Date of Award


Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Applied Cognitive Science and Human Factors (PhD)

Administrative Home Department

Department of Cognitive and Learning Sciences

Advisor 1

Shane T. Mueller

Committee Member 1

Elizabeth S. Veinott

Committee Member 3

Laura E. Brown

Committee Member 4

Leo C. Ureel II


Today’s intelligent software systems, such as Artificial Intelligence/Machine Learning systems, are sophisticated, complicated, sometimes complex systems. In order to effectively interact with these systems, novice users need to have a certain level of understanding. An awareness of a system’s underlying principles, rationale, logic, and goals can enhance the synergistic human-machine interaction. It also benefits the user to know when they can trust the systems’ output, and to discern boundary conditions that might change the output. The purpose of this research is to empirically test the viability of a Cognitive Tutorial approach, called Explicit Rule Learning. Several approaches have been used to train humans in intelligent software systems; one of them is exemplar-based training. Although there has been some success, depending on the structure of the system, there are limitations to exemplars, which oftentimes are post hoc and case-based. Explicit Rule Learning is a global and rule-based training method that incorporates exemplars, but goes beyond specific cases. It provides learners with rich, robust mental models and the ability to transfer the learned skills to novel, previously unencountered situations. Learners are given verbalizable, probabilistic if...then statements, supplemented with exemplars. This is followed up with a series of practice problems, to which learners respond and receive immediate feedback on their correctness. The expectation is that this method will result in a refined representation of the system’s underlying principles, and a richer and more robust mental model that will enable the learner to simulate future states. Preliminary research helped to evaluate and refine Explicit Rule Learning. The final study in this research applied Explicit Rule Learning to a more real-world system, autonomous driving. The mixed-method within-subject study used a more naturalistic environment. Participants were given training material using the Explicit Rule Learning method and were subsequently tested on their ability to predict the autonomous vehicle’s actions. The results indicate that the participants trained with the Explicit Rule Learning method were more proficient at predicting the autonomous vehicle’s actions. These results, together with the results of preceding studies indicate that Explicit Rule Learning is an effective method to accelerate the proficiency of learners of intelligent software systems. Explicit Rule Learning is a low-cost training intervention that can be adapted to many intelligent software systems, including the many types of AI/ML systems in today’s world.