Updated version of ``Randomized Evaluation of a Parental Education Training Program and its Impacts on Children Early Childhood Development Outcomes in Nepal: Pre-Analyis Plan'': Short- and medium-term impacts

The NGO developped a curriculum defining the topics to be presented and discussed during the parental education training sessions. The training sessions took place every three weeks, for a total of 15 sessions, over one year of intervention.
The sessions covered all aspects of early childhood developpement (ECD) that are known to be relevant to increase child cognitive and non-cognitive development. Each session are organized so that theory is delivered, and practices are given. The caregivers are encouraged to practice and experience the promoted attitudes from one session to the next.

2023-04-01

2024-04-30

Participation to the training sessions, parental ECD knowledge, aspirations for the children, gender equity opinions, opinion about violence against children, time spent with the children for each type of activities, diet regime, use of violence against children, hygiene practices, child development skills (physical/motor, cognitive, emotional, social, language skills)

During the baseline survey, the latitude and longitude of the place of residence of all sampled
households were collected. GRM is divided into six administrative wards. Within each ward,
we used the households’ GPS coordinates to create groups of ten households on average based
on their proximity. Overall, we defined 95 parental groups, the mean number of parents within
each group was 10.6 (median value: 10, min: 5, max: 15), given that only one parent/caregiver
per household was invited to attend the sessions.
We then assigned to each parental group a meeting place where to attend the parental
sessions if the group was assigned to the treatment. Prayas Nepal provided us with a total of
161 possible venues (50 schools and 111 venues of other types such ward offices, women’s group
halls or, if no alternative, private homes). We assigned each parental group to the closest venue
to the households residence. Some venues ended up being allocated to more than one group due
to strong population density (especially in Gajuri city) or because of lack of alternative venues.
We restrict the set of possible venues to the ones closest to the households.
Once the groups and associated venues were defined, venues were randomly assigned either
to the treatment group or the control group. The randomization has been stratified at the ward level (6 wards in total) and at the type of training venue (school vs. other type) level.
Randomization was conducted using Stata. Among the 82 venues, 45 were randomly assigned
to the intervention whereas 37 serve as control group, such that 50 parental groups ended up in
the treatment group and 45 in the control group. The sample consists of 1,000 households: 519
in the treatment group and 481 in the control group. Figure 2 shows the location of the treated
and control households (in red and blue respectively), and the location of the venues (schools
with a red flag, and other types with a star).

Not available

Randomization done in office by a computer, checking for the balance on the main outcome variables at baseline.

The unit of observations is the cluster, defined geographically. it is a type of ad hoc neighborhood. defined by the venue location.

Yes

82 clusters

1,000 households

Among the 82 venues, 45 were randomly assigned
to the intervention whereas 37 serve as control group, such that 50 parental groups ended up in
the treatment group and 45 in the control group.