'Hattie' and his meta-analysis: Do we need to rethink?
Hattie’s work on visible learning is such an impressive research on learning and education. Well such a huge meta-analysis is commendable and the 138 factors are such an interesting read. As I read more and more about the concept I found it even more interesting. I fell for it completely. But then nowadays, after my introduction to the very interesting concept of confirmation bias, I have become very skeptical of concepts and things I fell for easily.
Just to understand the concept of visible learning better and to feed my brain cells I searched more and more about the concept and I hit upon this Taylor and Francis article by Terhart, the full text for which I found at the VCC's digital library.
It is indeed an interesting read.
It criticizes the visible learning model by Hattie.
Most of my life I have dealt with teaching research methodologies and thus the part on the methodology was most interesting to me.
Simply putting it, when we talk of averages, they are often criticized for averaging out leaves you with losing out on information that is relevant. Meta-analysis is a technique in which we average out averages. And thus the quality of data or the original averages is very important for good results. Hattie's studies according to Terhart talk little about the quality of the studies he has selected for the meta-analysis. Terhart specifically quotes that and it is a well-known fact too, that in any meta-analysis the quality of the studies is crucial.
Terhart also talks about a lot of factors ignored and many a times over emphasized in Hattie's work. Further another criticism is the outdated nature of the data which has most studies from the 90's and the latest one from 2008.
Terhart argument that though the factors that Hattie talks about are stable but not the final answer or the Holy Grail as this pyramid of teaching is constantly being revised and reconstructed, is something that resonated with me.
