Early Childhood Research Quarterly 26 (2011) 124–133 Contents lists available at ScienceDirect Early Childhood Research Quarterly Effects of a preschool intervention on cognitive development among East-African preschool children: A flexibly time-coded growth model Lars-Erik Malmberg a,∗ , Peter Mwaura b , Kathy Sylva a a University of Oxford, UK b Madrasa Regional Research Programme, East Africa, Kenya article info Article history: Received 9 October 2008 Received in revised form 15 April 2010 Accepted 22 April 2010 Keywords: Cognitive development East-Africa Intervention Child-centred pedagogy Multilevel model abstract The aim of the study is to investigate the effects of the Madrasa Resource Center (MRC), a child-centered intervention program, on East-African (Kenya, Zanzibar, and Uganda) preschool children’s cognitive development. Altogether 321 children (153 non-intervention and 168 intervention) participated in a cross-sequential study over three time-points during preschool (mean ages 4.3, 6.0, and 7.1 years). A multilevel model (MLM; time-points nested within children nested within schools), in which time was coded flexibly (i.e., child’s age operationalized as months from start of the intervention), showed a bene- ficial curvilinear effect of the intervention program on children’s cognitive gains. A moderation analysis suggested that the effect of observed preschool quality (ECERS) was stronger in the intervention program. The findings are discussed within the context of East-African preschool policy. © 2010 Elsevier Inc. All rights reserved. Education in developing countries, and in East (sub-Saharan) Africa in particular, faces several challenges, such as insufficiently trained teachers, overcrowded classrooms, low levels of cognitive stimulation and, as a consequence, deteriorating educational out- comes, and a high level of dropout due to the lack of both material and human resources (Kholowa & Rose, 2007; King, 2007). As in industrialized countries, the introduction of preschool is seen as a way of improving educational equity and, when in parallel with other measures, is a way of combating poverty and societal exclu- sion, promoting health care, and advancing societal inclusion. As preschool is a relatively recent phenomenon in East (sub-Saharan) Africa, it is important to know more about its potential benefits on children’s cognitive development. As a region, sub-Saharan Africa is among the most disadvantaged areas in the world, with poverty, malnutrition, short life-expectancy, high prevalence of diseases, high child mortality, and stunted child development (Alderman & Engle, 2008; Grantham-McGregor et al., 2007; Iglesias & Shalala, 2002; Liddell & Rae, 2001). In a previous study, we reported on the effects of the Madrasa Resource Center (MRC) child-centered intervention on cogni- tive development of preschool children in disadvantaged Muslim communities in Kenya, Zanzibar, and Uganda (Mwaura, Sylva, & Malmberg, 2008). To discern the intervention effect between onset and half-way through preschool, multiple regression models were ∗ Corresponding author at: Department of Education, University of Oxford, 15, Norham Gardens, OX2 6PY, Oxford, UK. Tel.: +44 (0) 1865 274047. E-mail address: lars-erik.malmberg@education.ox.ac.uk (L.-E. Malmberg). carried out, using raw scores of school readiness outcomes, and controlling the effects of age and a range of covariates between the start of preschool and halfway through the program (Mwaura et al., 2008). The findings show that children in the MRC intervention had larger verbal, non-verbal, and mathematical gains than the com- parison group, and both groups outperform children who stayed at home. In the present study, we go beyond that study in three ways. First, the intervention effects up to the end of preschool are investigated. Second, in order to increase precision of the cognitive measures, age-residualized scores are used in conjunction with a flexibly coded time-variable (coded as time elapsed from the first measurement). Third, we use a three-level multilevel model (MLM; time-points nested within children, nested within preschools) to account for differences in classroom practices across preschools. 1. Stimulation of children’s cognitive development in preschool Following models of Bronfenbrenner (1979; Bronfenbrenner & Morris, 1998), development is viewed as an interactive pro- cess between the individual and the environment. The influences vary as a function of the individual, the proximal environmen- tal context (here the classroom in a preschool setting), the more distal environmental context (here the policy context), and time (longitudinal follow-up from beginning to end of preschool). In order to extrapolate how a preschool intervention might be suc- cessful in stimulating children’s cognitive development in the East African context, we draw on a range of literatures: those on effects of health interventions, preschool attendance, and quality 0885-2006/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ecresq.2010.04.003