Evaluating the Impact of a WhatsApp-Based Early Stimulation Program on Child Development and Parental KAP: A Randomized Controlled Trial in Rural and Urban India

Last registered on September 12, 2025

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0016756

Initial registration date

September 10, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

September 12, 2025, 10:37 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

India

Region

Fatehabad and Panipat, Haryana

Primary Investigator

Name

Bilal Afroz

Affiliation

Rocket Learning

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Namya Mahajan

PI Affiliation

Rocket Learning

Contact Investigator

PI Name

Nivedita Roy

PI Affiliation

Rocket Learning

Contact Investigator

PI Name

Dr. Dana McCoy

PI Affiliation

Harvard University

PI Name

Kenji Kitamura

PI Affiliation

Harvard University

Additional Trial Information

Status

On going

Start date

2025-04-01

End date

2026-06-30

Keywords

Behavior, Education, Gender, Other

Additional Keywords

child development

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

This study evaluates the impact of Rocket Learning’s birth-to-three early stimulation program, which aims to support parents and caregivers in fostering the healthy development of children during their first three years of life. Early childhood is a critical period for brain development, and responsive caregiving practices can significantly influence children’s cognitive, language, motor, and socio-emotional growth.

The program delivers weekly digital content via WhatsApp to parents of children aged 0–3 years. Messages focus on simple, age-appropriate activities that parents can do with their children to promote stimulation, bonding, and early learning. The goal is to strengthen parental knowledge, attitudes, and practices related to child development, while improving children’s developmental outcomes.

To measure impact, we are conducting a randomized controlled trial (RCT) across two districts in Haryana, India (Fatehabad and Panipat). Anganwadi Centres (AWCs) were randomly assigned to treatment or control groups. In treatment AWCs, all eligible households receive the digital parenting intervention, while control AWCs continue with existing services. The study follows a cluster randomization design, covering over 4,000 households with children aged 0–28 months.

The evaluation is designed to provide credible evidence on whether digital-first approaches can improve child outcomes at scale and inform future early childhood policies and programs in India.

External Link(s)

Registration Citation

Citation

Afroz, Bilal et al. 2025. "Evaluating the Impact of a WhatsApp-Based Early Stimulation Program on Child Development and Parental KAP: A Randomized Controlled Trial in Rural and Urban India ." AEA RCT Registry. September 12. https://doi.org/10.1257/rct.16756-1.0

Sponsors & Partners

Interventions

Intervention(s)

This study uses a cluster randomized controlled trial (RCT) to evaluate the impact of Rocket Learning’s birth-to-three digital parenting program. The intervention delivers weekly early stimulation content through WhatsApp to caregivers of children aged 0–3 years.

Clusters are defined at the Anganwadi Centre (AWC) level, and AWCs in the study districts were randomly assigned to either treatment or control groups. Caregivers in treatment areas receive the intervention, while control areas continue with services as usual.

The trial will measure program effects on child development (motor, cognitive, language, and socio-emotional skills) and parental knowledge, attitudes, and practices (KAP).

Intervention (Hidden)

This study is a cluster randomized controlled trial (RCT) designed to rigorously evaluate Rocket Learning’s birth-to-three digital early stimulation program in Haryana, India.

Unit of Randomization: Anganwadi Centres (AWCs) stratified by geography (rural/urban) within each district (Fatehabad and Panipat). A total of 309 AWCs were randomized (167 treatment, 142 control).

Sample: 4,281 households with children aged 0–28 months were surveyed at baseline. Only households with at least one smartphone were included, as the program relies on WhatsApp delivery.

Intervention: Treatment group caregivers receive weekly WhatsApp messages promoting responsive caregiving practices, early play activities, and stimulation techniques aligned to children’s developmental stages. Control group caregivers continue with existing services.

Measurement Tools:

Child outcomes measured using the Caregiver Reported Early Development Index (CREDI long form).

Parental outcomes measured using a 49-item KAP (Knowledge, Attitudes, Practices) scale, co-developed with academic experts.

Caregiver mental health module included for additional analysis.

Power & MDES: Sampling powered to detect a minimum detectable effect size of 0.15 SD in child outcome scores with ~80% power.

Analytical Strategy: Primary estimation will use an intention-to-treat (ITT) framework, with regressions controlling for child and household covariates, and standard errors clustered at the AWC level. Heterogeneity analyses are pre-specified by child age bands, caregiver education, gender, and geography (rural/urban).

The midline is scheduled for October/November 2025, with an endline to follow. This design will allow credible attribution of differences in child and caregiver outcomes to the intervention.

Intervention Start Date

2025-05-01

Intervention End Date

2026-05-31

Primary Outcomes

Primary Outcomes (end points)

Child Development Outcomes (CREDI Raw Scores), Parent ECD KAP (0-100 scale)

Primary Outcomes (explanation)

Child development will be measured using the Caregiver Reported Early Development Index (CREDI) long form, which captures developmental milestones across four domains — motor, cognitive, language, and socio-emotional. Scores are constructed as raw scaled scores on a 0–100 scale, with higher values indicating more advanced child development relative to age norms.

Parental outcomes will be measured using a 49-item KAP tool developed in consultation with early childhood experts. Items cover knowledge (e.g., understanding of child development), attitudes (e.g., beliefs about responsive caregiving), and practices (e.g., frequency of stimulation activities). Items are combined into standardized indices and reported on a 0–100 scale, where higher values indicate stronger knowledge, attitudes, and practices related to early childhood development.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

This study uses a cluster randomized controlled trial (RCT) design to evaluate Rocket Learning’s birth-to-three digital parenting program. The program delivers weekly early stimulation messages to caregivers of children aged 0–3 years via WhatsApp.

Randomization was conducted at the Anganwadi Centre (AWC) level in two districts of Haryana, India (Fatehabad and Panipat). Households in treatment clusters receive the intervention, while those in control clusters continue with existing services. The study will compare outcomes in child development and parental knowledge, attitudes, and practices (KAP) between the two groups.

Experimental Design Details

This study employs a cluster randomized controlled trial (RCT) to assess the impact of Rocket Learning’s birth-to-three early stimulation program. The intervention delivers weekly WhatsApp messages with age-appropriate activities, play routines, and responsive caregiving prompts to parents of children aged 0–28 months.

Clusters and Randomization: A total of 309 Anganwadi Centres (AWCs) across two districts (Fatehabad and Panipat) were stratified by geography (rural/urban) and randomly assigned to treatment (167 AWCs) or control (142 AWCs) groups. Randomization was conducted by the research team using computer-generated random numbers.

Sample: Within each selected AWC, all households with children aged 0–28 months were listed. Households with at least one smartphone were included (99.8% of eligible households). At baseline, 4,281 households were successfully surveyed.

Measurement:

Child outcomes: assessed using the Caregiver Reported Early Development Index (CREDI long form), covering motor, cognitive, language, and socio-emotional domains.

Parental outcomes: measured through a 49-item Knowledge, Attitudes, and Practices (KAP) tool developed in consultation with academic experts.

Additional module: caregiver mental health using the Hopkins Symptom Checklist (HSCL).

Study Power: The study is powered to detect a minimum detectable effect size (MDES) of 0.15 SD in child development outcomes, with 80% power.

Timeline: Baseline completed in April 2025, midline scheduled for Oct–Nov 2025, and endline planned for 2026.

This design allows for robust attribution of impacts to the intervention and will generate policy-relevant evidence on the effectiveness of digital-first parenting programs in low-resource settings.

Randomization Method

Randomization was carried out by the research team using a computer-based random number generator (Stata). Within each district (Fatehabad and Panipat), Anganwadi Centres (AWCs) were stratified by geography (rural/urban) and then randomly assigned to treatment or control groups.

Randomization Unit

The unit of randomization is the Anganwadi Centre (AWC). All eligible households with children aged 0–28 months within a selected AWC were assigned to the same study arm (treatment or control).

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

309 Anganwadi Centres (AWCs) across two districts (167 AWCs assigned to treatment; 142 AWCs assigned to control).

Sample size: planned number of observations

4,281 households with children aged 0–28 months (baseline surveyed households).

Sample size (or number of clusters) by treatment arms

Treatment: 167 AWCs (2,376 households surveyed)
Control: 142 AWCs (1,905 households surveyed)

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

The study is powered to detect a minimum detectable effect size (MDES) of 0.15 standard deviations on child development (CREDI) outcomes with approximately 80% power, accounting for the cluster-randomized design. Power calculations used baseline estimates (mean, SD) and an intraclass correlation coefficient (ICC) drawn from the Birth-to-Three Pilot-1 (Feb 2024). Primary analyses will cluster standard errors at the AWC level and report intention-to-treat (ITT) estimates.

Supporting Documents and Materials

IRB