Process-Based Measurement Toward Understanding Learning with Applications

Resumen:

Process data arise from measuring a construct over time to better understand how a learner achieves some outcome (i.e., explain or predict outcome performance). It is useful for instructional or assessment purposes.  For example, it may be used to adapt and individualize instruction. It may provide feedback to the learner or the instructor. It may also diagnose strengths and weakness in learning.  We first motivate the analysis of process data with a project in the area of simulation training of medical professionals.  We then discuss the application of process data analysis in the K-12 setting.  As part of a larger project developing enhanced assessment tools for measuring career readiness indicators, we developed several innovative assessment tasks that uses technology to improve our ability to infer high school students problem solving skills.  We instrumented these game-like tasks and implemented telemetry to collect in-game process information from a small pool of participants.  We combine the Bayes net and item response theory modeling as our data analytic approach.  We show that the resulting scores are strongly related to an externally administered reasoning measure.  We discuss implications for assessment development, practice, and research.

Posted on Aug 1, 2018 in Education, Seminars