The truth about percent correct in behavioral measurement
It seems everyone uses percent correct as a measure of behavior. Percent correct shows up in classroom and home behavioral implementations, theses and dissertations, and even books and journals. With so many people using percent correct, the measure must have many benefits. Cue the dramatic music and prepare yourself for a big reveal: percent correct monitoring seriously constrains information. In some cases, percent monitoring leads to compromised decision making. Join us to learn the truth about percent as a measure, percent evaluation, and percent monitoring. Also, we will discuss useful alternatives to percent for behavioral measurement.
What do the pro’s say about percent correct?
“academic frequencies (rates) recorded 40 times more effects of curricular changes than did percent correct...These and many other studies proved behavior frequencies significantly more sensitive to learning variables in the classrooms than percent correct and percent of time on task.”
Ogden Lindsley, 1990
“percentages cannot demonstrate proportional change (not to be confused with proportional values, which is the only demonstration they are capable of), percentages obscure trends or produce illusory trends, and percentages have low predictive power.”
Scott Miller, 2014
“Moving away from the percent correct measure not only more accurately reflects progress, but
also takes away the punishment element of measurement.”
Kent Johnson, 1992
Rick Kubina, Ph.D., BCBA-D, is Director of Research at CentralReach, cofounder of Chartlytics, Professor of Special Education at Penn State University, and author of The Precision Teaching Book.
Rick taught special education in a private schools specializing in post acute brain trauma rehabilitation for three years. During his doctoral program, Rick worked as a graduate teaching assistant for the Great Lakes Regional Resource Center. He was the editor of the Journal of Precision Teaching and Celeration and is the author of many publications.