Educational process factors for effective education in resource-constrained countries: a multilevel analysis

Abstract

Earlier conceptualisations of educational effectiveness magnified the importance of the need for significant amounts of fiscal and material resources to attain effective education. In the past, this has seemed to be justification for resource-constrained countries to seek mainly external support to fund their educational budgets in anticipation of attaining an effective education. Indeed,on many occasions any attempt to attain effective education in resource-constrained countries has been thwarted by the perceived lack of fiscal and material resources. Nonetheless, it is emerging that resource-constrained countries can actually have access to effective education. Using hierarchical linear modelling analysis, this study draws on the Southern and Eastern Africa Consortium on Monitoring Educational Quality (SACMEQ)database to generate an effective education model for resource-constrained countries, through a critical analysis of educational process factors that account for significant variations in educational outcomes.

The Rasch technique was used to construct most of the educational process indicators that were fed into the estimated multilevel models for reading and mathematics outcomes. On adjusting for pupil characteristics, contextual factors and school resource inputs, the process factors that significantly predict both mathematics and reading outcomes include opportunity to learn(OTL), school management competences, school-community relationships and school-based HIV/AIDS support. Further, for both mathematics and reading there is a significant interaction effect between teacher academic and professional capital (TAPC)and OTL; the effects of TAPC are completely mediated by OTL. On the other hand, whereas resource usage significantly predicts reading attainment, it does not predict mathematics attainment. Additionally, educational processes jointly explain more variance in mathematics attainment (16.5%) than that in reading(6%). Nonetheless, the preferred models explain about 25% and 26% of total variance in reading and mathematics, respectively. Overall, each of the two models explains more variance at Level 3(school level) than other levels. Unexpectedly, whereas there is inequity in the distribution of school inputs and opportunities for pupils to learn (OTL), there is limited evidence of inequity in the general distribution of learning outcomes by socio-economic status (SES)groupings.

The findings of this study extend the theory and practice of educational effectiveness,especially in developing countries where educational effectiveness research has always been limited to examining the potential impact of easily quantifiable educational inputs (using production functions) on educational outputs. Moreover, the study provides the various educational constituencies with sound evidence ofvarious educational process factors that could positively impact educational outcomes,and implores policy makers and practitioners to abandon input-output models for system-based models, simultaneously to pursue both quality and equity dimensions within educational outcomes and,most importantly, to refocus attention on the school and teaching processes

More information

Identifiers

ORCID for Daniel Muijs: orcid.org/0000-0002-0131-8921

Catalogue record

Export record