The objective of the research and development project, Structured Retrieval Practice to Enhance Conceptual Learning, is to develop a computer-based Structured Retrieval Practice program that guides students to practice retrieval while studying and enhances long-term retention of foundational conceptual knowledge. The research addresses fundamental questions about human learning and retrieval processes by testing the hypothesis that elaborative retrieval cues increase the effectiveness of retrieval practice. The first aim is to develop and enhance several aspects of the learning program in controlled experiments. The working hypothesis is that guiding students to practice active reconstruction of their knowledge will produce potent and robust learning and long-term retention. The goal is to identify the best ways to structure and scaffold students' retrieval practice to help them develop enriched mental models about science concepts. The second aim of this project is to establish the efficacy and effectiveness of Structured Retrieval Practice in two large undergraduate introductory psychology courses. The project addresses two gaps in cognitive science. First, the project is designed to contribute to the current research on retrieval and memory. Secondly, the project is designed to translate findings from the basic research conducted in a laboratory setting into application in the real world setting of the classroom. The set of studies is designed to inform the undergraduate STEM education community on how students learn and to apply the findings to the development of instructional strategies designed to enhance students' learning of foundational conceptual knowledge in STEM disciplines.
The long-term goal of our laboratory's research is to translate basic principles about how people learn, culled from cognitive science, into educational practices that promote meaningful learning and long-term retention of conceptual knowledge. The objective of this project was to develop a computer-based Structured Retrieval Practice program that guides students to practice retrieval while studying and enhances long-term retention of foundational conceptual knowledge. We carried out a series of controlled experiments to develop and enhance several aspects of the learning program, with the goal of identifying the best ways to structure and scaffold students' retrieval practice. An additional purpose of the project was to establish the effectiveness of Structured Retrieval Practice in large, introductory-level undergraduate STEM courses. Our general prediction was that guiding students to practice active retrieval of their knowledge would produce potent and robust learning and long-term retention. The Structured Retrieval Practice program uses an automated, computer-based method of scoring students' responses that we developed, which we call QuickScore. The scoring algorithm searches students' responses for specific keywords. Responses are considered "correct" if the necessary keywords are present. The purpose of the scoring algorithm is twofold. First, the automated scoring method allows us to control and adjust the presentation of particular concepts, depending on a learner's level of performance. Second, the scoring algorithm allows us to give students meaningful feedback about what they did or did not recall during retrieval practice. We conducted several experiments on the Structured Retrieval Practice program during the project period. For example, in one experiment, the scoring algorithm was used to determine which items would be presented for repeated retrieval practice or repeated studying. Students learned anatomy and physiology key term definitions. Each key term consisted of a muscle group (e.g., the deltoid muscle) and an important feature of that muscle group (e.g., its location, its function, or its innervation). Students learned by studying and recalling the terms across alternating trials. During study trials, students studied the key terms and definitions. During retrieval practice trials, students were given prompts (e.g., "How is the deltoid muscle innervated?") and asked to recall the target response by typing into an input field. Student answers were scored in real time using QuickScore, the automated scoring method we developed for this project. Students were assigned to one of two experimental conditions. One group of students repeatedly studied the key term definitions, while a second group (the retrieval practice condition) was given the opportunity to practice repeatedly retrieving the definitions. Thus, the repeated retrieval group recalled the definitions not just once but multiple times while learning. All of the students then took a final delayed test to assess what they retained from the learning experience. The results of this experiment showed a strong retrieval practice effect: Students in the retrieval practice condition showed about a 25% advantage over the students in the repeated study condition. This experiment is important because it demonstrates the effectiveness of repeated retrieval for long-term learning and it shows that our scoring algorithm is an effective and feasible way to implement retrieval practice. During the project period, we made significant progress on a broad range of projects aimed at identifying the "best practices" for implementing retrieval practice to produce optimal levels of meaningful, long-term learning. We are continuing our efforts to identify effective retrieval practice techniques, to develop and refine the QuickScore scoring algorithm, and to create a web-based interface that will allow students and instructors to use the Structured Retrieval Practice program in their courses. In sum, we have made substantial progress during the completed project period, and we look forward to continuing our extensive development of the Structured Retrieval Practice approach in the future.
