Effects of spaced digital retrieval practice on delayed vocabulary retention in EFL learners: A schedule-controlled classroom experiment

Abstract

Digital vocabulary tools are widely justified by appeal to the science of memory, yet most evidence compares a digital spaced condition against a non-digital massed one, confounding practice schedule with delivery medium, and typically reports the spacing effect as though it applied uniformly across words. This study isolated the scheduling manipulation by delivering spaced and massed retrieval practice through the same platform, with the same items, the same feedback, and the same total practice, to intermediate EFL learners (N = 60; first language Urdu). Item-level accuracy on 40 target words was assessed at pretest, immediately after practice, and after fourteen days, and analysed using mixed-effects logistic regression with crossed random effects for participants and items. Spaced practice increased the odds of correct retention overall, but the advantage was concentrated at delay: the schedules did not differ immediately (OR = 1.13) yet diverged after two weeks (OR = 2.10), producing a reliable schedule-by-time interaction. The advantage was uniform across word properties—frequency, concreteness, and cognate status each predicted retention, but none moderated the schedule effect. Under equated delivery, distributed practice yields more durable vocabulary knowledge irrespective of word type, supporting scheduling, but not item-selective scheduling, as the locus of benefit.