What is it about?
Randomised controlled trials (RCTs) are well-known from the medical world, for instance where a group of patients is divided at random into those who get the experimental drug and those who receive a non-active "placebo". A similar approach is becoming increasingly popular for assessing educational initiatives, especially in England. This is very contentious in the research community, because some researchers think RCTs can't reflect the richness of the classroom or are simply inappropriate.
Featured Image
Why is it important?
This article uses examples of two completed studies to argue that the most successful approach is often to use RCTs along with other types of research. Research methods should be chosen depending on their ability to answer a particular research question, not because of a researchers' allegiance to a certain approach. A combination of techniques can enable us to answer not only ‘what works?’ but also ‘why?’, ‘where?’ and ‘how?’.
Perspectives
Although the article may seem to be arguing for commonsense - "horses for courses" - it is surprising how often, to keep the equine analogy, researchers can get on their high horse about methodology. I have had well-respected researchers telling me RCTs can't possibly be used in education because there is so much variation, unlike the medical world. I'm sure medical researchers would have a lot to say about that, but it totally ignores the point that a well-constructed RCT is designed to address the problem of variation among schools, teacher, pupils etc. On the other hand, and this is becoming less frequent, some keen RCT-ers have been known to overlook the valuable insight measures other than academic or psychological tests (such as surveys, interviews and focus groups) can provide.
Dr Pam Hanley
University of Huddersfield
Read the Original
This page is a summary of: Reassessing RCTs as the ‘gold standard’: synergy not separatism in evaluation designs, International Journal of Research & Method in Education, February 2016, Taylor & Francis,
DOI: 10.1080/1743727x.2016.1138457.
You can read the full text:
Contributors
The following have contributed to this page
