Uptake of children's swimming lessons: an experimental study on the role of parents and children

This RCT examines whether, and to what extent, a child's agency matters for the final outcome in a decision to pursue an opportunity to develop their human capital. To do so, we experimentally vary which parent the child must give a consent form to for a signature that permits them to participate in free swimming lessons (through clear instructions in the form): either the mother (``Mother-only form'') or the father (``Father-only form''). The treatments are randomly assigned at the household level, with one child per household offered the consent form. By experimentally varying the range of options available to the child, we obtain exogenous variation in children's agency.

On a sub-sample of students, prior to the RCT, we conduct a survey choice experiment (CE) centred on a child's autonomous decision to pursue an offer to build their aquatic survival skills through free swimming lessons. Specifically, the choice involves deciding which parent to approach for permission to begin swimming lessons. Their preference is captured through their choice between a consent form that must be signed off by the mother (``Mother-only form”) and one that must be signed off by the father (``Father-only form"). We randomise children into four treatments using a 2 X 2 factorial design. Treatments on the `Risk' dimension vary by the framing of swimming lessons as a neutral vs risky activity, and treatments along the `Conflict' dimension vary by whether or not children are made to consider the possibility of inter-parental conflict in decision-making around swimming lessons. In this way, we examine the role of children’s beliefs about parental preferences in shaping their autonomous decision on which parent to involve in the decision process around their human capital accumulation.

2025-02-03

2025-12-31

Our main outcome variable of interest is a binary variable for whether or not the consent form is signed by the assigned parent and returned to the school. For the sub-sample of students who participate in the CE, we look at their form choice, whether it is the Mother-only or Father-only form.

The RCT is conducted in three stages.

In the first stage, following an introductory welcome to the session, each child is interviewed individually by a trained same-gender enumerator. This includes questions on their socio-demographics, attitudes to swimming, household characteristics, beliefs about parental preferences for swimming, beliefs about parents' risk preferences, the nature of the relationship with their parents, perceptions of the intra-parental relationship, etc.

The second stage is a survey choice experiment, which only a sub-sample of children participate in. The child is asked to choose between two consent forms. The forms differ on the gender of the parent that needs to be given the consent form for signing. At the end of the choice experiment in the second stage, the child must answer some additional survey questions - these are used as buffer questions between the choice experiment and the RCT. After completion, the child proceeds to the third stage.

In the third stage, we begin the RCT. The child is now given a consent form that is randomly assigned to them. We give each child clear instructions about what they must do with the form when they go home i.e. that the form needs to be signed by one of the child's parents and brought back to their class teacher within 1 week. If the consent form is signed and returned on time, the child will have a chance to be selected for free swimming lessons. One week after the consent form is given to the child, we collect the forms that have been returned to the school. The forms are evaluated by a local swimming training outfit before a shortlist of students are contacted with a guaranteed place in a 2-week swimming training course.

Not available

We randomise schoolchildren into treatments through our survey instrument which is a an XLSform deployed through KoboToolbox. The form automatically assigns schoolchildren into treatments by their ID number.

We randomise treatments at the indivual level (schoolchildren), with one schoolchild per household.

No

26 schools in the Zanzibar archipelago (Pemba and Unguja islands); 5 primary and 5 secondary in the island of Pemba, and 8 primary and 8 secondary in the island of Unguja.

Stage 2 - 832 schoolchildren. Stage 3 - 1248 schoolchildren, of which 416 schoolchildren skip Stage 2.

Stage 2 - 832 schoolchildren -> 208 schoolchildren per treatment in four treatment arms.
Stage 3 - 1248 schoolchildren -> 624 schoolchildren per treatment in two treatment arms, of which 208 schoolchildren skip Stage 2.