Assessing approaches to learning with nonparametric multidimensional scaling.

Area of Science:

• Educational Technology
• STEM Education
• Human-Computer Interaction

Background:

• Online gaming environments offer novel platforms for educational engagement.
• NASA Mars Mission STEM games in Whyville target middle school learning objectives.
• Understanding student learning approaches in digital environments is crucial.

Purpose of the Study:

• To assess learning approaches of middle schoolers using NASA STEM games in Whyville.
• To validate the use of trace data analysis for game-based learning research.
• To identify differences in learning strategies for curriculum improvement.

Main Methods:

• Trace-based assessment of 1361 interactions with NASA STEM games.
• Nonparametric multidimensional scaling for data analysis and visualization.
• Epistemic Network Analysis toolkit for post hoc data exploration.

Main Results:

• Distinct differences in learning approaches were identified among students.
• Multidimensional scaling proved effective for analyzing time-based behavioral data.
• The methodology provided statistical validation and visual examination of learning patterns.

Conclusions:

• The study demonstrates the value of trace data analysis in game-based STEM education.
• Findings can inform the design of online STEM content and educational games.
• Evidence-centered design principles are applicable to post hoc analysis of learning interactions.