Authors have claimed that Internet-based instruction promotes greater learning efficiency than non-computer methods. Objectives Determine, through a systematic synthesis of evidence in health professions education, how Internet-based instruction compares with non-computer instruction in time spent learning, and what features of Internet-based instruction are associated with improved learning efficiency. Data sources We searched databases including MEDLINE, CINAHL, EMBASE, and ERIC from 1990 through November 2008. Study selection and data abstraction We included all studies quantifying learning time for Internet-based instruction for health professionals, compared with other instruction. Reviewers worked independently, in duplicate, to abstract information on interventions, outcomes, and study design. Results We identified 20 eligible studies. Random effects meta-analysis of 8 studies comparing Internet-based with non-Internet instruction (positive numbers indicating Internet longer) revealed pooled effect size (ES) for time −0.10 (p = 0.63). Among comparisons of two Internet-based interventions, providing feedback adds time (ES 0.67, p = 0.003, two studies), and greater interactivity generally takes longer (ES 0.25, p = 0.089, five studies). One study demonstrated that adapting to learner prior knowledge saves time without significantly affecting knowledge scores. Other studies revealed that audio narration, video clips, interactive models, and animations increase learning time but also facilitate higher knowledge and/or satisfaction. Across all studies, time correlated positively with knowledge outcomes (r = 0.53, p = 0.021). Conclusions On average, Internet-based instruction and non-computer instruction require similar time. Instructional strategies to enhance feedback and interactivity typically prolong learning time, but in many cases also enhance learning outcomes. Isolated examples suggest potential for improving efficiency in Internet-based instruction.

Denne systematiske oversikten er utarbeidet på en god måte og forfatterne har vært forholdsvis tydelige på metoden de har brukt for å utarbeide oversikten. Inklusjons- og eksklusjonskriterier er beskrevet på en god måte, men de kunne med fordel dokumentert de ulike stegene i søkeprosessen på en bedre måte. Forfatterne har gjort en grundig vurdering av den metodologiske kvaliteten til de inkluderte studiene. Det er gjennomført meta-analyse og resultatene oppsummeres på en god måte i både teksten og med informative tabeller og figurer.