The Art and Science of Teaching by Robert Marzano is subtitled, "A Comprehensive Framework for Effective Instruction." I would suggest emphasizing "Framework" in that description over "Comprehensive", but that is not to say the book is bad. In fact, I recommend this book, especially for the big ideas it discusses in some detail on effective teaching.
I've been meaning to write a brief review of this book as it's been given to me as a gift a long time ago. It's a solid book of big ideas with adequate examples of how those ideas might be implemented in a classroom. Ideas regarding, for example, how a teacher can "establish and maintain effective relationships with students," and how to "help students practice and deepen their understanding of new knowledge."
The ideas are great, and examples are good, and the ideas are backed up by research (that's the science part of the book). I'd love to try employing more of these ideas into my own classroom in the future. There are, however, two major flaws in the book I should mention.
First, the first flaw. As I do computer science research, I'm happy the book includes numbers, lots and lots of numbers, purportedly showing how certain recommended actions are backed by research. I haven't tracked down the sources, so I can't critique whether the background research is any good, but as presented, the numbers as presented in the book are rather useless, for the most part.
For instance, in Chapter 2, Marzano makes recommendations regarding summarizing information taught. There is plenty of research cited to show how great this technique is, with average effect size and percentile gains from various studies shown. However, what's not clear is what the comparison is made against. Is the control group a bunch of students who did nothing for a week after the lesson? Is the control group a bunch of students who were given free time to study the topic for a week?
It would have been helpful to see more of the methodologies of the studies cited. As it is, the numbers cited raises many questions, and I get the feeling that the numbers are cited just to convince us that the recommendations are right (using the tried and true proof technique called "proof by overwhelming the reader with numbers").
Now, for the second flaw. As a math teacher, I have to complain about the lack of examples of how the recommended teaching techniques could be applied in a math classroom. There are a few examples of this type, but the vast majority are geared towards non-math classrooms. The recommendations are even good for the science classroom, but the math classroom does have certain challenges that the book doesn't adequately address.
This isn't a flaw in the book per se. Teaching math is just challenging, and the least like the other subjects (the outlier in the data set, so to speak). With some imagination and teaching experience, it's not hard to see how most of the book's
recommendations can be applied in the math classroom. So I recommend the book for math teachers too.
Lastly, the book, as I mentioned, is more about frameworks than it is about comprehensiveness. It's about giving you the big ideas, with examples to show how to apply those ideas. If you want comprehensiveness, I recommend trying Understanding by Design.