CoChat
Sign Up
Cited by 22Obsessive-Compulsive Spectrum DisordersAnxiety, Depression, Psychometrics, Treatment, Cognitive ProcessesNeural and Behavioral Psychology Studies

Uncertainty increases neural indices of attention in obsessive-compulsive disorder

Open access
Dieterich, Raoul, Endrass, Tanja, Kathmann, Norbert·2017·Depression and Anxiety
DOI 10.1002/da.22655
Open this paper in CoChat — free accountOpen PDF

Abstract

BACKGROUND: Patients with obsessive-compulsive disorder (OCD) experience abnormally high levels of uncertainty, and unpredictability is evaluated negatively and not well tolerated. The current study examined neural correlates of attentional processing in response to experimentally induced uncertainty in OCD. METHODS: Twenty-four OCD patients and 24 healthy controls performed a task where neutral and negative pictures were preceded by a cue, either being predictive (certain condition) or nonpredictive (uncertain condition) of subsequent picture valence. We examined prepicture anticipatory attention through α (∼8-12 Hz) suppression, and attentional allocation during picture presentation with the P1, N1, P2, N2, and late positive potential (LPP) of the event-related potential. Additionally, we tested how clinical measures related to these attentional markers. RESULTS: Subjectively, patients overestimated the frequency of negative pictures after nonpredictive cues. Patients, but not controls, showed upper α(10-12 Hz) suppression after nonpredictive and predictive negative cues relative to predictive neutral cues. Only patients showed increased P2 and decreased N2 amplitudes for pictures after nonpredictive cues, and, whereas both groups showed increased LPP amplitudes for pictures after nonpredictive cues, this modulation was more pronounced in OCD during the early LPP (<1,000 ms). In patients, P2 and LPP amplitudes for negative pictures were associated positively with anxiety and negatively with depression. CONCLUSIONS: These results suggest that OCD patients process anticipation of inevitable and potential threat similarly and highlight the substantial motivational impact of uncertain events to OCD patients. Finally, the correlation with anxiety implies that anxiety represents the source of hypervigilance during uncertainty resolution.

AI snapshot

Written from the paper’s full text.

TL;DR

In a cued-picture EEG/ERP study, OCD patients showed stronger anticipatory upper alpha suppression for both nonpredictive and predictive negative cues relative to predictive neutral cues, unlike healthy controls. Patients also showed OCD-specific increased P2 and decreased N2 for pictures after nonpredictive cues and a more pronounced early LPP predictability effect, with anxiety positively and depression negatively associated with these responses.

Key findings

  • Patients overestimated the frequency of negative pictures after nonpredictive cues: postexperiment estimates were 57.0% in OCD vs. 48.4% in controls, t(46) = 2.40, P = .021, d = 0.69; groups did not differ in trial-by-trial expectancy or subjective valence ratings.
  • For anticipatory alpha, the three-way alpha band × cue type × group interaction was significant, F(2, 92) = 3.69, P = .029, η²p = .07. In patients only, upper alpha power was smaller after nonpredictive and predictive negative cues compared with predictive neutral cues, F(1, 46) = 5.79, P = .020, η²p = .11, with no difference between nonpredictive and predictive negative cues.
  • For ERPs, P2 amplitudes were more positive and N2 amplitudes less negative for unpredictable vs. predictable pictures in patients (P2: t(23) = 3.11, P < .001, d = 0.69; N2: t(23) = 4.00, P < .001, d = 0.81), but not in controls (P2: t(23) = 0.55, P = .660, d = 0.07; N2: t(23) = 1.56, P = .132, d = 0.30). The early LPP predictability effect was larger in patients, t(23) = 5.15, P < .001, d = 1.06, than controls, t(23) = 3.10, P = .011, d = 0.66.
  • In patients, ΔP2 negative was associated positively with trait anxiety (STAI: B = 0.28, β = 1.22, t = 3.04) and negatively with depression (BDI-II: B = −0.46, β = −1.35, t = −3.64); posterior early ΔLPP negative was associated negatively with clinician-rated depression (MADRS: B = −0.41, β = −0.81, t = −2.69) and positively with trait anxiety at trend level.

Methods

  • Twenty-four OCD patients and 24 healthy controls (17 females/7 males per group; mean age 27.9 vs. 27.7 years) completed a cued-picture paradigm with 54 negative and 54 neutral IAPS pictures. Cues were predictive neutral (“O”), predictive negative (“X”), or nonpredictive (“?”); the task had 108 trials in six blocks of 18.
  • EEG was recorded from 60 Ag/AgCl electrodes at 250 Hz. Anticipatory attention was indexed by lower and upper alpha power modulation at PO1/PO2; picture processing was indexed by P1, N1, P2, N2, early LPP (400–1,000 ms), and late LPP (1,000–2,000 ms). Source localization used the CLARA algorithm in BESA.
  • Main analyses were three-way ANOVAs for ratings, alpha power modulation, and ERPs, plus multiple linear regressions within patients relating symptom measures (OCI-R, BDI-II, IUS, STAI; Y-BOCS, MADRS) to uncertainty-related ERP and alpha difference scores.

Limitations

  • Stated by the authors: comorbid diagnoses primarily encompassed intolerance-of-uncertainty-related internalizing disorders, limiting specificity; SSRIs may affect electrophysiological correlates of attention and emotion, making medication status a potential confound; alpha power modulation tended to be positive, so it did not always reflect actual suppression of prepicture relative to precue activity; the LPP is responsive to arousal, and future research should include positive pictures; symptom-ERP associations were lacking for neutral pictures.
  • Evident from the text: the sample was modest (24 patients, 24 controls), predominantly female, and included patients with comorbid diagnoses and some taking SSRIs, which may limit generalizability and specificity.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1002/da.22655

Cites

43 of 43 references verified0 citation errors · 0 not checkable

Statistics

24 tests recomputed · 0 with the reported and recomputed p on opposite sides of .05

Claims the paper makes

0 of 25 cited sources checked · 0 supported · 0 flagged

  • Findingnone given

    Anxiety rather than OC symptoms accompanied increased uncertainty-related attention to negative pictures.

    “Anxiety rather than OC symptoms accompanied increased uncertainty-related attention to negative pictures.”
    Quote in the paperSubstantial claim
  • Findingown data

    In OCD patients, upper alpha (10-12 Hz) power was smaller after nonpredictive and predictive negative cues compared to predictive neutral cues.

    “Patients, but not controls, showed upper 𝛼(10-12 Hz) suppression after nonpredictive and predictive negative cues relative to predictive neutral cues.”
    Quote in the paperSubstantial claim
  • Findingown data

    The early LPP predictability effect (unpredictable vs predictable) was larger in OCD patients than in healthy controls.

    “this modulation was more pronounced in OCD during the early LPP (<1,000 ms).”
    Quote in the paperSubstantial claim
  • Findingown data

    In OCD patients, postexperiment estimates of the proportion of negative pictures after nonpredictive cues were higher than in healthy controls.

    “Subjectively, patients overestimated the frequency of negative pictures after nonpredictive cues.”
    • Also stated in: § 3.1 Ratings · plain
      “OCD patients had higher postexperiment estimates of the proportion of negative pictures than controls, but groups did not differ in trial-by-trial expectancy and subjective valence ratings (see Table 1).”
    Quote in the paperNot a substantial claim
  • Findingown data

    Temporal, ventral prefrontal, medial frontal cortices, and cerebellum consistently contributed to uncertainty-related group differences.

    “Temporal, ventral prefrontal, medial frontal cortices, and cerebellum consistently contributed to uncertainty-related group differences, as detailed in Figure 4 and Table 4.”
    Quote in the paperSubstantial claim
  • May combine two claims. 9
  • Classified as probably not a substantial claim. 11

  • May combine two claims.

    Findingown data

    APM tended to be positive, thus not always reflecting actual suppression of prepicture relative to precue activity.

    “Furthermore, APM tended to be positive, thus not always reflecting actual suppression of prepicture relative to precue activity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Further sources included parietal (LPP) and occipital (P2, N2) cortices, cingulate (LPP, N2), parahippocampal gyrus/amygdala (N2), insula, and brainstem (P2).

    “Further sources included parietal (LPP) and occipital (P2, N2) cortices, cingulate (LPP, N2), parahippocampal gyrus/amygdala (N2), insula, and brainstem (P2).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    There was no significant group effect or interaction for SPN.

    “There was neither a significant group effect, F(2, 92) = 1.41, p = .242,nor an interaction, F < 1.00, p = .955.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The effects of predictability on P2 and N2 amplitudes were present in patients but not in controls.

    “Notably, as revealed by significant group × predictability interactions, these effects were present in patients (P2: t(23) = 3.11, P < .001,d = 0.69; N2: t(23) = 4.00, P < .001,d = 0.81), but not in controls (P2: t(23) = 0.55, P = .660,d = 0.07; N2: t(23) = 1.56,P = .132,d = 0.30).”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Alpha suppression appears sensitive to cues indicating threat, with stronger suppression while negative versus positive pictures or painful versus nonpainful stimuli were anticipated.

    “Alpha suppression presumably reflects preparatory activity in the visual cortex (Van Dijk, Schoffelen, Oostenveld, & Jensen, 2008) and appears sensitive to cues indicating threat, with stronger suppression while negative versus positive pictures (Onoda et al., 2007) or painful versus nonpainful stimuli (Babiloni et al., 2003) were anticipated.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Onoda et al., 2007 source

      Not checked in the free run.

    • cites Babiloni et al., 2003 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    Sources in the insula and amygdala corroborate imaging results, specifically uncertainty-related amygdalar hyperactivity in pediatric anxiety.

    “Particularly, sources in the insula and amygdala, known to process subjective salience (Davis, & Whalen, 2001;Straube, & Miltner, 2011), corroborate imaging results, specifically uncertainty-related amygdalar hyperactivity in pediatric anxiety (Sarinopoulos et al., 2010;Williams et al., 2015).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Sarinopoulos et al., 2010 source

      Not checked in the free run.

    • cites Williams et al., 2015 source

      Not checked in the free run.

    • cites Davis, & Whalen, 2001 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    The LPP is amplified by symptom provocation in OCD.

    “Since the LPP is amplified by symptom provocation in OCD (Paul, Simon, Endrass, & Kathmann, 2016), we excluded pictures from data analyses, if they were rated symptom relevant (rating > 1).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Paul, Simon, Endrass, & Kathmann, 2016

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    The early LPP increases in response to cues for threat of shock and during resolution of uncertainty, and decreases to signify uncertainty-related avoidance.

    “The early LPP increases in response to cues for threat of shock (Bublatzky, & Schupp, 2012) and during resolution of uncertainty (Dieterich, Endrass, & Kathmann, 2016), and decreases to signify uncertainty-related avoidance (Gole, Schäfer, & Schienle, 2012).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Bublatzky, & Schupp, 2012 source

      Not checked in the free run.

    • cites Dieterich, Endrass, & Kathmann, 2016 source

      Not checked in the free run.

    • cites Gole, Schäfer, & Schienle, 2012

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Lower alpha (8-10 Hz) may be more specifically involved in anticipation than upper alpha.

    “Previous findings suggest differentiating lower (8-10 Hz) and upper (10-12 Hz) 𝛼, as lower 𝛼 may be more specifically involved in anticipation (Onoda et al., 2007).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Onoda et al., 2007 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    Motivationally significant stimuli elicit blunted LPPs in depression.

    “motivationally significant stimuli elicit blunted LPPs in depression (Weinberg, Perlman, Kotov, & Hajcak, 2016).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Weinberg, Perlman, Kotov, & Hajcak, 2016

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    SSRIs affect electrophysiological correlates of attention and emotion.

    “SSRIs have been shown to affect electrophysiological correlates of attention and emotion (Kemp, Gray, Silberstein, Armstrong, & Nathan, 2004)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kemp, Gray, Silberstein, Armstrong, & Nathan, 2004

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Early (<1000 ms) and late (>1000 ms) portions of the LPP may reflect partially distinct attentional processes, obligatory attentional allocation and deliberate elaboration, respectively.

    “Its early (<1,000 ms) and late (>1,000 ms) portions may reflect partially distinct attentional processes, that is, obligatory attentional allocation (Weinberg, & Hajcak, 2010) and deliberate elaboration (Foti, & Hajcak, 2008), respectively.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Weinberg, & Hajcak, 2010 source

      Not checked in the free run.

    • cites Foti, & Hajcak, 2008

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The LPP is a broad, parietally distributed sustained positivity starting around 300 ms after stimulus onset.

    “The late positive potential (LPP) is a broad, parietally distributed sustained positivity starting around 300 ms after stimulus onset (Schupp et al., 2000).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schupp et al., 2000 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Anxiety and depression involve high sensitivity to uncertainty.

    “Apart from OCD, anxiety and depression involve high sensitivity to uncertainty (Mahoney, & McEvoy, 2012)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mahoney, & McEvoy, 2012 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The positive APM suggests considerable anticipation of the cues.

    “However, this suggests considerable anticipation of the cues, which is expected given that cues were highly informative and required a rating response.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    The LPP is responsive to arousal.

    “Lastly, since the LPP is responsive to arousal (Schupp et al., 2000),”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Schupp et al., 2000 source

      Not checked in the free run.

16 of 16

  • The paper relies on this without support.

    Uncertainty affects processing during anticipation, when expecting an unknown outcome, and resolution, when outcomes become available for evaluation.

    “Uncertainty affects processing during anticipation, when expecting an unknown outcome, and resolution, when outcomes become available for evaluation.”

  • The paper relies on this without support.

    Presenting cues predictive or nonpredictive of the valence of subsequent pictures allows examining anticipation and resolution of uncertainty, before and after picture onset, respectively.

    “Presenting cues predictive or nonpredictive of the valence of subsequent pictures allows examining anticipation and resolution of uncertainty, before and after picture onset, respectively.”

  • The paper relies on this without support.

    Anticipatory attention is reflected in suppression of occipitoparietal alpha power (8-12 Hz) relative to a less engaging reference period.

    “Anticipatory attention is reflected in suppression of occipitoparietal 𝛼 power (8-12 Hz) relative to a less engaging reference period.”
3 of 3

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

  • Inconsistent p-value: t(23) = 0.55, p = .660

    Minor

    The paper reports p = .66 here; recomputing from the test statistic and its degrees of freedom gives p = .588 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .66 · recomputed p = .588

  • Inconsistent p-value: F(2, 92) = 1.41, p = .242

    Minor

    The paper reports p = .242 here; recomputing from the test statistic and its degrees of freedom gives p = .249. The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .242 · recomputed p = .249

  • Inconsistent p-value: t(23) = 3.11, p < .001

    Minor

    The paper reports p < .001 here; recomputing from the test statistic and its degrees of freedom gives p = .005 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p < .001 · recomputed p = .005

3 of 3

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Doesn’t add upt(23) = 3.11reported p < .001 · recomputed p = .005

    “t(23) = 3.11, p < .001”

  • Doesn’t add upt(23) = 0.55reported p = .66 · recomputed p = .588

    “t(23) = 0.55, p = .660”

  • Doesn’t add upF(2, 92) = 1.41reported p = .242 · recomputed p = .249

    “F(2, 92) = 1.41, p = .242”

  • ConsistentF(1, 46) = 199.89reported p < .001 · recomputed p < .001

    “F(1, 46) = 199.89,p < .001”

  • Consistentt(46) = 0.839reported p = .406 · recomputed p = .406

    “t(46) = 0.839, p = .406”

  • ConsistentF(2, 92) = 2.54reported p = .085 · recomputed p = .084

    “F(2, 92) = 2.54, p = .085”

  • ConsistentF(2, 92) = 3.69reported p = .029 · recomputed p = .029

    “F(2, 92) = 3.69, p = .029”

  • ConsistentF(2, 92) = 0.52reported p = .599 · recomputed p = .596

    “F(2, 92) = 0.52, p = .599”

  • ConsistentF(2, 92) = 3.21reported p = .045 · recomputed p = .045

    “F(2, 92) = 3.21, p = .045”

  • ConsistentF(1, 46) = 5.79reported p = .02 · recomputed p = .020

    “F(1, 46) = 5.79, p = .020”

  • ConsistentF(1, 46) = 0.49reported p = .49 · recomputed p = .487

    “F(1, 46) = 0.49, p = .490”

  • Consistentt(23) = 5.15reported p < .001 · recomputed p < .001

    “t(23) = 5.15, p < .001”

  • Consistentt(23) = 3.1reported p = .011 · recomputed p = .005

    “t(23) = 3.10, p = .011”

    The paper states a correction for multiple tests

  • Consistentt(46) = 3.37reported p = .002 · recomputed p = .002

    “t(46) = 3.37, p = .002”

  • Consistentt(46) = 2.67reported p = .011 · recomputed p = .010

    “t(46) = 2.67, p = .011”

  • Consistentt(47) = 10.08reported p < .001 · recomputed p < .001

    “t(47) = 10.08,p < .001”

  • Consistentt(47) = 6.27reported p < .001 · recomputed p < .001

    “t(47) = 6.27,p < .001”

  • Consistentt(47) = 10.69reported p < .001 · recomputed p < .001

    “t(47) = 10.69,p < .001”

  • Consistentt(47) = 8.12reported p < .001 · recomputed p < .001

    “t(47) = 8.12, p < .001”

  • Consistentt(23) = 4reported p < .001 · recomputed p < .001

    “t(23) = 4.00, p < .001”

  • Consistentt(23) = 1.56reported p = .132 · recomputed p = .132

    “t(23) = 1.56,p = .132”

  • ConsistentF(2, 92) = 6.95reported p = .002 · recomputed p = .002

    “F(2, 92) = 6.95, p = .002”

  • Consistentt(47) = 2.99reported p = .015 · recomputed p = .004

    “t(47) = 2.99, p = .015”

    The paper states a correction for multiple tests

  • Consistentt(47) = 3.66reported p = .002 · recomputed p < .001

    “t(47) = 3.66, p = .002”

    The paper states a correction for multiple tests

24 of 24

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Y-BOCS coefficient has the opposite sign from its t value

    Minor

    In the second posterior early ΔLPP model, the reported coefficient is negative and its standard error is positive, so their ratio is approximately −1.62, not the reported positive 1.61.

    Reported 1.61 · recomputed -1.615385

  • Regression coefficient and standard error do not match the reported t value

    Minor

    For the first posterior early ΔLPP model, dividing the reported OCI-R coefficient, −0.4, by its standard error, 0.08, gives −5, not the reported −0.49.

    Reported -0.49 · recomputed -5

  • Endnote lists 53 aversive pictures rather than 54

    Minor

    The methods state that 54 negative pictures were selected, but the endnote's aversive-picture list contains 53 identifiers.

3 of 3

Cites

Flagged first, then by citations

1 of the 44 references couldn’t be matched to a known paper.

Research works printed but not linked: 5

  1. Hautzinger, 2006, Beck depressions-inventar (BDI-II). Revision
  2. Lang, 2008, International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Technical Report A-8
  3. Laux, 1981, Das State-Trait Angstinventar: STAI [The State-Trait Anxiety Inventory: STAI]
  4. Neumann, 1989, MADR-Skala zur psychometrischen Beurteilung depressiver Symptome (MADRS) [MADR-Scale for the psychometric assessment of depressive symptoms]
  5. Wittchen, 1997, Strukturiertes klinisches interview für DSM-IV [Structured clinical interview for DSM-IV]

Cited by

Sorted by citations

Related

Authors
Dieterich, R., Endrass, T., & Kathmann, N.
Affiliations
Humboldt-Universität zu Berlin (DE)
Corresponding author
Dieterich, Raoul
ORCID
Venue
Depression and Anxiety
Publisher
Wiley
Published
2017-05-25
Volume
34
Issue
11
Pages
1018-1028
ISSN
1091-4269 (ISSN-L)
DOI
10.1002/da.22655
OpenAlex
W2618156457
Semantic Scholar
faf3653746015eac25aafd691f4a4d09efd259db
Cited by
22
References
44
Citations by year
2026: 3 · 2025: 2 · 2024: 2 · 2023: 1 · 2022: 2 · 2021: 3 · 2020: 2 · 2019: 3 · 2018: 4
FWCI
3.04
Type
journal article
Language
en
Open access
Licence
Locations
  • Depression and Anxiety (publishedVersion)
  • PubMed (publishedVersion)
Funders
Evangelisches Studienwerk Villigst
Topics
  • Social Sciences › Psychology › Clinical Psychology › Obsessive-Compulsive Spectrum Disorders
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Anxiety, Depression, Psychometrics, Treatment, Cognitive Processes
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
Keywords
obsessive-compulsive disorder, uncertainty, anticipatory attention, alpha suppression, event-related potentials, P2, N2, late positive potential, threat anticipation, anxiety, depression, hypervigilance
MeSH terms
Adult, Attention, Cerebral Cortex, Cues, Electroencephalography, Emotions, Evoked Potentials, Female, Humans, Male, Obsessive-Compulsive Disorder, Pattern Recognition, Visual, Reference Values, Uncertainty, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review