Strategy Spotlight: So what did Hattie really say about learning?
A brief review of his findings.
The ongoing debate and long-standing influence regarding the findings of John Hattie and his infamous book Visible Learning (2009) are never-ending, and can be a powerful discussion to have among colleagues.
· What did Hattie really study and what did he not study?
· What did he really say about what impacts learning (causation) vs. what only influences learning (correlation)?
To tell you the truth, it is probably best to watch some of the dozens of Hattie videos available online (Hattie in his own words) than to try to make your way through the original Visible Learning book. Yes, I mean the white one sitting on your shelf that you didn’t read.
The book is great, though it really is a complex read filled with small type, loads of data, and statistical goobly-gook.
In short, Hattie did this: He took all the educational research at the time and summarized the findings into what impacted learning the most. He didn’t really comment much. He just sorted the most effective strategies and factors from top to bottom.
We know he used a research methodology called meta-analysis, which simply puts all the studies (and their findings) on a common scale using a number (effect size). The best stuff has a larger effect size and the stuff with a lesser impact has a smaller size.
The findings are statistically sound, as long as we understand the limits of any research method. With this in mind, it is important to note that even the most prominent meta-analyses we have on instructional practices produce effect sizes that only approximate the impact on students. It is best to think of these findings as the average impact on the average student. When we think of an effect size, think of it as a typical effect on the typical student in a typical classroom.
I’m sorry, but what is an effect size again?
Determining the good classroom techniques from the very good ones is a critical step in our professional learning. In statistical terms, effect size is one way of measuring the degree of impact when sorting out and rank-ordering good ideas from bad ones.
The way Hattie tells it, a positive effect size of .40 or above is good but has only a moderate impact on learning, while an effect size of .80 or above is better and has a much larger impact. Here are a few evidenced-based strategies that are all good, but giving quality feedback to students is more impactful to their learning than test-taking skills. Yes, we can do all three, but we have to get our feedback loops down perfectly because learning hinges on that technique.
feedback to students (0.70)
peer tutoring (0.53)
teaching test-taking skills (0.30)
Essentially, an effect size is a type of measurement (a number) that expresses in standard deviation units the increase or decrease in measurable achievement (positive or negative) for a group of students who are exposed to a strategy (variable) being studied. Simply put, an effect size of zero means that the strategy had no statistical impact on learning. It neither positively nor negatively affected students in their learning of the material.
In turn, strategies with effect sizes above zero were shown to have positive impacts and those below zero had negative impacts. In this manner, the research attempts to pinpoint the degree of impact along a number line from least impactful to most impactful. An effect size at .40 or above is generally viewed as having a meaningful impact on learning.
In keeping with our number line analogy, effect sizes of 1.0 (one standard deviation above the norm) are particularly impactful, producing an average improvement on student achievement of at least 34 percent. Those variables plotted well below the norm have less impact and might be described as detrimental to learning. No matter our personal preferences, the strategies showing the highest positive effects across multiple studies are powerful practices that we need to discuss and consider adding to our repertoire.
What Hattie did NOT study was the impact of factors outside of school like poverty and socio-economic status. We already know that these things impact student performance.
So, this is the part where Hattie did something extraordinary. In simple terms, he asked this question: No matter where kids come from, what school and classroom factors help them learn the most? Forget their home lives because we cannot impact that.
In fact, Hattie didn’t even research his own question. He just summarized the findings of hundreds and hundreds of studies conducted by others.
The most important conclusion Hattie reached wasn’t so much that discussion is a good thing (effect size of 0.82) and retention is a bad thing (effect size of -0.32), but that some things really make a difference (causation) and some things only sort-of make a difference (correlation).
Yes, things like making eye contact with kids is a good thing, but kids don’t learn math because we make eye contact. Only things like challenging tasks, summarizing and note-taking, and feedback cause learn to occur.
He doesn’t really tell us to focus on the high-impact things, but he challenges us to discuss them as colleagues and make sure we teach our lessons with purpose and with learning as the only outcome.
So what did Hattie actually say…
In the end, let’s use his own words because so many education experts like to describe the findings of researchers like Hattie in their own terms.
In his own words…
Here are a few quotes from Visible Learning that you might discuss with your school team:
“The major message is simple - what teachers do matters (p. 22).”
“The act of teaching requires deliberate interventions to ensure that there is a cognitive change in the student (p. 23).”
“Of course, it helps if these learning intentions and success criteria are shared with, committed to, and understood by the learner (p. 23).”
“A major aim is to develop ‘over-learning’ or fluency of achievement…when a student attains a high degree of fluency on a topic, they have more cognitive resources to devote to the next phase in learning (cognitive tasks) (p. 30).”
“What matters are conceptions of teaching…teachers having expectations that all students can progress, that achievement for all is changeable (and not fixed), and that progress for all is understood and articulated (p. 35.)”
“It is also what learners do that matters. So often learners become passive recipients of teachers’ lessons…the aim is to make students active in the learning process until the students reach the stage where they become their own teachers, they can seek out optimal ways to learn new material and ideas, they can seek out resources to help them in this learning, and when they can set more appropriate and challenging goals (p. 37).”
——————————————
See there…now you don’t have to read the book. 😄 And here’s a thanks to all the researchers who push our thinking on behalf of kids.
Dan Evans, Ed.D.
Post your thoughts below →