Computational Pedagogical Strategies to Help Learners Retain STEM Content Knowledge
Abstract: Computational pedagogical experiences were designed and implemented to help students retrieve content they learned in class, retain it, and apply it in different contexts to solve novel problems. Supported by professional development for teachers, these experiences ranged in complexity from simple electronic flashcards for basic retrieval strategies to low-stakes quizzes for spaced-out (exposure and retrieval effort are spaced out) and interleaved (two or more spaced-out topics are interleaved) practices. A mixed-methods approach is employed, using instruments with relevant psychometric properties for a large number of teachers (N=180) and their students. Action Research were conducted by a cadre of teachers who randomly selected control and target student groups within the same school, grade and course environment. Teachers self-selected an area of content within their respective disciplines or curriculum and created two different retrieval practices – a blocked practice that examines student knowledge and skills for applying a certain method to the solution of various questions on only one topic or type, and an interleaved practice that involves questions on two or more topics that need different methods to solve. Results show that students who learned math & science topics through interleaved practices consistently scored 5-30% better than those who learned it in the more traditional blocked practice. In many cases, the differences were statistically significant (p <0.05).