Visible Learning: A Synthesis of Over 800 Meta-Analyses Relating to Achievement Paperback – Nov 18 2008

There is an enormous amount of published educational research available, and it seems to support just about everyone's point of view. Hattie has done a lot of work gathering and summarising data to produce this book, which attempts to compare a large number of different ideas on an even footing. Aggregating research about education and turning it into actionable knowledge is really challenging, and the kind of project Hattie has undertaken is something very important. However, a number of issues keep it from being particularly useful.

There appears to have been no consideration of data quality or causality while putting the research together. Data quality is clearly an issue - the average effect size in all the interventions identified by Hattie is 0.4. It is highly implausible that trying anything at all will result in an expected improvement of 0.4 standard deviations when state & national average results barely change at all, and when they do they go down as often as they go up. Not attempting to deal with this leaves a reader to their own devices to guess what may or may not be relied on.

Causality is also very important. Students' expectations of their own grades are highly predictive, and this is accorded one of the largest "effect sizes" in the work. However, this is of little importance to teachers, because it is not the case that students' expectations are a major cause of their success or lack thereof - it's mostly the case that students, being intelligent people with access to a lot of information about their ability, motivation and the challenges of the curriculum, are able to make accurate predictions of their success in it. While this may be nice to know, and there may be a small causal component to student expectations, making no distinction between causal effects and things that just happen to be associated with success again leaves the reader to guess as to which things might actually be good ideas to try.

Finally, Hattie's written summaries are poor. There are few explanations of what went on in the studies that make up the data summarised and fewer connections from these studies to Hattie's grand theories of educations. Under the heading of "feedback", coming with one of the largest effect sizes identified, Hattie explains that he spend a long time misunderstanding feedback until he realised feedback from students to teachers was important. Why is he confused about the difference between teacher->student and student->teacher feedback? Why are they collected under the same heading? Most importantly, how does this help me work out how to incorporate more effective feedback into my classes... should I be spending most of my time on surveys?

Overall, it is a good concept, but I didn't actually get much from reading it. I think the Education Endowment Foundation's Teaching and Learning Toolkit is a much better execution of the same idea.

Quality, research-based teaching practices with outcomes focused on student achievement! I even took notes!

This book is packed with statistical analysis of various practices in the education system. It was eye opening to see that some of the practices we are using have a negative or moderate impact on students -based on the research.
The only reason I gave the book four stars is due to the fact my book came with two chapter 7s and no chapter 8. I’ve missed the return date so if you’re not planning to read it right away you might flip through to make sure you have all of the chapters.

Great content, but entire CHAPTER 8 is missing from the book. It jumps from p.126 to p. 159. Ironically, this was the chapter I was most interested in reading. Hope it's the only book like that and not the whole batch. Buyer beware!

Very interesting way to evaluate school-wide and student-based data. Reader will need a grasp of basic statistical analysis. Hattie presents material very well and will challenge the reader to do deeper thought about data.