Simplifying Complementary Feeding Practices: Experimental Evidence from Rural India

How to cite

Vaidya, Shardul; Radhakrishnan, Diksha; Mamidi, Pavan; Barnhardt, Sharon. 2021. "Simplifying Complementary Feeding Practices: Experimental Evidence from Rural India." AEA RCT Registry. AEARCTR-0007021. https://doi.org/10.1257/rct.7021-1.1

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

Shardul Vaidya* , Diksha Radhakrishnan , Pavan Mamidi , Sharon Barnhardt

* Primary investigator

Registry ID
AEARCTR-0007021
DOI
https://doi.org/10.1257/rct.7021-1.1
First registered
Last updated
Status
Completed
Trial dates
2020-10-31 – 2020-12-27
Countries
India (Madhya Pradesh)
JEL codes
I120
Registry record
http://www.socialscienceregistry.org/trials/7021

Abstract

In this randomised controlled trial study, we will evaluate the impact of two communication-based interventions, delivered remotely, on Complementary Feeding practices in rural India. Our interventions aim to simplify information for caregivers, make complementary feeding a social activity, and integrate relevant nutrition heuristics into communication being given to caregivers. We test for these using two experiments; in the first, our treatment is given simplified counselling over the phone, in our second, we assign mothers into moderated groups on WhatsApp. We deploy our experiments in rural Madhya Pradesh, with a sample of 699 mothers.

Intervention

We have two interventions that we test in our two separate experiments:

1. Phone Counselling: Mothers are given counselling on complementary feeding practices via a phone call from trained enumerators. The counselling consists of simplified instructions on CF practices, and calls last about 10 minutes in duration. Mothers are also asked to recall the practices to check their understanding of them. The call script also integrates the consequences of poor compliance of complementary feeding, to make the risks associated salient to mothers. In addition to calls, mothers are also sent three SMS messages on three consecutive days, highlighting the CF practices as well as the consequences of not adhering to appropriate CF practices.

2. WhatsApp Mothers Support Group: Mothers who have access to WhatsApp are randomly assigned to online WhatsApp groups of 3-4 mothers, where trained moderators provide information daily on complementary feeding. The group intervention combines simplified instructions on CF practices along with activities designed to encourage discussion and interaction amongst the mothers. We have also incorporated a few behavioral heuristics to make the need for CF salient to the mothers. We are including three mental models in this intervention:
a. Growth Mental Model: In this mental model, our counsellors -- through WhatsApp text posts, say that since the child is at a crucial stage in their life in terms of growth of the body and development of the mind, it is important that they get good nutrition - especially adequate quantity. If enough quantity is not fed, their growth can be impacted.
b. Starvation Mental Model: Here, we emphasise that all the parts of a child’s body are “hungry” - and not just the stomach. Therefore, it is important to feed the child different kinds of food that are made at home to babies now to feed all parts of their bodies.
c. Diet Diversity Mental Model: This model criticizes the similarities in the food that is given to sick adults and healthy children (porridge, rice/lentils) and also stresses that giving all kinds of food will not make the child sick, and in fact, should be encouraged. A healthy baby should be given the same food as a healthy adult.
Discussion in the groups was moderated for 14 days -- the treatment duration for this intervention.

Experimental design

Our research design is a straightforward randomised controlled field trial, where mothers with children in the age range 6-23 months are eligible for either the WhatsApp experiment or the Phone Counselling experiment -- depending on whether they have access to WhatsApp and consented to being part of a WhatsApp group.

Our treatments are bookended by a baseline and an endline, where we measure complementary feeding practices, attitudes and intentions, along with standard demographics.

Design details

Randomization method
Randomization was performed using the computational software R.
Randomization unit
For our phone counselling experiment, the randomisation is on a individual level, stratified by administrative block, age group child, and whether the mother has access to WhatsApp or not. The WhatsApp Mothers Support Group was stratified by the age group that the child falls under (6-11 months or 12-23 months), and whether the mother uses her own phone number for WhatsApp or has to depend on someone else for WhatsApp access.
Clusters
Clustering is applicable for the WhatsApp Support Group experiment. We have 25 WhatsApp Mothers Support Groups, as our treatment, consisting of 103 mothers.
Observations
Experiment 1 (Phone Counselling): 494 mothers Experiment 2 (WhatsApp Support Group): 205 mothers
Arms
Experiment 1 (Phone Counselling) Treatment: 235 mothers Control: 259 mothers Experiment 2 (WhatsApp Support Group) Treatment: 103 mothers Control: 102 mothers
Primary outcomes
We are most interested in measuring impact of our interventions on complementary feeding practices, measured using a 24-hour food recall survey. We will estimate treatment effects for 3 key indicator variables as defined by the WHO: Minimum Dietary Diversity (MDD), Minimum Meal Frequency (MMF) and Minimum Adequate Diet

Cited by

No citation data available for this registration yet.

Keywords