As growth mindset interventions increase in scope and popularity, scientists and policymakers are asking: Are these interventions effective? To answer this question properly, the field needs to understand the meaningful heterogeneity in effects. In the present systematic review and meta-analysis, we focused on two key moderators with adequate data to test: Subsamples expected to benefit most and implementation fidelity. We also specified a process model that can be generative for theory. We included articles published between 2002 (first mindset intervention) through the end of 2020 that reported an effect for a growth mindset intervention, used a randomized design, and featured at least one of the qualifying outcomes. Our search yielded 53 independent samples testing distinct interventions. We reported cumulative effect sizes for multiple outcomes (i.e., mindsets, motivation, behavior, end results), with a focus on three primary end results (i.e., improved academic achievement, mental health, or social functioning). Multilevel metaregression analyses with targeted subsamples and high fidelity for academic achievement yielded, d = 0.14, 95% CI [.06, .22]; for mental health, d = 0.32, 95% CI [.10, .54]. Results highlighted the extensive variation in effects to be expected from future interventions. Namely, 95% prediction intervals for focal effects ranged from -0.08 to 0.35 for academic achievement and from 0.07 to 0.57 for mental health. The literature is too nascent for moderators for social functioning, but average effects are d = 0.36, 95% CI [.03, .68], 95% PI [-.50, 1.22]. We conclude with a discussion of heterogeneity and the limitations of meta-analyses. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

中文翻译：

---

对成长心态干预措施的系统回顾和荟萃分析：此类干预措施对谁、如何以及为何有效？

随着成长心态干预措施的范围和受欢迎程度不断扩大，科学家和政策制定者开始问：这些干预措施有效吗？为了正确回答这个问题，该领域需要理解影响中有意义的异质性。在目前的系统回顾和荟萃分析中，我们重点关注两个有足够数据进行测试的关键调节因素：预计受益最大的子样本和实施保真度。我们还指定了一个可以生成理论的过程模型。我们纳入了 2002 年（首次心态干预）至 2020 年底之间发表的文章，这些文章报告了成长心态干预的效果，使用了随机设计，并至少具有一个合格的结果。我们的搜索产生了 53 个独立样本来测试不同的干预措施。我们报告了多种结果（即心态、动机、行为、最终结果）的累积效应大小，重点关注三个主要最终结果（即学业成绩、心理健康或社会功能的提高）。使用目标子样本和高保真度的学术成绩进行多级元回归分析，d = 0.14，95% CI [.06, .22]；对于心理健康，d = 0.32，95% CI [.10，.54]。结果凸显了未来干预措施预期效果的巨大差异。也就是说，学术成就焦点效应的 95% 预测区间为 -0.08 至 0.35，心理健康焦点效应的预测区间为 0.07 至 0.57。对于社会功能调节因素来说，文献还太新，但平均效应为 d = 0.36，95% CI [.03, .68]，95% PI [-.50, 1.22]。最后我们讨论了异质性和荟萃分析的局限性。（PsycInfo 数据库记录 (c) 2023 APA，保留所有权利）。