Making long-term memories in minutes: a spaced learning pattern from memory research in education

Abstract

Memory systems select from environmental stimuli those to encode permanently. Repeated stimuli separated by timed spaces without stimuli can initiate Long-Term Potentiation (LTP) and long-term memory (LTM) encoding. These processes occur in time scales of minutes, and have been demonstrated in many species. This study reports on using a specific timed pattern of three repeated stimuli separated by 10 min spaces drawn from both behavioral and laboratory studies of LTP and LTM encoding. A technique was developed based on this pattern to test whether encoding complex information into LTM in students was possible using the pattern within a very short time scale. In an educational context, stimuli were periods of highly compressed instruction, and spaces were created through 10 min distractor activities. Spaced Learning in this form was used as the only means of instruction for a national curriculum Biology course, and led to very rapid LTM encoding as measured by the high-stakes test for the course. Remarkably, learning at a greatly increased speed and in a pattern that included deliberate distraction produced significantly higher scores than random answers (p < 0.00001) and scores were not significantly different for experimental groups (one hour spaced learning) and control groups (four months teaching). Thus learning per hour of instruction, as measured by the test, was significantly higher for the spaced learning groups (p < 0.00001). In a third condition, spaced learning was used to replace the end of course review for one of two examinations. Results showed significantly higher outcomes for the course using spaced learning (p < 0.0005). The implications of these findings and further areas for research are briefly considered.

Keywords: education methods; encoding; long-term memory; long-term potentiation; spaced learning; spaced retrieval.

Figure 1

High-stakes test scores for teaching (four months) and spaced learning (one hour). There was no significant difference between the test scores for teaching (N = 127) and spaced learning (N = 46) groups in Condition 1. In Condition 2 there was also no significant difference between teaching test scores (N = 131) and spaced learning test scores (N = 21). All data are means ± SEM.

Figure 2

Test score percentage increase per hour of instruction. High-stakes test scores percentage increase per hour of instruction was significantly higher for spaced learning (N = 46) than for teaching (N = 127) in Condition 1 ***P < .00001. In Condition 2 the test score increase for spaced learning (N = 21) was also significantly higher than for teaching (N = 131) ***P < .00001. All data are means ± SEM.