Scaling Early Childhood Development – what to read this month (September 2024)

In this monthly blog – also available as a newsletter – we highlight recent advances in research, materials, tools and practices related to how to design, implement, monitor and evaluate scalable early childhood development (ECD) programmes in low- and middle-income countries (LMICs) worldwide. The inspiration for this new series came from Ugo Gentilini’s excellent Weekly Social Protection Links.   

Scaling Early Childhood Development – What to Read this Month is curated by Bet Caeyers (Lead Editor, Chr. Michelsen Institute), Gemma Knights (Editor, Oxford Policy Management) and Meghan Taylor (Editor, Oxford Policy Management). 

First up: What is the impact and potential scalability of home-installed growth charts and nutrient supplements on child growth and development, and is there any risk in bundling these two? Fink et al. conducted a randomised control trial (RCT) across three districts in Zambia to find out. Direct distribution of nutrient supplements alone has positive impacts on children’s nutritional status and development across all outcomes (stunting, anaemia and development). Home installation of a wall chart – containing an easy-to-use growth monitoring tool along with key nutritional and stimulation messages – seems to have little effect in reducing linear growth faltering but does seem to increase children’s caregiver-child interactions and development. Interestingly, the combination of both interventions underperforms across most outcomes in comparison to the standalone interventions. The evidence suggests that participation in two interventions lowered parental efforts to comply with either intervention. This result stresses the importance of careful integration of interventions if they are to be combined. 

Sticking to bundling interventions but now moving to water, sanitation and hygiene (WASH). Let’s look at the effects of an intervention combining behaviour change messaging with hardware provision to improve safe child faeces management in Odisha, India. Sclar et al.’s cluster RCT evaluates the Movement and Action Network for Transformation of Rural Areas (MANTRA) programme, targeting caregivers of children under 5 who had access to WASH facilities at baseline. The community-based behaviour change component provided caregivers with information on safe child faeces disposal and child latrine use. Hardware provision focused on items that aid caregivers with safe WASH practices (e.g. a bucket with a lid). Results indicate that the intervention increased safe disposal by caregivers with the lowest safe disposal rates at baseline and also increased child latrine use among the child age group that is developmentally ready to toilet train (children under 3). While future research is required to assess the sustainability of these effects, these are encouraging findings for a potentially scalable safe faeces management intervention. 

Can picture books improve learning in pre-primary education in low-resource settings? Maruyama and Igei explore the impact of introducing seven Japanese-originated ‘Kamishibai’ in pre-schools in Madagascar. Kamishibai is based on a simple picture book with scripts, purposefully designed for use by unqualified teachers in 10-minute sessions for groups of children (approximately 20). Despite the small sample size and the relatively short intervention period of two and a half months, the results are positive. For example, the proportion of children who can count numbers from 10 to 20 increased by 24.8 percentage points. Check out pages 5–7 for a full breakdown of all 29 outcomes. Although larger sampling over a longer period should take place to verify results, given the low intensity of the intervention, there is significant potential for scaling, especially in low-resourced education settings.

What lessons can we learn from behavioural science to make parenting interventions more effective, scalable and sustainable? Building on results from a meta-analytic review (which we discussed in our March newsletter) in a recent Nature Correspondence article, Ahun and Bacon argue that, first, researchers need to engage in a more intentional and transparent intervention development process. This should be done by determining the correct approach, analysing the target behaviour and conducting formative research before diving straight into testing the intervention. Second, policymakers, NGOs and researchers should test the feasibility and efficacy of an intervention before attempting to scale. Finally, more systematic reporting about the content and how the intervention is implemented is needed.

Can centre-based parenting – arguably more sustainable and accessible than home-based parenting – be successful in reducing caregiver anxiety and stress levels? This new paper by Yue et al. uses an RCT to answer this question, focusing on caregivers of lower socioeconomic status in rural China. The paper finds that parenting centres are, indeed, an effective avenue to reduce caregivers’ anxiety and stress levels, with effects persisting two years post-intervention. Primary impact mediators seem to be increased parenting time investment, alterations in parenting styles and changes in the level of trust in others. This research contributes valuable insights into addressing mental health challenges in rural settings. 

In keeping with the theme of mental health, in a rare cost-effectiveness analysis for perinatal depression care in sub-Saharan Africa, a new paper by McBain et al. calculates the economic cost of maternal depression treatment for women living with HIV. The cost-effectiveness analysis of the M-DEPTH intervention evaluated its value in reducing perinatal depression by comparing its costs to health outcomes, measured in disability-adjusted life years averted. Using a Markov chain Monte Carlo  approach with 100,000 simulations in TreeAge Pro, the analysis estimated cost-effectiveness across a range of willingness-to-pay thresholds, focusing on US$964 (Uganda’s median GDP per capita) as the key threshold, consistent with WHO-CHOICE guidelines. The result was an estimated incremental cost-effectiveness ratio of M-DEPTH versus care as usual of US$397, suggesting that M-DEPTH was cost-effective. The authors maintain that based on the analysis, the cost of the M-DEPTH interventions would be less than 0.5% of the Ugandan health budget, essentially paying for itself by reducing the burden of perinatal depression.

Pivoting to the subject of intimate partner violence (IPV), this new systematic review by Mercier et al. looks at IPV interventions in the perinatal period – distinguishing between four main intervention types: education, cash transfers, counselling and home visits. The review reveals that cash transfers are most prevalent in LMICs, whereas home visits are more common in higher-income countries. IPV interventions most commonly seek to reduce revictimisation and improve maternal mental health. Only very few IPV studies consider the impacts of IPV interventions on neonatal and child health outcomes. The review calls for more research to close this evaluation gap. 

On the topic of measurement, in this paper, Wolf et al. introduce a new observational tool designed to assess support for engaged learning in early childhood classrooms. Unlike existing tools, the Playful Learning Across the Years (PLAY) early childhood education observational tool focuses on interactions between caregivers and students that promote children’s engagement in the learning process. Using video data from Ghanaian pre-primary classrooms the authors assess concurrent validity by considering how the tool relates to more widely used measures of quality (e.g. the Teacher Instructional Practices and Processes System – TIPPS). Also helpful is an analysis of the measure’s sensitivity to intervention impacts – in this case, confirmed through an RCT of a teacher training development programme. This new measure is expected to be very valuable in improving our understanding of how engagement can be supported in early childhood education classrooms. It remains to be seen, however, whether the tool – which in the study was used by specially trained enumerators with experience in collecting data – can be useful at scale. 

Finally, for those interested in whether and how policy influences practice, the inspirational What Works Hub for Global Education has published a new paper by Angrist and Dercon, looking at the education policy-practice gap across 50 countries during COVID-19. For policy to matter it must be implemented. During COVID-19, almost all countries promptly implemented distance learning policies, but distance learning did not take place at the same rate. The largest policy-practice gap was observed in sub-Saharan Africa where 90% of countries rolled out a distance learning policy and fewer than 30% of students received any distance education. Beyond COVID-19, using data from Systems Approach for Better Education Results (SABER), the authors find large gaps between ECD policies that are introduced and their implementation, with a difference of roughly 50 percentage points on average. Most of the gap can be explained by ineffective service delivery.