Model Human Learners: Computational Models to Guide Instructional Design
Christopher J. MacLellan
TL;DR
This work addresses the challenge of instructional design in sprawling design spaces by introducing the Model Human Learner, a unified computational framework that simulates human learning to predict the outcomes of instructional interventions. It validates the approach through two human A/B experiments, showing that the Apprentice Learner Architecture with the Trestle model can predict main effects and generate learning curves without prior human data. The model not only aligns with human performance across different tasks but also provides theoretical insights into why interventions work, including a challenging finding about constrained versus unconstrained problems. Overall, this approach offers a scalable, theory-driven tool to pre-screen instructional designs, reducing reliance on costly human experiments and guiding design across diverse tasks and interventions.
Abstract
Instructional designers face an overwhelming array of design choices, making it challenging to identify the most effective interventions. To address this issue, I propose the concept of a Model Human Learner, a unified computational model of learning that can aid designers in evaluating candidate interventions. This paper presents the first successful demonstration of this concept, showing that a computational model can accurately predict the outcomes of two human A/B experiments -- one testing a problem sequencing intervention and the other testing an item design intervention. It also demonstrates that such a model can generate learning curves without requiring human data and provide theoretical insights into why an instructional intervention is effective. These findings lay the groundwork for future Model Human Learners that integrate cognitive and learning theories to support instructional design across diverse tasks and interventions.