In Zambia, 13% of the 15 to 49 year old population lives with HIV. The highest number of new HIV infections is among young people. To counter the spread of the disease, developmental and governmental actors are increasingly relying on educational behavior change tools. A particularly widely used tool, implemented by the German Development Corporation (henceforth, GIZ), is the so-called "Join-In-Circuit on AIDS, Love Sexuality" (JIC). The tool aims to improve a) HIV and sexual reproductive health knowledge, b) HIV testing uptake, and c) demand for health services. Previous research has investigated the direct effect of the JIC on knowledge about Sexually Transmitted Infections (STIs) as well as self-reported sexual behavior in Zimbabwe, and has found positive effects in both domains.

The research project evaluates the JIC in Zambia. The study randomly assigns 170 participating schools to five different JIC treatment arms. The first two arms represent control schools. Here, no JIC will be implemented. The third arm implements the JIC among a random subset of students. The fourth arm implements the JIC among indegree central students. The fifth arm implements the JIC among edge betweeness central students. In each school, the JIC will be implemented in one pre-determined grade. Within each school at least 30 students will be selected. For larger schools, 20 percent of students in the selected grade are selected.

The study's population is schools in the Choma and Livingstone provinces of Zambia (a final set of 170 schools). The schools were provided by the Zambian Ministry of Education. The schools are split between primary (grade 1 to 7) and secondary schools (grade 8 to 12). In order to not interfere with exam periods, the Ministry of General Education and GIZ decided to implement the JIC among children in grade 6 in primary schools, and grade 11 in secondary schools.

On average, the schools have 80 students and 74 percent of the sample are in Choma. 88 percent of schools are considered rural, i.e., they do not lie in a major town. 4 percent of the sample have received a JIC before.

The 170 schools were randomly assigned to one of five treatment conditions.

• Control 1 (No intervention) (34 schools)
• Control 2 (No intervention) (34 schools)
• Treatment 1 (Random selection) (34 of schools)
• Treatment 2 (Indegree centrality) (34 of schools)
• Treatment 3 (Betweenness centrality) (34 of schools)

The Columbia University (CU) team advised GIZ on the design of the study, and the intervention will be solely run by GIZ. The CU team is not involved in the implementation of the study. Data will be collected by the American Institutes of Research (AIR). The CU team will only analyze fully de-identified data produced by the study.

Inclusion Criteria:

• Students must be enrolled in the selected grade in one of the study schools and present on the day of the interview (for piloting) or on the day of baseline and endline (for main evaluation)
• Students must have informed consent
• Students must speak English or Tonga

Exclusion Criteria:

• There are no exclusion criteria within the pre-defined schools