Abstract
This paper provides an overview of the various ways in which mixing qualitative and quantitative methods could add value to monitoring and evaluating development projects. In particular it examines how qualitative methods could address some of the limitations of randomized trials and other quantitative impact evaluation methods; it also explores the importance of examiningprocessin addition toimpact, distinguishing design from implementation failures, and the value of mixed methods in the real-time monitoring of projects. It concludes by suggesting topics for future research -- including the use of mixed methods in constructing counterfactuals, and in conducting reasonable evaluations within severe time and budget constraints.
Highlights
In recent years there have been increasing demands to measure the effectiveness of international development projects
In the last decade there has been an explosion of quantitative impact evaluations of programme interventions in international development
Some development economists (e.g., Banerjee, 2007; Duflo and Kremer, 2005) have gone as far as to argue that randomised control trials (RCTs) should be central to development practice, and that knowledge claims based on alternative approaches are not merely inferior, but inherently suspect
Summary
In recent years there have been increasing demands to measure the effectiveness of international development projects This demand has emerged in response to two concerns: (1) heightened criticism that most development agencies report only their outputs (e.g., number of teachers trained, kilometres of roads built) rather than outcomes; and (2) concerns that, despite assurances that development resources have contributed to the reduction of illiteracy and poverty, little reliable information has been presented to show that this is the case.. By restricting themselves to the econometric analysis of survey data, development economists are boxed into a Cartesian trap: the questions they ask are constrained by the limitations inherent in the process by which quantitative data from closed-ended questions in surveys are collected (Rao, 2002; Rao and Woolcock, 2003) As such, they are limited in their ability to ask important questions about the social, cultural and political context within which development problems are embedded.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
