Exploring Initiative as a Signal of Knowledge Co-Construction During Collaborative Problem Solving.

Area of Science:

• Educational Technology
• Computer Science Education
• Cognitive Science

Background:

• Peer interaction supports learning, with knowledge co-construction (KCC) as a key mechanism.
• KCC is abstract, posing challenges for developing supportive instructional software.
• Identifying measurable indicators of KCC is crucial for software development.

Purpose of the Study:

• To analyze peer dialogs in introductory Computer Science to find measurable indicators of KCC.
• To investigate the relationship between task initiative shifts, KCC, and learning outcomes.
• To explore the potential of task initiative shifts for modeling productive peer collaboration in software.

Main Methods:

• Analysis of a corpus of peer dialogs from introductory Computer Science courses.
• Identification and quantification of "task initiative shifts" within dialogs.
• Correlation analysis between task initiative shifts, KCC episodes, and student learning (pre-test scores).

Main Results:

• Task initiative shifts correlate positively with both KCC episodes and learning.
• Task initiative shifts occur more frequently during KCC episodes.
• Task initiative shifts within KCC episodes correlate with learning for low pre-testers; total shifts correlate for high pre-testers.

Conclusions:

• Task initiative shifts serve as a measurable indicator of productive peer collaboration and KCC.
• Task initiative shifts are a more accessible construct than KCC for modeling in instructional software.
• This research provides a foundation for developing intelligent tutoring systems that foster collaborative learning.