REF-399 · Draganich C & Erdal K, 2014 · Journal of Experimental Psychology: Learning, Memory, and Cognition, · Paper · PubMed
Placebo sleep affects cognitive functioning
Participants
n=164 participants across 2 studies (undergraduate students — full text; not checked against the abstract)
Protocol
Participants reported their previous night's sleep quality
Random assignment to 1 of 2 sleep-quality conditions or 2 control conditions
"Above average" condition: told they had spent 28.7% of total sleep time in REM; "below average" condition: told only 16.2% in REM (the REM figure was presented as a measurement — full text; not checked against the abstract)
Cognitive tests: Paced Auditory Serial Addition Test (PASAT), Controlled Oral Word Association Task (COWAT), Digit Span, Symbol Digit Modalities Test (SDMT)
Key Results
Assigned (placebo) sleep quality — not self-reported sleep quality — significantly predicted scores on the PASAT and the COWAT
Assigned sleep quality did not predict Digit Span (as expected) and did not predict SDMT (unexpected)
Control conditions showed the findings were not due to demand characteristics
Authors' conclusion: mindset can influence cognitive states in both positive and negative directions
Mechanism
An expectancy (placebo/nocebo) pathway: being told you slept well or badly shifts performance on attention-demanding tasks in the direction of the message, while the participant's own report of how they slept did not predict performance. The effect was task-specific — two of four tests moved.
Practical Application
In 164 students told, falsely, that their REM sleep had been above (28.7%) or below (16.2%) average, the assigned sleep quality — not how they said they had slept — predicted performance on two attention-heavy tests (PASAT, COWAT), while two other tests (Digit Span, SDMT) did not move. What people believe about their sleep can shift cognitive performance in either direction.
Inference, not tested: a wearable's "poor recovery" or "poor deep sleep" readout is exactly this kind of assigned sleep quality — the athlete's belief in the number may cost more than the sleep itself; the study measured cognition, not physical performance
Coaches should consider limiting athletes' access to daily sleep scores before key sessions
A morning self-assessment ("How rested do I feel?") did not predict test scores here either — the finding is about the power of the assigned label, not the superiority of self-report
Limitations
Students, not athletes
Cognitive tests, not physical performance; two of four tests unaffected
Single session — no longitudinal tracking
How this entry was checked and changed
Content correction 2026-09-27
"If WHOOP/Garmin shows 40% recovery..." -> illustrative number not from the paper; removed, brand names dropped.
"Morning self-assessment is more predictive of actual performance than device scores" -> the abstract says self-reported sleep quality did NOT predict scores; corrected.
Added the abstract's null results (Digit Span, SDMT), the COWAT result and the control conditions; the undergraduate population is labelled as a full-text detail.