Leveraging Parents and Teachers to Improve Foundational Numeracy

T1: Targeted Teacher Training – Grade 1–3 teachers in treatment schools will receive training to improve their mathematical content knowledge and pedagogical skills. The training will emphasize contextualized instruction, integrating real-world examples and practical problem-solving into numeracy lessons.

T2: Structured At-home Math Learning Activities – Parents of Grade 1–3 students will receive structured, easy-to-understand instructional materials to engage their children in simple, targeted numeracy activities at home. These materials will be disseminated digitally via WhatsApp and customer relationship management (CRM) software.

T3: Combined Teacher and Parental Interventions – Schools in this group will receive both T1 and T2.

2025-10-20

2027-06-30

Numeracy skills

We will assess the numeracy skill using an online assessment tool called PEMANTIK (Literacy and Numeracy Self-Assessment) developed by Pusat Studi Pendidikan dan Kebijakan (PSPK).

Students’ math anxiety, parental attitude and engagement in educational activities, and teachers’ understanding of mathematical concepts and pedagogical content knowledge.

Math anxiety will be measured using the Children's Mathematics Anxiety Scale UK (CMAS-UK), a widely used and validated tool for capturing students’ anxiety levels related to mathematics. We will also collect children's learning practices at school and home. For parental intervention, we will assess parental confidence in supporting their child’s learning, engagement in educational activities, awareness of their role in their child’s education, and beliefs about involvement in learning at home. These will be measured through a combination of phone interviews and computer-assisted self-interviewing. We will also measure teaching practice and teachers’ understanding of mathematical concepts and pedagogical content knowledge, which will be assessed through surveys and structured classroom observations. These instruments will be developed to be aligned with intervention content, with the structured classroom observation drawing from the TEACH Primary tool (World Bank, 2022).

The impact evaluation employs a 2x2 factorial design clustered Randomized Controlled Trial (RCT) to evaluate the individual and combined effects of two interventions: (1) Teacher training, and (2) Home-based numeracy activities.

The study involves 10,560 students in grade 1 and 2 at 264 primary schools across four districts in four provinces in Indonesia, their parents, and their homeroom teachers.

Eligible schools will be proportionally sampled and randomly assigned to one of the following four groups:
- Teacher training group: Grade 1-3 teachers in these schools will be trained to improve their mathematics teaching practices in the classrooms, especially in delivering the mathematics curriculum by using examples and problems drawn from everyday life.
- Home-based activities group: Parents of Grade 1- 3 students in the group will receive easy-to-read, specific instructions on how to conduct simple, targeted numeracy learning activities at home with their children.
- Bundled teacher-parent intervention group: Grade 1-3 teachers of these schools will receive the same intervention as the teacher training group, while at the same time, parents of these students will also receive the home-based activities program.
- Pure control group: Students and teachers in these schools did not receive any programs during the study period.

Not available

Randomization was done in the office using Stata.

School

Yes

264 schools

- 10,560 students - 10,560 parents - 528 teachers

- Teacher training: 66 schools
- Home-based numeracy activities: 66 schools
- Combination of both treatment: 66 schools
- Control: 66 schools

We determined our sample size based on a minimum detectable effect size (MDES) of 0.16 standard deviations when comparing each treatment arm, consistent with Kraft’s (2020) “medium” benchmarks for educational interventions. The power calculations adopt standard parameters commonly used in randomized controlled trials, such as a significance level (α) of 0.05 and statistical power (1-β) of 0.80. The Intra-cluster Correlation (ICC) is calculated based on the Early Grade Mathematics Assessment (EGMA) data from the Tanoto Foundations’ PINTAR program in 2023, which consists of 1811 students in grades 2 and 3 in 156 schools across 5 provinces. The EGMA consists of 96 questions across 9 subtasks, from Number identification to word problems, to measure a student's foundation skills in numeracy and mathematics in the early grades. Using several specifications to calculate the ICC, we found that the ICC is around 0.08-0.1, and 0.1 ICC is considered more conservative.