Understanding Personal Data Tracking and Sensemaking Practices for Self-Directed Learning in Non-classroom and Non-computer-based Contexts

Title of the Paper

Understanding Personal Data Tracking and Sensemaking Practices for Self-Directed Learning in Non-classroom and Non-computer-based Contexts

Paper Information

• Subject Area: Artificial Intelligence and Educational Technology, specifically self-directed learning (context: non-classroom and non-computer-based scenarios)
• Keywords: Personal Informatics, Self-tracking, Non-classroom Learning, Non-computer-based Learning, Autonomous Learning, Behavior Change

Research Background and Issues

1. Identified Problems or Challenges:

   • Students engaged in self-directed learning may face challenges such as distractions, lack of motivation, and difficulty in conducting metacognitive analysis.
   • Current learning tracking technologies are primarily designed for classroom and computer-based learning scenarios, making them ineffective for supporting non-classroom and non-computer-based learning activities, such as tasks involving paper-based materials.
   • Existing systems lack contextual information and personalized learning feedback, focusing more on group data.

2. Significance of the Problem:

   • Autonomous learning is a critical theory for enhancing learning ability, problem-solving skills, and a proactive learning attitude. However, achieving effective autonomous learning in non-classroom environments, such as at home or in libraries, remains challenging.
   • The COVID-19 pandemic has underscored the importance of online and autonomous learning, exposing the limitations of traditional learning tracking tools.

3. Research Motivation:

   • There is a need to explore how to design learning tracking tools that adapt to non-classroom and non-computer-based learning activities.
   • Understanding what types of learning data hold personal significance for learners and how they utilize this data for reflection and improvement.

4. Research Objectives:

   • Investigate the motivations, data types, and reflective practices of students recording and interpreting learning data using Timing (a popular learning tracking application).
   • Provide design insights to improve future learning tracking tool designs.

Solutions

1. Methods and Research Design:

   • A qualitative approach was adopted, conducting semi-structured interviews with 24 Timing users.
   • Analyzed the types of data users tracked, their behaviors, and the reflective strategies they employed.

2. Innovations:

   • Emphasized learning tracking scenarios in non-computer and non-classroom environments.
   • Explored the role of qualitative data (e.g., emotions, photos, and video recordings) in promoting metacognition and reflection.
   • Proposed design recommendations to support self-directed learning and address the shortcomings of existing learning tracking technologies.

3. Implementation Steps and Key Techniques:

   • Users manually tracked learning data in Timing (e.g., study plans, learning videos, learning outcomes, and emotions).
   • Data was categorized progressively (e.g., pre-study planning, behavioral trajectories during learning, and learning outcomes), observing how users derived insights from diverse data types.

Research Findings

1. Specific Findings:

   • User Motivations: Primarily to support self-discipline and maintain learning motivation through data tracking.
   • Types of Tracked Data:
     • Study Plans: Including daily, weekly, monthly, or annual plans.
     • Learning Process: Time tracking and dynamic learning videos (e.g., study tasks, environment, facial expressions).
     • Learning Outcomes: Photos of completed tasks, notes, and accelerated versions of edited videos.
     • Emotions: Represented by markers (e.g., emojis, phrases) indicating different emotional states during learning.
   • Reflective Strategies:
     • Reviewing dynamic learning videos to observe behaviors and learning states.
     • Briefly browsing static photos of plans and outcomes.
     • Cross-referencing different data types to analyze learning challenges.
     • Manually recording reflective summaries for future reference.

2. Comparison with Existing Solutions:

   • Current learning tracking systems focus mainly on automated quantitative data, whereas this study highlights the unique value of qualitative data (e.g., learning videos, emotion records) in supporting deep reflection.
   • Emphasized the role of integrating multiple data types to enhance metacognition.

3. Experimental and Evaluation Results:

   • Data tracking and reflection using Timing significantly helped learners with self-management, self-monitoring, and self-adjustment.
   • Compressed video playback of learning behaviors inspired and enhanced users' learning motivation.

4. Limitations and Future Directions:

   • Limitations:
     • The sample was concentrated on Chinese students (particularly those preparing for high-stakes exams like college entrance or postgraduate exams), which may limit the generalizability of the findings.
     • Non-student users were not examined.
   • Future Directions:
     • Validate findings in broader, cross-cultural contexts, such as learners from other countries (e.g., South Korea, Japan).
     • Explore whether other learning tracking applications with different features can reveal new user behavior patterns.
     • Deepen the analysis of dynamic videos to optimize users' interpretation of rich contextual data.

Design Implications

1. Support for Emotional Data Recording and Recall:

   • Provide privacy-friendly emotion tracking designs while avoiding additional data collection burdens on users.

2. Balancing Learning Video Recording and Privacy Protection:

   • Allow users to control which sensitive information is recorded, offering automatic privacy masking features.

3. Facilitating Efficient Interpretation of Dynamic Videos:

   • Introduce delayed playback for certain summary videos and incorporate computer vision analysis to mark specific behaviors (e.g., signs of distraction).

4. Reducing the Burden of Cross-data Reflection:

   • Offer text or voice prompt features to help users integrate insights from multiple reflective data sources in real-time, enhancing the efficiency of secondary reviews.