| Abstract: | Young children often learn less from video than face-to-face presentations. Meta-regression models were used to examine the average size of this difference (video deficit) and investigate moderators. An average deficit of about half of a standard deviation was reported across 122 independent effect sizes from 59 reports, involving children ages 0–6 years. Moderator analyses suggested (a) the deficit decreased with age, (b) object retrieval studies showed larger deficits than other domains, and (c) there was no difference between studies using live versus prerecorded video. Results are consistent with a multiple-mechanism explanation for the deficit. However, the analyses highlighted potential quality and publication bias issues that may have resulted in overestimation of the effect and should be addressed by future researchers. |
