Natural rates of teacher approval and disapproval in grade-7 classrooms.
Middle-school teachers naturally scold twice as much as they praise — a gap you can close with simple tracking tools.
01Research in Context
What this study did
Researchers watched 12 grade-the teachers over the study period. They counted every “good job” and every “stop that.” No one told the teachers what to say. The goal was to see how much praise and reprimand happen naturally.
What they found
Teachers handed out two scoldings for every one praise. Praise rarely matched when kids were on task. Most teacher talk was correction, not celebration.
How this fits with other research
Justus et al. (2023) gave teachers a $3 hand counter. Teachers who clicked for each praise doubled their rate. The 1978 low-praise baseline shows why the counter helps.
May (2019) paired student choice with quick praise. On-task behavior jumped only when praise came right away. The 1978 data explain why praise feels rare to kids.
Joslyn et al. (2020) ran the Good Behavior Game even when teachers skipped steps. Disruption still dropped. Low natural praise may let sloppy GBG still feel better than usual class.
Why it matters
Your classroom probably sounds like 1978: more “no” than “nice.” Start every class with a 5:1 praise goal. Use a counter or tally sheet. Track praise for one week. You will reset the baseline that the 1978 study captured.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Put 10 pennies in your left pocket; move one to the right each time you give behavior-specific praise before 10 a.m.
02At a glance
03Original abstract
The natural rates of teacher verbal approval and disapproval in ten grade-seven classrooms were determined and compared with those described by White (1975). Although there were differences in the observation techniques used and the behavioral, cultural, and ethnic groups sampled, the results were similar. The majority of the teachers displayed individual rates of disapproval that were higher than their individual approval rates. The correlations between levels of on-task behavior and approval and disapproval rates were low. The issues raised by these findings are discussed in terms of directions for further research.
Journal of applied behavior analysis, 1978 · doi:10.1901/jaba.1978.11-91