Conversation exploring whether the experiment is a research method that is failing to provide insights for educational inquiry
Conversational context. Professor Gary Thomas recently published a challenging paper in the British Educational Research Journal. The title might be provocative for some Education students/researchers – ‘Experiment’s persistent failure in education inquiry, and why it keeps failing‘. As pointed out in this article, the term ‘experiment’ can be used quite loosely when describing methods of inquiry. It may sometimes mean no more than a deliberate effort towards ‘finding something out’. But typically ‘experiment’ has a stricter meaning: one based upon a structured observing of interventions and their different outcomes. From exploring such contrasts the experimenter seeks to identify causal forces. Although well established in the natural and biological sciences, Gary’s paper suggests this method has failed to deliver in educational research. The purpose of the conversation was to provide an ‘advance organiser’ for those who might go on to read the paper and/or discuss the issues raised.
Conversational themes. The discussion covers how the term ‘experiment’ is understood in educational research; the prominence of the randomised control trial; the relevance of good theoretical grounding to RCT designs, the relevance of ‘natural’ or ‘quasi’ experiments; small scale experiments that critically explore grand theory; connections to multi-variate modelling; the status of ‘local’ research inquiry in the individual classroom ecology.
Download conversation transcript in Microsoft Word format or in PDF format
Question posed
01.30: A question to confirm the distinction between experiment as ‘finding out’ and Experiment as a formal research procedure?
05.05: What do you see the USP (‘unique selling point’) of the experiment as a support for educational research?
05.15: Where is the experiment flourishing as a method in Education – is the random control trial (RCT) the main site of activity?
10.30: The journal issue carrying the present paper does not have any articles that are experiments – does this suggest the method is of minority interest?
13.23: Is there a contrast between the bulk of (published) research and the body of funded research – whereby the former tends not to involve experiments but the latter expects them?
15.34: Could it be that RCT research is relatively immature – in the sense that its designs are often not grounded on a sufficiently theoretical rich ‘substrate’?
19.35: Rephrasing of previous question to clarify “immaturity” – are RCT designs often theoretically unready?
23.00: Is there a danger is accepting RCT outcomes when showing an intervention does not work while resisting the RCT as a valid approach to finding out what does work?
25.10: Would you rule out ‘natural’ experiments (example of Luria research on Schooling in Soviet Union) because they involved the same kind of comparisons?
29.05: Would Piaget illustrate the kind of preferred intellectual disposition in relation to educational inquiry?
31.40: Scepticism about Piagetian stage theory would seem to have been the result of simple experiments (Donaldson, Hughes, Olson) – does this not vindicate the experiment?
33.22: Do you rule out multi-variate studies (example given) that infer causal relations from modelling patterns of data co-relation?
38.30: Would we not want to discourage the practitioner’s approaches to understanding a local context if it was based on experimental comparisons arranged in that context?
41.15: What might be the implications of this perspective for how we think about undergraduate and postgraduate teacher education?
Authored by:
Gary Thomas
Gary Thomas is Emeritus Professor of Inclusion and Diversity at the School of Education, University of Birmingham
Charles Crook is Emeritus Professor of Education and a member of the Learning Sciences Research Institute at the University of Nottingham
Nice and helpful information shared. . Good Work. keep it up.