Parental Misperceptions on Child Nutrition in India: Implications for Child Feeding Practices and Growth

Treatment Arm 1: Provide information on child's true relative nutritional status

Treatment Arm 2: Provide information child's true relative nutritional status AND effects of child undernutrition on long-term health, education, and labor market outcomes

2024-09-18

2025-02-28

1. Average willingness-to-pay for bundle of protein-rich foods
2. Difference between true and perceived height-for-age percentile relative to WHO standards
3. Difference between true and perceived weight-for-age percentile relative to WHO standards
4. Knowledge score on returns to child nutrition (Binary)
5. Minimum frequency of meals
6. Minimum dietary diversity
7. Height-for-age z-score
8. Weight-for-age z-score
9. Weight-for-height z-score
10. Consumption of Balamrutham
11. CREDI child cognition scale z-score

1. Average willingness-to-pay for protein-rich food bundle
All mothers who participate in the survey will be entered into a lottery to win a bundle of protein-rich food items for their child or, alternatively, a randomly chosen cash prize (amount may range from Rs. 100 to Rs. 2000). 25 lottery "winners" will be chosen randomly at the end of the baseline survey. Mothers will be asked to state their preferences between the food bundle and several potential cash prize amounts, using a multiple-price-list elicitation method. One cash prize amount will be randomly chosen for each mother, and their choice for that amount will be implemented in case they win the lottery. WTP will be measured by the mid-point of the interval of two cash amounts at which a mother switches from preferring to receive cash to preferring to receive food. Possible values range from 0 to 2000. Average willingness-to-pay will be compared between mothers in the treatment groups and the control group.
[Time Frame: Baseline]

2. Difference between true and perceived height-for-age percentile relative to WHO standards
The difference between the child's true height-for-age percentile relative to the WHO reference population and the mother's perceived percentile rank. Values may range from 0 to 100.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

3. Difference between true and perceived weight-for-age percentile relative to WHO standards
The difference between the child's true weight-for-age percentile relative to the WHO reference population and the mother's perceived percentile rank. Values may range from 0 to 100.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

4. Knowledge score on returns to child nutrition (Binary)
Binary variable coded "1" if at least half the prompts (3 out of 6) about the returns to child nutrition are answered correctly, and "0" otherwise. This is a binary indicator constructed based on the knowledge score scale that may range from 0 to 6, with higher scores representing better knowledge.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

5. Minimum frequency of meals
Binary variable coded "1" if the child consumed the minimum recommended number of meals in the last 24 hours, based on their age, and "0" otherwise
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

6. Minimum dietary diversity
Binary variable coded "1" if the child consumed food from at least 5 of the 8 specified food groups in the last 24 hours, and "0" otherwise. This is a binary indicator constructed based on the World Health Organization "Minimum Dietary Diversity - Infant and Young Child Feeding" (MDD-IYCF) scale. Scores may range from 0 to 8, with higher scores representing better outcomes.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

7. Height-for-age z-score
Height-for-age z-score at the time of the endline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

8. Weight-for-age z-score
Weight-for-age z-score at the time of the endline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

9. Weight-for-height z-score
Weight-for-height z-score at the time of the endline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

10. Consumption of Balamrutham
Binary variable coded "1" if the child consumed Balamrutham (government-provided therapeutic food) in the last 24 hours, and "0" otherwise
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

11. CREDI child cognition scale z-score
The Caregiver-Reported Early Development Instruments (CREDI) Short Form is a validated set of 20 population-level measures of early childhood development (ECD) for children from birth to age three (0-36 months). The responses on this 20-point scale (based on age) will be converted to a norm-referenced standardized Z-score for overall development. The z-scores may range from -6 to +6, with larger scores representing better outcomes.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

1. Change from baseline in height-for-age z-score
2. Change from baseline in weight-for-age z-score
3. Change from baseline in weight-for-height z-score
4. Change from baseline in height
5. Change from baseline in weight
6. Episodes of illness in last 14 days (binary)
7. Household food expenditure in last calendar month
8. Knowledge score on returns to child nutrition (Continuous)
9. Diet adequacy
10. Grams of protein consumed in last 24 hours

1. Change from baseline in height-for-age z-score
Difference between height-for-age z-score between the endline survey and the baseline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

2. Change from baseline in weight-for-age z-score
Difference between weight-for-age z-score between the endline survey and the baseline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

3. Change from baseline in weight-for-height z-score
Difference between weight-for-height z-score between the endline survey and the baseline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

4. Change from baseline in height
Difference between height between the endline survey and the baseline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

5. Change from baseline in weight
Difference between weight between the endline survey and the baseline survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

6. Episodes of illness in last 14 days (binary)
Binary variable coded "1" of the child experienced any episodes of illness in the 14 days prior to the survey
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

7. Household food expenditure in last calendar month
Total household expenditure on food in the last calendar month
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

8. Knowledge score on returns to child nutrition (Continuous)
The number of correctly answered prompts about the returns to child nutrition. Scores may range from 0 to 6, with higher scores representing better knowledge.
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

9. Diet adequacy
Binary variable coded "1" if the child meets the minimum frequency of meals AND diet diversity criteria
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

10. Grams of protein consumed in last 24 hours
Total grams of protein consumed by the child in last 24 hours
[Time Frame: During endline survey, an average of 4 months (or 17 weeks) from baseline]

The trial involves an individual-level field experiment with 1500 mothers of children aged 7 to 24 months, with two treatment arms and a control arm:

- Treatment arm 1: Update mothers' beliefs on the height-for-age and weight-for-age percentiles of their child relative to a reference group of healthy children based on WHO standards

- Treatment arm 2: Treatment 1 + information on the impacts of child undernutrition on long-term health (risk of chronic and infectious diseases, mortality), education (high school test scores, years of education), and labor market (earnings) outcomes, synthesized from existing literature

- Control arm: Status-quo, no intervention

Not available

Participants will be sampled and randomly assigned to one of the three study arms using a block randomization technique after stratification by sex and a baseline measure of nutrition based on anthropometric z-scores. The randomization will be done using statistical software (Stata or R) on a computer.

Individual

No

150 Anganwadi centers

1500 mothers

500 mothers control, 500 mothers Relative Nutritional Status, 500 mothers Relative Nutritional Status and Returns

Assuming a significance level of 0.05 and power of 0.8, a sample size of 1500 participants (500 in each treatment arm) will allow us to detect effect sizes between 0.10-0.20 standard deviations for all outcomes, including absolute differences of 3.5-5 percentile points (relative effect ~0.15 SD) for outcomes related to the beliefs of mothers on their children’s relative nutritional status, 4-7 percentage points (relative effect ~0.15 SD) for outcomes related to child feeding behaviors, and 0.14-0.20 points for the anthropometric z-scores (relative effect ~0.11 SD).