Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Early Childhood Development in the Slums of Cuttack, Odisha, India. A follow up
Last registered on December 28, 2020


Trial Information
General Information
Early Childhood Development in the Slums of Cuttack, Odisha, India. A follow up
Initial registration date
March 09, 2020
Last updated
December 28, 2020 4:48 PM EST

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Other Primary Investigator(s)
PI Affiliation
PI Affiliation
PI Affiliation
Yale University
Additional Trial Information
On going
Start date
End date
Secondary IDs
This is a follow-up study of the registered experiment “Early Childhood Development in the Slums of Cuttack, Odisha, India” (AEARCTR-0000169). The intervention that was randomized across 54 slums, consisted on weekly home visits delivered by local women and targeting children between 10 and 20 months old during 18 months. Evidence from the first follow up reported positive effects in cognition (0.35 SD), receptive language (0.25 SD) and expressive language (0.20 SD). Yet, effects were larger for boys (cognition: 0.45 SD, receptive language: 0.45 SD and expressive language: 0.51 SD).One of the objectives of this project is to estimate long run impacts six years after the intervention, i.e., when children are between 6 and 8 years old. A second objective is to estimate spillover effects on siblings younger than 16 years old. Finally, we would like to understand intra-household allocation of resources in this context. We aim to follow the initial sample of 421 target children and their families. There is still a set of outstanding questions in the literature about whether or not impacts of very early interventions are sustained and which interventions are the ones leading to longer-run impacts. One of the aims of the current study is to examine whether the intervention had sustained impacts on child development six years after the randomization and to quantify them. Also, we would like to test whether there are spillover effects of the intervention on developmental outcomes of siblings of the original study children. Similarly we would like to examine intrahousehold allocation of resources of study households and their neighbours.
External Link(s)
Registration Citation
Attanasio, Orazio et al. 2020. "Early Childhood Development in the Slums of Cuttack, Odisha, India. A follow up." AEA RCT Registry. December 28. https://doi.org/10.1257/rct.5444-2.0.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
We carried out a stimulation programme aimed to improve the interaction between mothers and their infants in everyday activities by introducing a systematic week-by-week curriculum that followed the natural developmental stage of the child. The curriculum was delivered by local women (home visitor) during weekly individual home visits to 421 children, aged 10 - 20 months at the start of the intervention, and their caregivers. The key organising element of the home visit were the curriculum and protocols originally developed by Sally Grantham-McGregor specially adapted for the context of Odisha. The curriculum was progressive and enabled the facilitator to let the child dictate the pace. Emphasis was placed on improving maternal-child interaction, using positive reinforcement, and on using everyday child care activities and household tasks to teach the child new words and concepts. During the visits, the facilitator interacted with the mother and her young child and demonstrated play activities using home-made toys and objects around the home. She also discussed child development with the mother and emphasised the importance of chatting and play. As part of this research project, we ensured the curriculum was both age relevant and adapted to the cultural context. This is a follow-up study in which we are going to evaluate whether the initial effects were sustained six years after the randomisation took place.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Children cognition and language.
Primary Outcomes (explanation)
1. Children between 6 months and 3 years old: cognitive and language development (assessed using the Bayley-III and the Caregiver Reported Early Childhood Instrument, CREDI).
2. Children between 3.1 and 16 years old: cognitive and language development (assessed using the Wechsler Preschool and Primary Scale of Intelligence, WPPSI-IV and a set of tasks developed at the Spelke lab at Harvard University.
Secondary Outcomes
Secondary Outcomes (end points)
1. Parental beliefs on returns to educational investment given different endowments.
2. Intrahousehold allocation of resources given different levels of ability and health.
3. Quality of the home stimulation environment (as measured by the 'play activities' and 'play materials' subscales of the Family Care Indicators) in the home.
4. Maternal time spent on high stimulation activities with children (as measured by time use module in household questionnaire) in the home.
5. Educational expenditure: we will ask about educational expenditure in the previous month and year.
Secondary Outcomes (explanation)
Parental beliefs on returns to educational investment: the respondent is asked to estimate the returns to educational investments and how these vary by the initial conditions (or endowments) of the child. These initial conditions will be specified in terms of child health or schooling ability. The respondent will answer questions about different scenarios that vary in terms of 1) initial skills (low vs high) and 2) amount invested (low vs high). In each scenario, the respondent will be asked to estimate child’s wage during adulthood and child’s educational attainment.
Intrahousehold allocation: We formulate a game in which the primary caregiver is asked to allocate a given amount of resources between two children. The respondent will go through different scenarios that will be characterised by: 1) resources to be spent on children’s education, 2) children initial ability level, and 3) children initial health status. After presenting each scenario, the enumerator will distribute a set of beans to the respondent who will be asked to divide them between the two children.
Parental investment: The following instruments will be used:
1. Household environment (based on Family Care Indicator (FCI) questionnaire developed by UNICEF and the HOME): adapted to the age of the child.
2. Time use: we will ask about how many hours the child spends on a set of different activities in a typical week day of the previous week.
3. Educational expenditure: we will ask about educational expenditure in the previous month and year.
Experimental Design
Experimental Design
Intervention group:
27 clusters received psychosocial stimulation through weekly home visits - lasting around one hour - to mothers/primary care givers of children aged 10 - 20 months. The home visitors were local women and followed the Grantham-McGregor's curriculum and protocols specially adapted for the context of Odisha. The home visitors interacted with carers and children and discussed the importance of stimulation and play for child development with the carer.

Control group:
27 clusters without intervention.
Total duration of intervention: 18 months (2 periods of 1 month of data collection will precede and follow the intervention).
Experimental Design Details
Not available
Randomization Method
The randomisation was conducted by a computer in the office, using the statistical program Stata version 12. The steps followed were:
1. Identification of a sample of eligible slums: this means those with at least 7 children in the defined age range (10 - 20 months), excluding children with physical or mental disabilities and twins.
2. Stratification of the sample of slums by slum size, i.e. the stratification variable was an indicator = 1 if the slum has more than nine children in our target age range according to the initial listing conducted in the slums, 0 otherwise.
3. Within strata,
3.1 We randomly assigned a number between 0 and 1 to each slum;
3.2 The number of slums was sorted by the randomly assigned number;
3.3 We assigned the first half of the slums to treatment and the second half to control.
Randomization Unit
Slums with at least 7 children aged 10 - 20 months.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
54 slums
Sample size: planned number of observations
421 target children. 770 spillover children.
Sample size (or number of clusters) by treatment arms
27 slum intervention group, 27 slum control group.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The sample is designed to detect a minimum effect of 30% and 36% of one standard deviation (SD) of cognitive development on children cognition as assessed by the tasks developed at the Spelke lab at Harvard and the WPPSI-IV, respectively. The level of significance is fixed at 5%, power is fixed at 80%. The intra cluster correlation is fixed at 0.09 for the WPPSI-IV based on work done by members of the research team in India and at 0.1 for the Spelke’s tasks, based on previous experience in Ghana. Sample size requirements are 27 slums per treatment arm (54 total) and 7.5 children per slum on average. 8 children per slum on average were included to allow for some sample loss between baseline and follow ups.
IRB Name
University College London
IRB Approval Date
IRB Approval Number