The “research based” label still carries some weight in education. To me, the term means that a given suggestion has been offered knowing that researchers have collected data warranting a positive opinion on whatever is being recommended.
One thing that tends to frustrate practitioners is that impression that advocates can always find research to support their recommendation. This is kind of true – a hedge. One attempt to deal with the complexity of research is to make use of techniques that aggregate multiple students. Aggregation can be done in multiple ways. A popular statistical technique is called “meta-analysis” and makes use of the statistical effect sizes (could be negative or positive) generated from multiple students. An easy way (not exact) would be to think of this as finding the average effectiveness of a technique.
Different folks have made doing aggregation studies their thing. In my own classes, I have long assigned meta-analyses conducted by Bangert-Drowns. If you have read much educational research, you may recognize this name. A new and influential player in this space is John Hattie. Dr. Hattie has generated lists of effective and ineffective educational tactics that some have found quite helpful. He has also written a book – Visible Learning for Teachers – that explains his findings.
As a prof, I work very hard to take a research-based approach. I have my students read original research associated with the different topics I want students who work with me to think about. Is all of this necessary when referencing a popular list as provided by Dr. Hattie necessary?
I recently encountered a blog post on this topic that encouraged me to offer my own opinion. Consider a position I sometimes find myself in. I talk about project based and problem based learning in my Instructional Design and Technology grad classes. These tactics appear on Hattie’s low effectiveness list and this ranking does not surprise me. I am aware of other aggregation efforts that have reached a similar conclusion. On the other hand, I am aware of specific, very well implemented studies (see the work of Deanna Kuhn) that show such methods can be quite successful. How do you recognize such findings? One approach might be to suggest that such tactics are complicated – some versions are not well implemented and some situations may not be suited to such approaches. It might be fair to suggest that as a general “best practice” any version of such techniques cannot be counted on. This is not the same as “this approach does not work”. Why this is the case becomes another topic for researchers. I know this sounds like the mantra – more research is needed. It may also suggest those who implement must give some careful thought to what makes a good use of some specific tactics.