According to Hattie, different teaching practices have different effect sizes—here’s how to supersize visible learning.
In 2015, researcher John Hattie updated his seminal research Visible Learning. Hailed as “teaching’s Holy Grail,”1 Hattie synthesized 15 years of research on more than 800 meta-analysis about what works in the classroom. His goal was to focus educators around the idea that all students should make at least a year’s worth of progress for a year’s input.
Hattie found that most of the classroom activities we engage in have some effect on student achievement and even memorably noted “perhaps all you need to enhance learning is a pulse!”2 but he was also able to determine the average effect of classroom practices.
Hattie argues that unless a factor provides more impact than the average teaching activity, it shouldn’t be used to make decisions about what happens in classrooms.
Effect Size in Education
For those of us who aren’t statisticians, effect size works like this: Imagine you’re taking a road trip from Boston to Chicago. If you drive an average of 60 MPH, you’ll spend about 17 hours covering those 1,000 miles. Now imagine you can drive as fast as you like; 85 MPH cuts the trip down to 12 hours. Double it to 120 MPH and you’re rolling into Chicago in about eight hours.
Teaching practices work the same way. Cooperative learning, providing enrichment and afterschool programs have an effect size around the average of 0.4 (average impact). Things like charter schools, student gender and teacher’s level of education are around 0.1 (almost no impact,) while feedback, acceleration and formative assessment are around 0.7 (better impact).
Hattie’s goal was for us to use his research to develop practices that drive improved instruction and results.
Best of the Best
The 2015 update to Hattie’s original research uncovered some interventions that eclipse every other classroom activity with their effect on student achievement.
Conceptual change programs, self-reported grades and collective teacher efficacy all have effect sizes greater than 1.15. To put that into perspective, if you compared collective teacher efficacy at 1.57 to student control over learning at 0.01, 95 percent of your students in the “control” group would perform worse than the average student in the efficacy group.