This systematic review and meta-analysis examined the effects of analogy instruction (ANA) on motor performance and knowledge declared (KD) compared with explicit learning (EXP) and control conditions. Five databases were included. The study analyzed 16 randomized controlled trials. Subsequent analysis was performed for moderators variables as age, skill, retention, stress situations number of rules, specificity and number of trials. The ANA instruction demonstrated greater efficacy than the control (ES = 0.32, p = 0.03) or EXP condition (ES = 0.29, p = 0.02) in motor tasks performance in general terms. ANA instructions also showed superiority in motor performance when compared to control conditions in retention (ES = 5.72, p = 0.004), and a trend towards significance was found under stress (ES = 1.18, p = 0.05). ANA also showed superiority in motor performance when compared to EXP instruction (ES = 0.29, p = 0.02). ANA demonstrated greater effects than EXP in retention (ES = 7.25, p = 0.01), but not under stress (ES = 0.62, p = 0.18). Sub-analyses demonstrated that children (all p < 0.01) and novices (all p < 0.01) are more likely to benefit from ANA instruction when compared to control or EXP. A subgroup analysis based on quantity of information comparing ANA versus EXP condition shows that ANA is more effective for enhancing motor performance than EXP when the number of rules are similar. Sub-analyses comparing ANA versus CNT shows that as the number of repetitions increases and the task becomes less specific, ANA instruction significantly improves performance. A comparison between ANA and EXP indicates no significant differences in performance regarding the number of repetitions and task specificity. A secondary analysis examined KD of different instructions. KD was greater in EXP instructions (ES = −1.48, p < 0.001) when compared to ANA. Findings suggest that analogy instruction may improve motor performance, especially in novice and child populations. However, caution is needed due to concerns when comparing with other instructional types and environments, as well as due to high heterogeneity in most of the comparisons and high risk of bias in included studies.
Read full abstract