Charting the Future of AI in Project-Based Learning: A Co-Design Exploration with Students

Title of the Paper

Charting the Future of AI in Project-Based Learning: A Co-Design Exploration with Students

Paper Information

• Subject Area: Educational Technology and Artificial Intelligence
• Keywords: AI in Education, Project-Based Learning (PBL), Co-Design, Qualitative Research, Generative AI

Research Background and Problem

• Identified Issues and Challenges: Students are increasingly using Artificial Intelligence (AI) in Project-Based Learning (PBL), posing new assessment challenges for educators, such as how to fairly evaluate student learning outcomes and avoid merely measuring AI capabilities instead of students' abilities.
• Significance: PBL focuses on students' ability to solve real-world problems, with its processes and outcomes showcasing higher-order thinking skills. However, AI involvement may alter traditional outcome-based assessment methods, impacting future educational goals and skill development.
• Motivation and Related Work: While there is extensive research on AI applications in education, few studies explore how student-AI interaction data can inform educational assessments. Additionally, the rise of generative AI introduces new demands and directions for educational processes.

Solution

• Methods and Solutions:

  • The paper designed a series of co-design workshops, collaborating with 18 university students to explore how AI can enrich PBL scenarios and how AI data can support learning assessment.
  • Students were encouraged to envision potential future AI use cases and, through design activities, create AI usage reports to support assessment.

• Innovations:

  • Deeply integrating generative AI with educational assessment, the study innovatively proposed student-centered designs for future PBL AI data reporting models.
  • Explored how student-AI interaction data could be used to analyze higher-order thinking skills, moving beyond traditional outcome-based assessment methods.

• Implementation Steps and Key Techniques:

  1. Designing Co-Creation Exploration Activities: Included reviewing past PBL experiences, envisioning AI-enhanced learning processes, and designing personalized AI usage reports.
  2. Activity Framework: Three stages—envisioning future scenarios of AI in learning, imagining ideal future student assessment standards, and designing visual reports to showcase AI usage data.
  3. Data Analysis Techniques: Utilized qualitative coding and thematic analysis to process data generated during the workshops, with triangulation ensuring the accuracy of research findings.

Research Findings

Specific Findings and Advantages

1. Students' Imagined AI Applications:

   • Automating repetitive and time-consuming tasks, such as data collection, debugging, and documentation.
   • Supporting divergent thinking and creative generation.
   • Providing options and feedback to help students directly implement solutions.
   • Guiding students in learning new knowledge, even taking on partial teaching roles.

2. Transformations in Future Educational Assessment:

   • Proposed new methods for assessing traditional skills (e.g., re-evaluating creative thinking) and standards for assessing new skills (e.g., effective use of AI).
   • Emphasized analyzing students' ability to leverage AI functionalities and their leadership roles in projects.

3. Practical Design of Student AI Usage Reports:

   • Reports designed by students included analyses of task allocation across different project stages, the effectiveness of student-AI interactions, and how students integrated AI suggestions into project processes.
   • Highlighted the influence of different AI roles (tool, teammate, or expert) on report design.

Experimental or Evaluation Results

• Results Display: Seven workshops generated six key sub-themes of AI usage scenarios and seven analysis themes related to student-AI interactions (e.g., task allocation, interaction effectiveness).
• Diverse Report Designs: Students used pie charts, flowcharts, cumulative tables, and other design methods to summarize AI contributions, showcasing their ability to calculate and understand skill development.
• Limitations and Future Directions:
  • The current study involved 18 participants over a short period, lacking large-scale and long-term validation.
  • Future research could incorporate more teacher perspectives and conduct long-term PBL experiments with actual AI usage.

Summary and Discussion

• This paper proposed a novel perspective of using student-AI interaction data to assess learning, advancing the exploration of AI-enhanced education.
• The study provided qualitative insights into students' participation in designing and imagining future learning scenarios, emphasizing their active role in educational data analysis.
• It outlined future research directions for educational technology and HCI, including customized AI tool design, promoting students' self-regulated AI learning, and evaluating the effectiveness of student-AI collaboration.

Recommendations for Next Steps

• Further investigate differences between student and teacher perspectives, exploring how to integrate diverse needs into comprehensive AI data reporting tools.
• Combine quantitative research and long-term practice, validating findings with more educational scenarios and large-scale student data.