Evaluators from Health and Development Africa, Social Surveys, and Aka Associates sought to assess the process and impact of South Africa's Talk to Me, a nationwide initiative that includes a one-hour television special, a radio campaign, and a print component. This campaign, which emerges from the educational children’s TV series Takalani Sesame, is designed to "encourage caregivers and young children to communicate constructively about HIV and AIDS,” and to provide skills to support this dialogue and to reduce the stigma often faced by those infected with and affected by HIV and AIDS.

Between September and December 2004, Talk to Me materials were tested using a classic experimental design (a controlled trial). The study was conducted in Escourt, Kwa-Zulu Natal. Altogether, 349 young children and their caregivers completed interviews at two points in time (baseline and again 2 weeks after the intervention). A total of 1400 interviews was conducted, as well as two gender-based focus groups (with female and with male caregivers).

Communication between the caregiver and any child living in the household was significantly positively associated with the intervention.

• Overall, 54% of the control group and 75% of the intervention group reported having spoken to any child in their household about HIV in the previous two weeks at follow-up.
• After adjusting for communication at baseline, caregivers in the intervention group were twice as likely as those in the control group to report that they communicated with a child in their household about HIV in the previous 2 weeks ([odds ratio] OR=2.08; 95%CI [confidence interval] 1.08, 4.02).
• This association was statistically significant and remained so, even after adjusting for other potential confounders and effect modifiers (OR=2.46; 95%CI 1.14; 5.32).

When considering absolute numbers of children spoken to about HIV, the intervention effect was greatest in the 7-9 year old category. When considering the number of household children that the caregivers reported speaking to about HIV during the past two weeks by intervention category and age of child, the difference between baseline and follow-up in the control group (ages 7-9) was 26. In contrast, the difference between baseline and follow-up in the intervention group (ages 7-9) was 72. [The intervention effect was very small for the 3-6 year group on this measure, and also very small for the older age groups].

With respect to caregiver attitudes towards communicating about HIV with young children - The overall responses to the attitudinal questions were not consistent and did not show clear intervention effects. Measures that asked about caregivers' attitudes with regard to the different age groups showed no intervention effect for attitudes towards HIV communication with 3-6 year-old children but did show some evidence of effect for improved attitudes towards communicating with the 7-9 year-old age group.

With respect to children's attitudes towards various family arrangements - The main theme of the story booklet was alternative family structures and composition, and encouraging openness and acceptance of such structures. To measure children’s attitudes toward family arrangements, they were asked if they agreed or disagreed with the statements: "A proper family has a mother, father and two children" and "Many different people can take care of children". Children's attitudes using these measures showed no marked change between baseline and follow up and no intervention effect.