A meaningful experience opens questions. Testing what follows requires distinct measures. This comparison separates reported awareness, confidence accuracy, everyday decisions and lasting care.
AI-assisted editorial synthesis; independent methods review pending. Older studies are labelled by original date. This page does not announce four new studies or a complete systematic review. Approximate reading time: 7 minutes.
Four claims, four evidence requirements
Claim
Measure needed
Unresolved transition
I noticed my thoughts.
Clearly attributed reports, repeated sampling and a defined practice.
Does reporting more noticing reflect a change in experience, vocabulary or expectations?
I judged my performance more accurately.
Confidence linked to correct and incorrect responses, accounting for task performance and response bias.
Does the effect transfer between memory, perception and decisions?
I made better decisions.
Prespecified behavioural outcomes with meaningful consequences.
Does accuracy improve choices outside the test?
I acted with lasting care.
Repeated observed helping, consent-aware follow-up and assessment of harms.
Of 137 recruited, 124 remained: 61 practitioners and 63 non-practitioners. Confidence–recall correlations were 0.52 and 0.53 respectively, without a significant group difference. Practitioners reported higher mindfulness. Self-selection, age/education differences and one prompted memory task constrain inference; this is not a randomised intervention.
The paper reports preregistration at OSF k8hcv; timing/files were inaccessible. Bayesian tests were additional analyses. Removing confidence-response trials over five seconds (3.5% of trials) was not preregistered. Multiplicity was not alpha-corrected. Numerical group-effect intervals/Bayes factors remain unverified.
Scope: abstract and selected methods/results/discussion, not complete tables, supplement or data/code audit. DFG/Bern funding; conflict statement unverified. Open-access article; repository reuse unverified. Relevance 5/5; provisional quality 3/5; moderate confidence. Added assessment: separate registration, exploratory analysis and trial exclusions. This null result does not establish equivalence.
A reported preregistered pre/post study compared 31 days of mindfulness with control training. It found no training effects on intertemporal decisions, confidence, metacognitive sensitivity or precommitment against preference reversals. Authors report Bayes factors favouring the null.
Analysis N, allocation/blinding details, attrition, numerical Bayes factors, intervals, registration timing and data/code remain unverified. No competing interests declared. Scope: indexed primary abstract/metadata; publisher full text blocked. Reuse unverified. Relevance 5/5; quality withheld pending methods; moderate confidence in the abstract, limited confidence in generalisation. Added assessment: a direct test of decision-monitoring transfer, not a universal verdict on mindfulness.
FOUNDATIONAL EVIDENCE · POSITIVE AND NULL OUTCOMES
A positive memory result with a perceptual boundary
Fifty students were randomised to two-week meditation (26) or active nutrition training (24). Six did not attend; task exclusions followed. Memory analysis N=42 and perception N=43 are inferred from reported ANOVA degrees of freedom. Memory monitoring improved relative to control: interaction F(1,40)=4.98, p<.05. Perceptual monitoring did not: F(1,41)=0.10, p=.75. No improvement in primary memory sensitivity was detected.
Small sample, unequal dropout (1 versus 5), instructor co-authorship and no verified prospective registration or independent direct replication limit inference. Intervals/data/code unverified. US education/NSF funding disclosed; conflict statement unverified. Scope: publisher PDF selected methods/results/discussion, not raw-data audit; copyright retained, link only. Relevance 5/5; provisional quality 3/5; moderate confidence. Added assessment: denominator reconstruction and a domain-specific comparator. The newer studies use different tasks/designs, so are not direct failed replications.
CONTEXTUAL REASSESSMENT · SAME PRIOR LEAD, NOT NEW DISCOVERY
The reported preregistered trial randomised 5,049 US/UK adults to brief mindfulness, self-compassion, gratitude or control. Negative affect improved modestly. Predictive modelling suggested differential responses; prospective algorithm-guided assignment remains untested.
Scope: primary indexed abstract/highlights, not full methods. Effect sizes/intervals, exclusions, blinding, registration timing, conflicts/funding and independent replication unverified. Data/code link uninspected; reuse unverified. Relevance 4/5; quality withheld pending methods. Added assessment: separates wellbeing prediction from proven personalisation, metacognitive accuracy and compassion. The DOI already appears in the corpus ledger.
What would deepen the BQs?
Connection
Proposed next test
Advancement gate
BQ002 · Thought and awareness
Compare spontaneous noticing with prompted confidence judgements across tasks.
Operational definitions, reliable measures and competing explanations; no metaphysical inference.
BQ006 · Intelligence and wisdom
Test whether improved calibration changes consequential choices.
Observed decisions and independent replication; confidence alone is insufficient.
BQ010 · Attention and flourishing
Compare participant-chosen, random and algorithm-assigned practices prospectively.
Held-out prediction plus randomised evaluation of benefit, harms and follow-up.
BQ011 · Interconnection into care
Measure repeated helping alongside experience, expectations and opportunities to act.
Demonstrate each proposed transition; do not infer societal or ecological benefit from mood.
