Individual Submission Summary
Share...

Direct link:

Presentation #4: Letting communities take charge: Key areas of analysis and results

Tue, April 16, 10:00 to 11:30am, Hyatt Regency, Floor: Street (Level 0), Plaza

Proposal

We analyzed a number of outcome variables that we classify into the following 12 themes of interest. The first five themes represent our primary outcome variables: (i) access and attendance, (ii) children’s learning achievement, (iii) sustainability, (iv) civic engagement and village institution legitimacy, and (v) national government legitimacy. The next seven themes indicate “mechanism” outcomes that we use to understand the causal pathways through which any treatment effects may have come about: (vi) barriers to school access, (vii) the amount spent per eligible child, (viii) ways that funds are applied in communities, (ix) community involvement in managing education, (x) community satisfaction with education services, (xi) teacher motivation, and (xii) teacher performance.

We also analyzed a number of moderator variables that allow us to assess conditions that affect the functioning of the sustainability model. We estimate the moderating effect of the following variables on attendance, learning, and sustainability: (i) capacity of the administrating organizations (NGO vs community institutions), (ii) teacher qualifications, (iii) levels of community engagement, (iv) quality of NGO services prior to handover, (v) barriers to accessing school among household, (vi) village population size, (vii) household economic conditions, (viii) children’s age, and (ix) social institutions and ethnicity. Finally, we estimate the moderator effect of gender on attendance and learning.

We tested two types of hypotheses: equivalency hypotheses and non-equivalency hypotheses. We run equivalency hypotheses when the goal was to check whether the sustainability model operates in a way that is no worse, or at least not substantially worse, than NGO administration. This applies to the access and learning outcomes, for example. It also applies to certain moderator effects, such as the moderating effects of gender and ethnicity. In all of these cases, we were less concerned as to whether the sustainability model is superior to NGO administration; we wanted to check if it is not substantially worse.

We choose non-equivalency hypotheses when our interest is to assess whether the sustainability model was superior to NGO administration. This applies to the government legitimacy outcomes as well as the mechanism outcomes. For example, we hypothesized that the civic engagement and village institution legitimacy among villagers would have increased in the sustainability model communities relative to communities where NGOs continued CBE administration, because, (a) now the village community institutions are directly managing the CBE classes in the village, (b) civic engagement and mobilization was one of the topics covered in the institutional capacity training in the sustainability model communities. We also tested whether teachers recruited based on qualifications contributes to the sustainability using non-equivalency hypotheses. This presentation will share the key outcomes of these analyses.

Author