Investing in early childhood development (ECD) is critical for individual and societal development. Variable-centered research on ECD has shown that family wealth, maternal education, and parenting practices predict childhood outcomes overall. However, little is known about differences in the ECD patterns and their predictors. This study examined the latent classes of ECD using data from three waves of the Multiple Indicators Cluster Surveys (MICS) conducted in 29 low- and middle-income countries (LMICs) between 2010 and 2020 (MICS 4, 5, and 6) and identified their predictors at different ecological levels. The total sample size for analyses was 226,374 ( = 70,082, = 91,652, = 64,640; = 47.23(months), = 6.87). Three classes, , and , were consistently identified across MICS 4–6 using latent class analysis. Three variables, all at the microsystem level, predicted class membership with acceptable effect sizes in one or more waves of the MICS data: preschool attendance, number of books at home, and maternal education. The study has implications for future research and the development of policies aimed at monitoring and supporting ECD in LMICs.

中文翻译：

---

低收入和中等收入国家儿童早期发展的潜在类别及其预测因素：2010 - 2020 年多指标类集调查的结果


投资儿童早期发展 (ECD) 对于个人和社会发展至关重要。以变​​量为中心的儿童早期发展研究表明，家庭财富、母亲教育和养育方式可以总体预测儿童结局。然而，人们对 ECD 模式及其预测因素的差异知之甚少。本研究使用 2010 年至 2020 年间在 29 个低收入和中等收入国家 (LMIC) 进行的三轮多指标类集调查 (MICS) 数据（MICS 4、5 和 6）检查了 ECD 的潜在类别，并确定了他们在不同生态水平上的预测因子。分析总样本量为 226,374（= 70,082，= 91,652，= 64,640；= 47.23（月），= 6.87）。使用潜在类别分析在 MICS 4-6 中一致确定了三个类别 、 和 。三个变量均处于微系统水平，在一轮或多轮 MICS 数据中以可接受的效应大小预测了班级成员资格：学前班出勤率、家庭书籍数量和母亲教育程度。该研究对未来研究和旨在监测和支持中低收入国家儿童早期发展的政策制定具有影响。