Poster #29 - Beyond family characteristics: Effectiveness of a home visiting program to stimulate child development

• In Event: 3-001 - Poster Session 2 during Breakfast

Abstract

Training programs for caregivers on children stimulation have been shown to effectively improve the cognitive and language development of children (Baker-Henningham & Lopez Boo, 2010; Engle et al., 2007). A large-scale home visiting program in Colombia during 2010 and 2011 was effective at promoting the cognitive and language development of children at socioeconomic risk (Attanasio et al., 2014). In order to understand the populations more likely to benefit from this program, and help to design more cost-effective interventions, I analyzed whether the effectiveness of this intervention was moderated by four factors: Depressive symptoms of the mother, difficult temperament of the child, learning resources and activities available at home, and the corresponding cumulative risk.
I analyzed data from 1267 children of ages between 12 and 24 months from 96 municipalities across Colombia collected by Attanasio et al. (2014) from the UK Data Service repository (Fitzsimons, 2015). 639 children randomly received the home visiting intervention and 628 were part of the control group. I performed multilevel analyses on three dependent variables (cognition, receptive language and expressive language) as a function of the intervention, four moderators (difficult temperament, maternal depression, learning resources and activities, and their cumulative risk index), and three co-variables (baseline levels of dependent variables, age, and house’s wealth index). All variables of interest were introduced simultaneously in the models. In a second set of multilevel models, I introduced the interaction terms between the intervention and the four moderators.
Table 1 shows all multilevel models’ results, including direct and conditional effects of each independent variable. Direct effects suggest that the intervention was positively associated with results in cognition and receptive language, but not with expressive language. Besides, the learning materials and activities were positively associated with the three dependent variables. However, none of the interaction terms resulted significant.
The characteristics of this intervention are discussed in order to explain why none of the moderators were significant. Results support the importance of the implementation of this program in low-income families from countries like Colombia, as it is likely to be equally effective for children across the range of studied characteristics. This intervention may help to reduce children’s income-achievements gaps in cognition and language.

Author

• Tatiana Plata-Caviedes, Universidad de Los Andes
  Presenting Author