The purpose of this meta-analysis was to examine the current effectiveness of physical activity (PA) interventions to change affective judgements (AJ) and subsequent behaviour and explore potential moderators. Eligible studies were published in a peer-reviewed English journal and included an experimental design in the PA domain with a measure of AJ as the dependent variable, among adults (>17 years). Literature searches concluded in July 2017 using 11 common databases, with additional hand searching conducted in February 2018. The search yielded 32 independent studies. Random-effects meta-analysis showed positive changes in AJ favouring intervention over control groups, g = 0.43 (95% CI = 0.26–0.60). These changes predicted (β = 0.64) positive changes in PA, g = 0.38 (95% CI = 0.16–0.60), among a sub-sample (k = 14) of studies that also provided behavioural data. Moderator analyses showed the effects were inflated by potential publication bias, participant gender, baseline PA and focus of the intervention. AJ may show change from intervention but larger sample studies are required to obtain a more reliable effect size estimate. Further, few studies have employed behaviour change techniques that would align with the theoretical reasons for changes in AJ, so our evidence for practical intervention content is limited.

这项荟萃分析的目的是检验体育活动（PA）干预措施当前用于改变情感判断（AJ）和随后行为的有效性，并探讨潜在的主持人。合格的研究发表在同行评审的英语期刊上，其中包括成年人（> 17岁）在PA域进行的实验设计，以AJ作为因变量。文献检索于2017年7月使用11个通用数据库完成，并于2018年2月进行了额外的手工检索。该检索产生了32项独立研究。随机效应荟萃分析显示，AJ的积极变化优于对照组，g  = 0.43（95％CI = 0.26-0.60）。这些变化的预测（β  = 0.64）中PA积极变化，克= 0.38（95％CI = 0.16-0.60），其中 还提供了行为数据的研究子样本（k = 14）。主持人分析显示，这种影响因潜在的出版偏见，参与者性别，基线PA和干预重点而膨胀。AJ可能显示干预后的变化，但需要进行更大的样本研究才能获得更可靠的效应量估计值。此外，很少有研究采用与AJ变化的理论原因相符的行为改变技术，因此我们对实际干预内容的证据是有限的。