Data and Code for: Informing Mothers about the Benefits of Conversing with Infants: Experimental Evidence from Ghana

Despite the well-established importance of verbal engagement for infant language and cognitive development, many parents in low-income contexts do not converse with their infants regularly. This paper reports on a randomized field experiment evaluating a low-cost intervention designed to boost verbal engagement with infants. The intervention entails showing recent or expectant mothers a 3-minute informational video and providing them with a themed wall calendar. Six to eight months later, mothers who participated reported a stronger belief in the benefits of verbally engaging with infants, more frequent parent-infant conversations, and more advanced language and communication skills of their infants. Treatment effects on objective measures of parent-child conversation (from a recording device) and infant language and cognitive skills (from surveyors' observations) were consistently positive but statistically insignificant. We find larger effects on objectively measured parent-child conversation immediately after the intervention, suggesting scope for a larger long-term effect had the behavior change stuck more. The intervention's potential for low-cost implementation via health clinics makes it a promising strategy for early childhood development in low-income contexts, particularly if complemented by efforts to support habit formation.

This is the data and code accompanying the article. None Response Rates: Of the 1,765 women approached, 1,462 were eligible. 283 were ineligible because of their child's age or due date, 17 because of their age, and 3 because they did not speak English or Dagbani. Of the 1,462 eligible women, 1,408 completed the survey and were administered the intervention. Thirty-eight chose not to participate partway through the baseline survey, 15 refused to participate, and 1 did not pass the COVID symptom screening.

We conducted the endline activities from September to December 2021, on average 6.4 months after the intervention. We completed interviews with 89\% of respondents, with no differential attrition between the treatment and the control group.
In March 2021, we employed a team of surveyors from Innovations for Poverty Action (IPA) Ghana to enroll a sample of prenatal and postnatal patients from the health clinics. Surveyors approached patients before/after their prenatal or postnatal clinic visits and, if the patients were interested, screened them for eligibility. In order to participate, women had to (1) be aged 18-40 years old, (2) have an infant or be pregnant with a child who would be 2-18 months six months later at the time of the follow-up survey, and (3) speak English or Dagbani (the main language in Tamale). Women visiting public health clinics in Tamale, Ghana for prenatal or postnatal checkups.

To be eligible, women had to (1) be aged 18-40 years old, (2) have an infant or be pregnant with a child who would be 2-18 months six months later at the time of the follow-up survey, and (3) speak English or Dagbani.
. coded on-site observation; cognitive assessment test; on-site questionnaire;