This preprint investigated whether sleep supports the retention and long-term maintenance of newly formed ability self-beliefs. The experiment used a between-subjects Sleep versus Wake design with a reported sample size of N=54. Participants formed new task-related ability beliefs based on experimentally provided feedback and then either slept or remained awake during a 12-hour retention interval following belief formation. The article is a preprint and has not been peer reviewed; additional methodological detail beyond the abstract should be consulted in the full manuscript.
Belief formation and revision were implemented using the Learning Of Own Performance (LOOP task). In the first session participants received mocked feedback intended to establish either relatively positive or relatively negative ability self-beliefs. In a later revision phase participants encountered contradictory feedback that allowed previously formed beliefs to be updated. The design therefore permitted measurement of both initial belief formation and the capacity to revise those beliefs when faced with counterevidence.
Belief states were assessed at four time points: immediately after the initial belief formation session; after the 12-hour retention interval that followed formation (during which participants either slept or stayed awake); after the revision phase in which contradictory feedback was presented; and a final assessment three weeks later to evaluate long-term maintenance. The key experimental contrast was whether sleep during the 12-hour interval affected short-term retention, resistance to revision, or long-term changes in recalled beliefs.
According to the reported results, participants successfully formed and revised ability self-beliefs using the LOOP task. However, contrary to the authors' hypotheses, sleep following belief formation did not enhance short-term retention across the 12-hour interval relative to wakefulness. Likewise, sleep did not appear to stabilize initially formed beliefs against later revision when participants encountered contradictory feedback.
These findings indicate that, in this paradigm, the immediate consolidation processes attributed to sleep did not translate into measurable protection of newly formed self-beliefs over a single 12-hour interval or into reduced susceptibility to belief updating in the subsequent revision phase.
Although no short-term benefits of sleep were observed, the authors report a selective long-term effect at the three-week follow-up. Participants who slept after belief formation showed a shift toward more positive recall for the initially low-ability self-beliefs when assessed after three weeks. This effect was described as a positivity shift specific to the initially negative/low-performance condition rather than a generalized strengthening of all formed beliefs.
The authors interpret this as evidence that sleep may facilitate a delayed, preferential enhancement of positive aspects of recalled ability beliefs, even though it did not promote short-term retention or stabilization against revision in this task.
The study authors are Julia Carbone and colleagues from the Institute of Medical Psychology and Behavioral Neurobiology, University of Tuebingen, and the Social Neuroscience Lab, University of Luebeck. The manuscript is a bioRxiv preprint (posted September 21, 2026) and has not undergone peer review; readers should interpret findings accordingly.
The authors declared no competing interests. Funder information reported in the source lists the Deutsche Forschungsgemeinschaft. The abstract does not report full participant demographics, detailed sleep measures (for example, sleep-stage composition), statistical effect sizes, or specific task parameters; those details are not reported in the abstract and should be checked in the full preprint and supplementary materials for evaluation and replication.
From the available summary, the study suggests that the role of sleep in consolidating complex, self-referential beliefs differs from its well-established role in consolidating many forms of declarative and procedural memory. Specifically, sleep did not produce short-term consolidation benefits for newly formed ability self-beliefs, nor did it protect them from revision, but it was associated with a selective positive change in memory for low-ability beliefs at a three-week delay. These findings prompt several research questions left open by the abstract: which sleep stages (if any) mediate the delayed positivity shift; whether the effect generalizes to clinically relevant self-beliefs (for example, in depression or anxiety); and what neural mechanisms underlie selective long-term valence-specific changes.
This rewrite is based solely on information reported in the article abstract and front matter. The abstract does not provide detailed methods (e.g., participant inclusion/exclusion, randomization, exact timing and nature of feedback, sleep monitoring procedures), statistical values, or effect sizes. Those items were not reported in the source abstract and must be consulted in the full preprint for definitive methodological appraisal and for translation of findings into clinical or experimental contexts.