Mapping multiple dimensions of student learning: the ConstructMap program.

Area of Science:

• Educational Measurement
• Psychometrics
• Educational Technology

Background:

• Traditional assessments often use items with single correct answers, limiting insight into student capabilities.
• Performance-based assessments offer richer data but present scoring and interpretation challenges, including potential bias.
• Advances in psychometrics and assessment design enhance the feasibility of evaluating performance-based tasks.

Purpose of the Study:

• To develop a tool assisting educators in interpreting and representing complex performance-based assessment data.
• To address the technical difficulties and potential biases associated with scoring performance tasks.
• To provide a method for understanding students' knowledge and application of that knowledge.

Main Methods:

• Development of the ConstructMap computer program.
• Inputting student scores on performance-based items.
• Utilizing multidimensional item response theory (IRT) models for proficiency computation.
• Generating graphical representations of student proficiencies.

Main Results:

• ConstructMap facilitates the interpretation of multidimensional performance data.
• The program computes student proficiencies using advanced psychometric models.
• Graphical outputs provide clear visualizations of student performance across variables.

Conclusions:

• The ConstructMap program enhances the effective and efficient assessment of student performance.
• It aids teachers in making informed interpretations of what students know and can do.
• This technology supports a more nuanced understanding of student learning beyond traditional metrics.