01
Praise, criticism, praise. Does it actually work?
flip
The evidence is split — and that is the useful part.
Some studies find the sandwich helps. Others find it changes nothing but how people feel. Neither camp is wrong. They were testing different situations.
02
More than a third of feedback attempts make performance worse .
flip
Worse than saying nothing at all.
Kluger & DeNisi, 1996: 607 effect sizes, 23,663 observations. Feedback helped on average (d = .41) — but over a third of interventions backfired. Positive feedback was no safer than negative.
03
People rate the sandwich highly. Do they then perform better?
flip
No. Perceptions, not performance.
Parkes, Abercrombie & McCarty (2013) put both outcomes side by side. Recipients liked the sandwiched version more. Their subsequent work did not improve.
04
But one experiment found it worked — using fake praise.
flip
Generic, unearned, pre-written — and it still helped.
Prochazka et al. (2020): 91 students, math problems. The sandwich group outperformed both the corrective-only and no-feedback groups. Yet the two positive statements were written before the study and had no link to anyone's actual work.
05
So what decides whether feedback lands?
flip
Where the listener's attention goes.
Attention can land on the task, on motivation, or on the self. Effectiveness drops as it climbs toward the self. “Wrong figure on slide three” keeps attention on the slide. “You're careless” moves it to the person.
06
When should you use the sandwich?
flip
When the bookends hold attention on the work.
The person is new to the task · another attempt is coming soon · the relationship is thin and the correction is small · you're writing, not speaking · they're already braced for judgment.
07
When should you skip it entirely?
flip
When the bread becomes a warning label.
They know the pattern · the message must not be missed · you do it every time · the praise is invented and specific enough to check · the relationship could carry plain directness.
08
One question worth asking before the conversation starts.
flip
Is the bread for their attention, or for my comfort?
If it keeps them focused on the work, use it. If it's there to make the conversation easier on you, drop it and say the thing plainly. That one question sorts most cases correctly.