| Sumario: | The challenge of differing parameter estimates from studies of the same treatment has been a focus of attention in the development literature, where randomized controlled trials (RCTs) have become the research method of choice. This is the usual set-up of formal synthesis strategies, which judge the quality (internal validity) of individual studies and then construct weighted averages of estimates across studies, implicitly assuming that each constitutes an independent estimate of that parameter. The CBO built its model by developing parameter estimates grounded in its own synthesis of the research literature. One way to think of these studies is as a series of estimates of the same parameter, with the results varying for methodological reasons (data quality, the use of simulated instruments or other techniques, questions about parallel trends).
|
|---|