novel multidimensional analysis schema and an automated sequential analysis method to better understand collaborative problem-solving (CPS) behaviors in online discussions. By combining temporal and sub-problem sequences, the study reveals that deep, exploratory discussions driven by conflicting viewpoints and evidence-based rationalization lead to improved outcomes, offering a more nuanced view of CPS than previous social or cognitive-only approaches.

What Problem Does This Paper Address?

Collaborative problem-solving (CPS) is a critical 21st-century skill, especially in online learning environments. While existing research has explored CPS through social or cognitive lenses, there's a gap in understanding the interactive dynamics. Furthermore, traditional sequence analysis methods often struggle to integrate both the temporal flow of discussions and the progression through different sub-problems within a task. This makes it difficult to capture the full complexity of how groups collaboratively solve problems. The authors aim to bridge this gap by developing a more comprehensive approach to analyze CPS behavior in online discussions.

Methods

The core of this study lies in its novel methodology. The authors propose a multidimensional analysis schema for CPS, integrating both cognitive and social dimensions. This schema is then operationalized through an automated sequential analysis method. This method combines:

  • Time sequences: Analyzing the order of events as they occur.
  • Sub-problem sequences: Accounting for the progression through different stages or aspects of the problem being solved.

This combined approach allows for a more granular understanding of behavioral patterns. The discussion transcripts were coded using a detailed schema that categorizes behaviors into "Statement," "Negotiation," "Asking questions," "Management," and "Sharing feelings," with further subcategories. An improved Apriori algorithm was used to mine frequent sequential patterns from these coded behaviors.

Datasets

The study utilized discussion transcripts from an online collaborative learning activity involving 191 undergraduate students from two universities in northern China. These students were organized into 38 teams of four to six students each. The data collected was from a compulsory "Data Structure" course, where students worked on designing an algorithm for a collection management program. The discussion data was collected over a 2-hour period.

Key Insights and Ideas

  • Beyond Social/Cognitive Dichotomy: The study emphasizes that effective CPS is not just about individual cognitive contributions or social interactions, but the dynamic interplay between them.
  • Conflict as a Catalyst: Deep exploratory discussions, particularly those arising from conflicting viewpoints, are crucial for promoting better problem-solving outcomes and perceptions.
  • Evidence-Based Rationalization is Key: The use of evidence to support arguments and rationalize decisions is a strong motivator for effective collaborative behavior.
  • "Statement" Dominance: The "Statement" category (proposing, explaining, and revising opinions) was the most frequent behavior, highlighting the importance of knowledge construction through discourse.
  • "Revise Opinions/Solutions" as an Indicator: The frequency of revising opinions and solutions was identified as a significant indicator of high-level knowledge construction and group performance.
  • Management Matters: Effective group management (coordinating, organizing) is crucial for high-performing groups, preventing them from getting bogged down in administrative issues and allowing for more productive discussions.

Key Results

  • Behavioral Distribution: The most frequent behaviors observed were "Statement" (46.78%), "Asking questions" (21.84%), and "Negotiation" (17.52%).
  • High vs. Low Performance: High-performance groups showed a higher proportion of "Disagree, give evidence" (C24) and "Agree, give evidence" (C22) compared to low-performance groups. They also demonstrated a higher frequency of "Revise opinions/solutions" (C13).
  • Sequential Patterns:
    • High-performance groups exhibited frequent transitions like "Propose opinions/solutions" (C11) → "Revise opinions/solutions" (C13) and "Disagree, give evidence" (C24) → "Propose opinions/solutions" (C11).
  • Low-performance groups, in contrast, showed a tendency to get stuck in management loops (e.g., "Organize/assign tasks" (C41) → "Organize/assign tasks" (C41)), indicating ongoing management challenges.
  • Evidence Use: High-performance groups used evidence more frequently in negotiation, suggesting a deeper engagement with the quality of arguments.

Limitations and Future Work

  • Single Course Context: The study's data mining method was validated in only one collaborative discussion course. Future research should replicate and extend these findings across different collaborative discussion activities.
  • Epistemic Dimension: The epistemic dimension of CPS (how students construct new knowledge) was not deeply explored. Future work could investigate this aspect to understand how students respond to learning challenges.
  • Group Composition: The influence of group composition and size on CPS processes was not extensively studied. Future research could explore these factors.
  • Behavioral Transformation Stages: Investigating behavioral transformation modes at different stages of activities could provide a deeper understanding of strategies.
  • Intervention Strategies: The authors suggest using these findings to develop appropriate intervention strategies and timing to support students' problem-solving and knowledge construction.

Summary updated

Abstract

Collaborative problem solving (CPS) is an influential human behavior affecting working performance and well-being. Previous studies examined CPS behavior from the perspective of either social or cognitive dimensions, which leave a research gap from the interactive perspective. In addition, the traditional sequence analysis method failed to combine time sequences and sub-problem sequences together while analyzing behavioral patterns in CPS. This study proposes a developed schema for the multidimensional analysis of CPS. A combination sequential analysis approach that comprises time sequences and sub-problem sequences is also employed to explore CPS patterns. A total of 191 students were recruited and randomly grouped into 38 teams (four to six students per team) in the online collaborative discussion activity. Their discussion transcripts were coded while they conducted CPS, followed by the assessment of high- and low- performance groups according to the developed schema and sequential analysis. With the help of the new analysis method, the findings indicate that a deep exploratory discussion is generated from conflicting viewpoints, which promotes improved problem-solving outcomes and perceptions. In addition, evidence-based rationalization can motivate collaborative behavior effectively. The results demonstrated the potential power of automatic sequential analysis with multidimensional behavior and its ability to provide quantitative descriptions of group interactions in the investigated threaded discussions.
BibTeX
@inproceedings{zheng2020investigating,
  author = {Yafeng Zheng and Haogang Bao and Jun Shen and Xuesong Zhai},
  title = {{Investigating Sequence Patterns of Collaborative Problem-Solving Behavior in Online Collaborative Discussion Activity}},
  booktitle = {Sustainability},
  year = {2020},
  doi = {10.3390/su12208522},
}

Loading related content...

Figure