# Spanking and Children’s Approaches to Learning: Estimates from a Longitudinal Matched-Sample Design

## TL;DR

Spanking at age 5.5 is linked to less positive learning approaches in children by ages 6.5 and 7.5, even with low-frequency spanking.

## Contribution

This study provides causal estimates of spanking's impact on learning approaches using a matched-sample design.

## Key findings

- Spanking at age 5.5 is associated with less positive approaches to learning at ages 6.5 and 7.5.
- Results remain significant even when limited to low-frequency spanking (once per week).

## Abstract

One form of corporal punishment commonly used in the United States is spanking. Spanking is a well-known risk factor for adverse child development, although its influence on children’s approaches to learning (ATL) has been largely overlooked. Existing research is particularly limited in inadequately considering multiple confounds and selection biases in children’s exposure to spanking. This study examined the links between spanking and children’s ATL, using a matched-group design to strengthen causal estimates among children aged 5 to 7.5 (N = ~12,800) from the Early Childhood Longitudinal Study, Kindergarten Class of 2010–2011. Entropy-balanced matching mitigated selection and confounding biases, controlling for a wide array of covariates. The sensitivity of spanking’s effects on ATL was also tested by limiting the sample to low-frequency spanking (once in the past week) to address concerns that primarily higher-frequency spanking predicts ATL. Findings indicated that spanking at age 5.5 was associated with less positive ATL at ages 6.5 and 7.5. These results remained significant when limited to low-frequency spanking. This study’s findings suggest that spanking may hinder children’s development of positive approaches to learning, which holds significant implications for lifelong well-being. This study contributes to the growing literature on the potential negative effects of physical punishment on child development across multiple domains.