Metacognition
Can we train ourselves to judge our own minds better?
Open access · cc by · source: Europe PMC
Providing targeted feedback on the accuracy of people's confidence ratings can improve their overall ability to monitor their own performance, even on completely untrained tasks.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Metacognitive self-evaluation is partially domain-general, meaning an individual's ability to monitor their performance correlates across visual perception and memory tasks.
Evidence for the claim as stated.
Metacognitive abilities appear to be largely domain-general, as an individual's self-evaluation accuracy and confidence ratings are significantly correlated across different types of mental tasks.
Evidence for the claim as stated.
Metacognitive accuracy can be enhanced through targeted, adaptive feedback training on confidence ratings, and these improvements can transfer to entirely untrained tasks.
Evidence for the claim as stated.
Providing targeted feedback on the accuracy of confidence ratings can train and improve metacognitive calibration, and this improvement can transfer to untrained tasks.
Evidence for the claim as stated.
Study at a glance
- Design
- RCT — Eight training sessions of staircase perceptual discrimination: confidence-accuracy feedback vs first-order performance feedback
- N
- N=61 · Final analytic sample after 102 started; 35 women, mean age 38.1 (20–64)
- Population
- Adult online volunteers completing a 10-session protocol
- Outcome
- Metacognitive efficiency (confidence discrimination) after confidence-targeted vs performance-targeted feedback
Structured fields used in claim comparison tables when every cited study has a complete layer.
Inspect source passages
Can we train ourselves to judge our own minds better?
These passages occur verbatim in the stored paper text. They support the points listed below; they do not verify every statement in the explanation.
61 participants
Final analyses were carried out on a dataset of 61 participants (35 women, 26 men; M age = 38.1 years, age range = 20–64 years).
10 sessions
The experiment was divided into three phases: Phase 1, pretraining (one session) → Phase 2, training (eight sessions) → Phase 3, posttraining (one session), resulting in 10 sessions in total.
Key findings
The training program successfully enhanced metacognitive calibration (how well confidence ratings matched actual performance) exclusively in the experimental group, while the control group actually showed a decrease. This training effect successfully transferred to a completely untrained recognition memory task and a different stimulus type. While the experimental group's overall confidence level shifted rapidly at the start of training, their metacognitive efficiency (how well they distinguished right from wrong answers) improved more gradually over the course of training.
Methodology
Researchers recruited a group of adult online participants, ultimately analyzing data from 61 individuals who finished the entire study. Over a series of 10 sessions, participants completed visual perception tasks and recognition memory tasks. For eight training sessions, the experimental group received specialized point-based feedback based on the accuracy of their confidence ratings, while an active control group received points based on their perceptual performance. To keep task difficulty stable and fair, a computer program adjusted the difficulty for each person to keep their performance at roughly 75% correct.
Limitations
The study does not establish whether these metacognitive training effects persist over long periods, as participants were only tested immediately after training. Additionally, because the training and testing were restricted to computerized, simple multiple-choice laboratory tasks, it remains unclear if the improvements would transfer to real-world learning or clinical environments. Finally, while the composite calibration scores showed clear transfer, when analyzed separately, the evidence for domain-general transfer of metacognitive efficiency and confidence bias remained inconclusive.
Related papers in this topic
Same topic cluster — not a recommendation engine.