RANDOMIZED CONTROLLED TRIALS

Randomized Trials are the Simplest and Best Way of Assessing the Impact of a Development Intervention

Randomized Trials are the Simplest and Best Way of Assessing the Impact of a Development Intervention

Introduction

The word 'Randomized Control Trial' has its origins in clinical research. While, suggestions for the use of controls and other forms of Randomization developed as early as the 18th century, the theoretical and methodological foundation implemented in the earlier part of the 20th century. Furthermore, the wider adoption of RCTs for clinical studies by governments and the private sector did not begin until the 1950s. Clinical experiments have their origins in the experimental designs used for agricultural experiments in the 1920s by R. Fisher. It has been more than several years since the preference of Randomized Control Trial (RCT) methods and its use for assessing the impact of development intervention studies. Most of the time, researchers use this method as a requirement. Those who are in favor of using RCTs suggest that this method sets a methodological benchmark or 'gold standard' which ensures rigor and scientific validity. There are various arguments, which the proponents of RCT put forward causing debate on the subject.

Discussion

Fisher (1935) proposed Randomization of treatments among experimental units for the following reasons:

To protect from the use of judgment or systematic arrangements, which leads to one treatment at disadvantaged (i.e., to avoid bias)

To provide a foundation for the standard methods of statistical analysis, such as significance tests.

Fisher dealt with agricultural experiments where the experimental units were mostly plots of land. Up till now, these remain the reasons for the Randomization of treatments among experimental units in this trial.

There are various reasons for carrying out an impact evaluation and various views about what an impact evaluation intends to do. The significant reasons for conducting impact evaluations are as below:

1.Generate evidence on what works and what does not.

2.Measure impacts and relate the changes in dependent variables to developmental policies and programs.

3.Produce information that is relevant from an accountability perspective.

4.Benefit from individual and organizational learning.



It should be evident that, in order to achieve these goals, there needs to be a broad range of methodologies and the evaluator's toolbox should contain more than the RCT design. However, there has been an argument that the RCT approach is the best way of finding out what works, and what does not. RCT designs enable an estimation to be made of the difference in the average effects of the treatment and the control, under very stringent conditions that isolate the experiment from contamination by influential factors other than the treatment and the control (Scott, Anderson & Spaulding 2008, Pp. 139). However, there are two issues when attempting to use RCTs for the impact evaluation of development intervention:

1.The need to limit the indicator of impact to quantities that can be measured with enough accuracy.

2.For the methodology to work, the level of control needs to be good enough to render any contamination effect negligible

Within the context of development interventions, the solution to ...