Learning Through Play and Early Childhood Education: Evidence from The Gambia

The intervention includes the piloting of a new play-based curriculum in selected structured public ECDs of The Gambia. The new play-based approach will be modelled on the Teyari ECE program successfully tried in Kenya. The Tayari (which is a Swahili word meaning “ready,”) is an Early Childhood Development (ECD) programme in Kenya is a four-year ECD and Education (ECDE) intervention funded by the Children’s Investment Fund Foundation (CIFF) and implemented in four counties in Kenya (Nairobi, Laikipia, Siaya, and Uasin Gishu. The intervention had four components: child-centred learning; child health support; teacher training; books and teacher guides (Ngware et al., 2018). The Gambia version of this model will focus on just component the child-centred learning and teacher training component with the introduction of a play-based approach – this will be the main novelty of this intervention. Specifically, preschool children in levels 2 and 3 (ages 3 and 4 respectively) in the targeted ECDs will be introduced to a new play-based curriculum which will put more emphasis on learning through play. Additionally, early childhood educators in the treated school will be provided in-service training on the new curriculum. Both interventions will be rolled out in targeted schools only.

2023-03-06

2024-01-02

The primary outcomes of interest are school readiness for children and classroom performance for teachers.

Following, Attanasio et.al (2019), school readiness is measured using the International Development and Early Learning Assessment (IDELA) toolkit prepared by save the children proposed by Pisani et.al (2015). This measure of school readiness is composed of 5 domains: numeracy, literacy, fine motor skills, executive functioning, and socio-emotional skills. The school readiness score is computed by counting the average percentage of right scores over all indicators of a domain.

In terms of the capacity of the ECD workforce to use play-based approaches, the intervention will include in-service teacher training on the use of play approaches developed for the targeted schools. In order to assess the performance of ECD educators, an evaluation tool for teachers will be used. This evaluation focuses on various relevant aspects of their teaching practices. It includes indicators of pedagogical expertise in the content of the ECD curriculum; the ability of teachers to establish a safe, inclusive, and respectful play-based learning environment for a diverse population of students; and teachers’ ability to plan and deliver effective instruction and create an environment that facilitates learning through play for their students. In principle, the evaluation of teachers will follow standard teacher evaluation tools focused on three areas: planning and preparation, classroom environment and instruction.

The improvement in learning and teacher capacity is expected to improve enrolment in public ECDs. Teachers' outcomes include knowledge of ECD, hours spend in the classroom and preparing activities, and classroom practices (responding to students). We will adapt the instruments proposed by Wolf et. al (2017) to assess the capacity of the teachers. Once the performance is assessed, feedback will be given to the teachers. This shall be the mechanism through which the teacher intervention will improve teacher performance. Some secondary outcomes such as improvement of the home environment, and health outcomes, among others, will also be of interest.

To evaluate the impact of the new curriculum on preschool children's readiness for primary education, a cluster randomized control trial approach was adopted. In particular, the targeted subjects for the evaluation are put into two groups: a treatment group that will be exposed to the intervention (will receive a new curriculum and the capacity of their facilitators will be strengthened on learning through play pedagogies) and a pure control group who will not receive the intervention. The assignment of children to each of the two arms was done using cluster randomization. They are first grouped into clusters and further stratified based on cluster size and district. Clusters will be randomly assigned to each of the two arms at equal proportions. For each cluster, a random number of the evaluation units (children and teachers or facilitators) are selected.

Randomisation of schools into treatment and control groups will be done using Stata. Data on the total number of structured public ECDs and student enrolment is available with MoBSE. This data will be used to select centres for the evaluation and randomly assign them to control and experimental groups. In each centre, 35 children will be randomly selected for evaluation. The children will be assigned to the 2 level classes of the centres. Where 2 level class size in a centre is less than 35, random assignment of the students to the 2-level classes will be done using probability proportionate to size randomization.

Randomisation of schools into treatment and control groups was done using Stata. Data on the total number of structured public ECDs and student enrolment was obtained from the Ministry of Basic and Secondary Education (MoBSE). The data were used to select centres for the evaluation and randomly assign them to control and treatment groups.

Randomization was done at the ECD centre or school level. Schools were further stratified based on districts.

Yes

100 school will be used for the evaluation. This was determined using the power calculation.

The total number of observations for the study was determined using power calculation (see details below). Accordingly, 35 pupils per school were obtained as the sample size per school leading to a total of 3500 observations overall.

50 schools treatment and 50 schools control; 35 students per school.

For the power calculation, the study assumed an average mean of school readiness of 26.1 percent as observed in Ghana (Attanasio et.al, 2019), and assuming an ICC of 0.08, the Minimum Detectable Effect (MDE) was 8.8 percent. This MDE is approximately 0.20 (20%) of the standard deviation.