Stability of Self-Reported Child Abuse: Implications for Global Development and Policy Evaluation

Introduction: Data Quality as the Cornerstone of the Sustainable Development Goals

In the global development agenda, eliminating all forms of violence against children is an explicit commitment of the Sustainable Development Goals (SDGs), particularly Target 16.2. However, monitoring progress toward this goal is highly dependent on effective and reliable data. The measurement of child abuse typically relies on two approaches: prospective reports (e.g., official records) and retrospective self-reports. The latter, due to its low cost and ease of implementation, is widely used in large-scale surveys and clinical studies. However, retrospective self-reports have long been considered susceptible to memory bias and changes in subjective evaluation, raising doubts about their stability. A recent systematic review and meta-analysis published in Nature Mental Health (Coleman et al., 2026) provides quantitative evidence on this debate, with far-reaching implications for global development research, policy formulation, and ESG investing.

Core Findings of the Study: Overall Stable, but with Key Heterogeneity

This meta-analysis integrated 49 studies involving 38,332 participants, with an average follow-up of 2.4 years, and found that retrospective self-reports were generally highly stable (correlation coefficient r = 0.79). This result challenges the traditional assumption that "self-reports are inherently unstable." However, stability was not uniform: it was lower in population-representative samples than in clinical or convenience samples; lower for neglect reports than for abuse reports; lower in children than in adults; and within the child group, stability declined significantly with longer follow-up periods. From a neurobiological perspective, the immaturity of the hippocampus and lower memory consolidation efficiency during childhood may be the mechanisms underlying this lower stability.

Implications for Global Development: How Measurement Error Affects Policy Effectiveness

For developing countries or resource-limited regions, retrospective self-reports are often the only feasible data source. This study indicates that relying on children's own reports, especially after longer intervals, may overestimate or underestimate the incidence of abuse, leading to misallocation of policy resources. For example, a survey on violence against children in a certain African country using single-time-point child self-report data might miss exposure changes over time, thereby affecting the baseline setting for SDG targets. The meta-analysis points out that stability decays with follow-up time in child populations, meaning the earlier the intervention, the more likely memory is to be shaped; therefore, early detection and intervention are crucial.

In addition, stability for neglect-type abuse is lower than for physical abuse, suggesting that existing tools may be more fragile in capturing emotional and care deprivation. This calls for improvements in survey questionnaires designed by global development agencies such as UNICEF: modular, scenario-based questions targeting neglect need to be added, and a mixed mode combining interviews and questionnaires should be considered to reduce measurement error.

Implications for ESG and Investment Assessment in the Social DimensionIn ESG investing, the social pillar (S) is increasingly focusing on child protection in supply chains. When conducting human rights due diligence, companies often rely on self-reported data from workers or communities to assess risks. This study reminds investors that if the assessment targets adolescents or uses longer recall periods (e.g., the past five years), the stability of self-reported data may decrease, leading to bias in risk evaluation. It recommends that assessors combine official records or third-party observations, or adopt questionnaires with shorter recall periods, to improve data reliability. Meanwhile, stability differences across gender and age groups also highlight the need for segmented assessment.

International Cooperation and Standardization of Research Design

In this study, sample characteristics and measurement tools are key factors explaining heterogeneity. The global research community should promote the standardization of child maltreatment measurement, especially in cross-cultural comparisons. For example, differences in memory reporting patterns may exist between developed and developing countries, but current evidence is still insufficient. International organizations such as WHO and UNICEF could update their "Guidelines for Measuring Violence against Children" based on these findings, recommending the priority use of adult retrospective recall over child self-report in representative surveys, or appropriately increasing the frequency of follow-ups.

Long-term Trends and Judgment of Governance Capacity

With the global aging population, the long-term impact of childhood maltreatment on adult mental health is receiving increasing attention. This study shows that the recall stability of subjective experiences is higher in adults, suggesting that epidemiological data based on adult retrospective recall (e.g., ACE studies) have good reliability. However, low stability in childhood implies that policy interventions relying solely on adult recall may miss critical windows of change during childhood. Governance systems need to shift from "post-hoc recall" to "real-time monitoring," such as integrating child protection screening into school health systems and using short-interval repeated measurements.

Conclusion: Data Robustness Determines the Credibility of Development Outcomes

This meta-analysis provides a key quantitative baseline for the stability of self-reported child maltreatment, revealing its overall reliability while pointing out vulnerabilities in specific contexts. For global development practitioners, this serves both as affirmation and a warning: affirmation of the practicality of retrospective data under resource-limited conditions; a warning that data interpretation and tool selection must be adjusted based on sample type, abuse type, and age. In the pursuit of SDG goals, only rigorous measurement can produce reliable evidence, thereby guiding effective policy and investment. Future research should focus on long-term stability (beyond 12 years) and cultural differences to further improve the global data infrastructure for child protection.*Reference: Coleman, O., Al-Jaber, M., Aponsu, G., et al. (2026). Stability of childhood maltreatment self-reports: a systematic review and meta-analysis. Nature Mental Health. https://www.nature.com/articles/s44220-026-00677-7*