John Hattie’s Visible Learning (2008) is an oddity in educational publishing. It is an international bestseller in spite of being full of numbers and tables. What Hattie has done is he has compiled hundreds of meta-analyses related to student achievement, and combined them thematically to come up with 208 pages of “what works”. Now this approach sounds like a dull, but ambitious work, but Hattie has gone out of his way to make it accessible. He has created simplified graphics and short, snappy write-ups to make interpretation straightforward as possible.
First, the easy stuff. Almost everything works in the classroom. Give a teacher some students, some subject matter and some time and the kids will learn. Boring but true. It takes a superhuman effort to stop kids from learning. One of the few programs that achieved the nearly-impossible feat of slowing down student learning was Whole Language Instruction for reading. At least, according to the meta-analyses and Hattie’s synthesis. You can take that one to the bank—don’t offer whole language instruction, it’s just a bad, bad idea.
The big idea of Visible Learning, then, is that it helps teachers, schools, school authorities and governments pick and choose amongst a number of possible ways to invest their time and money. Here’s another unsurprising result—students benefit greatly from direct instruction in mathematics. If you want kids to be able to do math, it’s beneficial to give them clear, worked examples, accompanied by clear directions on how to proceed.
Teachers reading this are likely to be nearly jumping out of their skins to object. Surely our goals in education are greater than information retention and the acquisition of simple, testable skills. Of course. But as we struggle to teach children to be citizens, to understand scientific concepts, to apply their knowledge to like in the word, it’s only sensible that they possess the skills, understanding and dispositions that underwrite all these higher-level activities. The sorts of things that are amenable to statistical analysis, to meta-analysis and Hattie’s synthesis are not the full breadth of education. But they are important, and they should be taken into our deliberations.
The book is both ambitious and oddly modest. It’s ambitious in that Hattie has taken a wide array of educational research and has distilled into 392 pages of light prose, clever graphics and (mostly unhelpful) tables. Hattie appears to have provided a cookbook for educational reform. But he hasn’t. He is careful to note that you cannot use this book as a guide to educational planning, educational reform or daily teaching. The book does not provide the correct level of analysis for this. Rather, the book is a terrific first resource. Look at Hattie’s synthesis. Then go to the Bibliography and find the source studies. There is no way on this earth that a book of this magnitude does not contain errors, misinterpretations and debatable conclusions. This is not a shot at Hattie, it’s a reality of interpreting research. And in the end, teachers have to rely on experience and practice. It is foolish to dismiss the book and its results; it is equally foolish to accept every statement uncritically.
I have already written a short response to a few criticisms of Hattie focusing on a gruesome statistical error repeated throughout the book. As you can read, Hattie is very fortunate that this error does not damage his overall method and conclusions. Hopefully future editions will correct this. In future entries I will address a few other issues in greater detail. I’m a fan of Hattie’s work, but I have some serious misgivings about his methods and his conclusions. Stay tuned.