Simulations of Thought: The Role of Computational Cognitive Models in Assessment ACT-R CBT AD
- Author(s):
- Moon, Jung Aa; Finn, Bridgid; LaMar, Michelle M.; Katz, Irvin R.
- Publication Year:
- 2018
- Report Number:
- RDC-26
- Source:
- R&D Connections, n26, 2018
- Document Type:
- Periodical
- Subject/Key Words:
- Rater Cognition, Computational Cognitive Models, Test Taker Cognition, Test Development, Simulations, Test Validity, Adaptive Control of Thought-Rational (ACT-R), Markov Decision Processes, Learning Curves, Forgetting Curves, Human Cognition, Computer-Based Testing (CBT), Item Format, Item Design, Solution Strategies, Human Raters, Assessment Development (AD)
Abstract
This essay describes computational cognitive models and gives examples of how they are used at Educational Testing Service (ETS). Computational models are grounded in theories of cognition. They seek to understand the mental processes of humans interacting with assessments. Researchers and assessment developers at ETS are, for example, interested in knowing if the way in which an assessment task is presented can influence how test takers react and think, and ultimately, how they answer. Such models play an important role in many aspects of assessment, including assessment development (e.g., creating test items), evaluation of validity evidence, and improving the efficiency of the raters who score test responses. The article highlights how computational cognitive modeling informs research on student and rater responses. This is the 26th issue in the R&D Connections series.