Shifting parental belief to boost early parental engagement - RCT

How to cite

Ravà, Matteo Santangelo. 2025. "Shifting parental belief to boost early parental engagement - RCT." AEA RCT Registry. AEARCTR-0015112. https://doi.org/10.1257/rct.15112-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Matteo Santangelo Ravà*

* Primary investigator

Registry ID
AEARCTR-0015112
DOI
https://doi.org/10.1257/rct.15112-1.0
First registered
Last updated
Status
In development
Trial dates
2025-01-22 – 2026-03-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/15112

Abstract

This is an impact evaluation of a light-touch parenting program in the Philippines that aims to increase parental investment, targeting caregivers of newborns. The program is implemented alongside routine immunization. It leverages findings from behavioral science to build enduring habits in caregivers to engage in infant-directed speech (IDS) with their children early in life. The rationale for the intervention is that while simple verbal stimulation can enhance children's language and cognitive development, many parents are unaware of the potential returns of this simple practice for their children's growth. The intervention is inexpensive, and without any additional visit component, it is highly scalable. The design of the experiment is an RCT with no clustering and individual randomization. Our main outcomes measure children's development and the home environment. To the traditional self-reported measures we add long-form recording measurements. Sampling children's vocal production and their exposure to speech in the home using long-form recordings, collected via small recording devices worn on the child's clothes and analyzed using an advanced speech-processing algorithm, we aim to validate self-reported measure with high-quality data, that suffer less of potential reporting bias and measurement errors. We will examine heterogeneity in the effects of the intervention across caregivers' backgrounds, and the child's charachteristics, to determine the dominant factors influencing changes in parental behavior and subsequent child outcomes.

Intervention

The intervention comprises three main components. First, a 5-minute video is shown at the health facility, designed to welcome participants into the program. This video provides an overview of basic principles in early childhood development (ECD) and emphasizes the importance of communication with children and playful serve-and-return interactions. The second component of the intervention is a behaviorally informed calendar, which serves as a tangible resource for participants. It contains important information shared in the video and suggests daily activities that can be done with their child. The third component consists of a series of text messages, sent to the caregiver’s primary phone. These messages are coordinated with the calendar; because the calendars cannot be personalized for individual children, the texts focus on supporting the parent with personalized encouragement and specific activities linked to the child's age.
The intervention will last the full year of 2025. Most families will receive the intervention at the end of January and February for logistical reasons, so being effectively enrolled for a bit more than 11 months to a minimum of 10 months.

Experimental design

The experimental design is an RCT with individual randomization, with data collection at health facilities. The baseline will take place in January-February 2025, with an endline planned for the end of 2025.

The experimental design is an RCT with individual randomization, with data collection at health facilities at baseline and home visit at endline. The baseline will take place in January-February 2025, with an endline planned for the end of 2025.

The study is designed as a randomized control trial aiming to study the impact of the proposed intervention, as well as how parental beliefs and preferences affect parental engagement with children. Caregivers in the city of Antipolo, Rizal will be enrolled with their 0-1-year-old children in a 12-month study. In the study, we will collect information at 2 points in time. At the beginning of the study, we will have a baseline survey for caregivers and an assessment of their children. The data collection will also entail, for a subsample of 1000 participant, a long-form audio recording, which is an audio sampling done with a small recorder, that is worn in the t-shirt of the infant. Half of the study population will receive an intervention consisting of a short video showing, a calendar, and a series of text messages. At the endline, 12 months later we will collect information on parental engagement, belief, and knowledge. We will also have an assessment of the child, and we will conduct a second long-form audio recording.

Design details

Randomization method
The randomization is done through a survey platform that generates a random number which assigns caregivers to either the treatment or control group. If treated caregivers consent to long-form audio recording, they will receive the treatment after the recording devices are collected during a home visit. The randomization is done through a survey platform that generates a random number which assigns caregivers to either the treatment or control group. If treated caregivers consent to long-form audio recording, they will receive the treatment after the recording devices are collected during a home visit. There is a minimal chance of contamination for those in the control group, that may see someone else receiving the intervention at the health facility. This is minimal though because, only 200 people receive the intervention at the clinic. The enumerators are instructed to interview patients when they are about to leave the health facility. The contamination would be only for one part of our intervention, the video. Considering other similar experiment, this does not seems to be an issue. Also information alone, without the calendar and the SMS is unlikely to bring sustained behavior change. In any case we will measure knowledge and expected return or both control and treatment group, to measure any potential spillover.
Randomization unit
Individual randomization
Clusters
10 health facilities, with the possibility to expand if take up rate is low.
Observations
1600 caregiver-child dyads
Arms
800 caregiver-child dyads in the control group - 800 caregiver-child dyads in the treatment group
Primary outcomes
Full sample Caregiver-reported measure of children's vocabulary - adaptation of the MB CDI Self-reported index of child engagement MICS validated module Multidimensional Child assessment CREDI. (Depending on funding) Sub sample with recording Metrics of speech production derived from long-form audio recordings, including total adult word counts, total child speech-like vocalizations, and total adult-child conversational exchanges Caregiver-reported measure of children's vocabulary Self-reported index of child engagement
Secondary outcomes
Perceived return of early engagement Knowledge child development Female adult word counts Male adult word counts Words produced during instances of infant-directed speech (indicated by e.g., above-average pitch)

Cited by

No citation data available for this registration yet.

Keywords