Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study
Open accessPublished version of a preprint
The DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index; 7 of the 36 sources its claims cite were checked for support.
Abstract
BACKGROUND: Voice assistants allow users to control appliances and functions of a smart home by simply uttering a few words. Such systems hold the potential to significantly help users with motor and cognitive disabilities who currently depend on their caregiver even for basic needs (eg, opening a door). The research on voice assistants is mainly dedicated to able-bodied users, and studies evaluating the accessibility of such systems are still sparse and fail to account for the participants' actual motor, linguistic, and cognitive abilities. OBJECTIVE: The aim of this work is to investigate whether cognitive and/or linguistic functions could predict user performance in operating an off-the-shelf voice assistant (Google Home). METHODS: A group of users with disabilities (n=16) was invited to a living laboratory and asked to interact with the system. Besides collecting data on their performance and experience with the system, their cognitive and linguistic skills were assessed using standardized inventories. The identification of predictors (cognitive and/or linguistic) capable of accounting for an efficient interaction with the voice assistant was investigated by performing multiple linear regression models. The best model was identified by adopting a selection strategy based on the Akaike information criterion (AIC). RESULTS: For users with disabilities, the effectiveness of interacting with a voice assistant is predicted by the Mini-Mental State Examination (MMSE) and the Robertson Dysarthria Profile (specifically, the ability to repeat sentences), as the best model shows (AIC=130.11). CONCLUSIONS: Users with motor, linguistic, and cognitive impairments can effectively interact with voice assistants, given specific levels of residual cognitive and linguistic skills. More specifically, our paper advances practical indicators to predict the level of accessibility of speech-based interactive systems. Finally, accessibility design guidelines are introduced based on the performance results observed in users with disabilities.
Overview
CoChat Summary
TL;DR
In a mixed-methods study with 16 users with motor, linguistic, and cognitive disabilities, performance with Google Home was about 58.5% accurate on average and was best predicted by MMSE and sentence repetition ability. The paper argues that commercial voice assistants can be usable for some disabled users if cognitive and linguistic abilities are sufficient, and it proposes design implications for accessibility.
Key findings
- Participant accuracy in interacting with the voice assistant was 58.5% (SD 18.6%) on average.
- The most frequent errors were phrasing errors (75/182, 41.2%) and timing errors (74/182, 40.7%); less frequent were comprehension errors (19/182, 10.4%) and pronunciation errors (14/182, 7.7%).
- In the final regression model, MMSE and repetition of sentences significantly predicted accuracy: MMSE (β=3.70, t13=3.26, P=.006) and repetition of sentences (β=22.06, t13=4.16, P=.001); the model had AIC=130.11 and R2=.77.
- Participants’ overall evaluation of the system was highly positive, with a mean score of 9.4 (SD 1.2), and all participants said they would like a voice assistant in their own home.
Methods
- Design: Two-phase mixed-methods study in a living laboratory: phase 1 involved group sessions where participants performed voice-assistant tasks; phase 2 involved neuropsychological and linguistic assessment. Sessions were video recorded and later coded.
- Sample/data: 16 participants with ascertained motor impairments and related language difficulties who needed daily caregiver assistance; mean age 38.3 years and mean education 11.8 years.
- Analysis: Video-coded task attempts were summarized into a performance index, then related to cognitive/linguistic measures using multiple linear regression with AIC-based model selection.
Limitations
- The paper states that the sample size was limited to 16 participants, and calls for larger and more heterogeneous samples.
- The paper states that the study examined a group situation in a living-lab setting; it says future work should test individual use to better examine person-device interaction.
- Evident from the text: the study used a single commercial voice assistant (Google Home) in a controlled setting, so generalization to other systems and real-world home environments is limited (evident from the setting described).
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 6 Not valid · 12 Not checked · 55 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“Indeed, the studies investigating user experience and usability of voice assistants mainly involved able-bodied users [3,8-11], thereby neglecting a broad community of users with disabilities.”
§ Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study > Introduction > Background
What the sources say
- Not checked — no full text of the sourcecites [3] source
- Disputedcites [8] sourceExperimental — accuracy not yet measured
The Methods and Participants section (page unavailable) says the paper's usability sessions involved people with motor impairments, not mainly able-bodied users. The Preliminary User Evaluation section likewise describes recruiting participants with motor impairments. This paper therefore runs contrary to its use as an example of the studies described in the citing sentence; it does not establish the broader pattern across other studies. (LLM, conf=0.86)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites [9] source
- Not checked — no full text of the sourcecites [10] source
- Not checked — no full text of the sourcecites [11] López, G, 2018, Advances in Human Factors and System Interactions source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“These abilities involve memory functions, specifically long-term memory and working memory, both crucial when interacting with voice interfaces [25].”
§ Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study > Introduction > Speech and Cognitive Factors Accounting for User Performance
What the sources say
- Partly supportedcites [25] sourceExperimental — accuracy not yet measured
Section 2.2.1 (p. 3) explains why auditory options place demands on working memory, supporting its importance in voice-interface use. Section 2.2.2 (p. 3) discusses forgetting information from earlier in a long dialogue, but does not identify long-term memory as a crucial function; that part of the claim is not established. (LLM, conf=0.86)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This was probably due to the lack of saliency of the feedback provided by Google Home after the waking command [35], which consists only of dim lights moving on top of the device.”
§ Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study > Discussion > Principal Findings
What the sources say
- Partly supportedcites [35] sourceExperimental — accuracy not yet measured
The paper’s barriers table identifies a lack of on-screen feedback and context, and suggests a companion screen to show status. It does not discuss feedback after the waking command, dim lights on the device, or timing errors caused by that feedback. (LLM, conf=0.76)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“As reported in a previous study, speech disfluency can represent an accessibility barrier to voice assistants.”
§ Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study > Introduction > Speech and Cognitive Factors Accounting for User Performance
What the sources say
- Partly supportedcites [23] sourceExperimental — accuracy not yet measured
The Abstract reports that pauses and hesitations in older adults’ speech were associated with cognitive scores during voice-system interactions. It also mentions differences in handling recognition errors, but does not establish that speech disfluency itself caused those errors or posed an accessibility barrier. (LLM, conf=0.76)
plainstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“For instance, long hesitations or pauses can be misinterpreted by the system as a sentence delimiter [23], causing an alteration of speech segmentation.”
§ Investigating the Accessibility of Voice Assistants With Impaired Users: Mixed Methods Study > Introduction > Speech and Cognitive Factors Accounting for User Performance
What the sources say
- Could not verifycites [23] sourceExperimental — accuracy not yet measured
The Abstract reports pauses and hesitations as vocal characteristics related to cognitive scores, but does not say that the system treats long pauses as sentence delimiters or changes speech segmentation. Searches for “sentence delimiter” and for pauses with speech recognition and segmentation found no passage addressing that mechanism in our copy. (LLM, conf=0.72)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
47 more claims · 2 with their cited sources checked
Create a free account to see all 77 claims and every source verdict.
- Probably not a substantial claim. 12
- It may state two separate claims as one. 17
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?
The cited source says what the claim attributes to it.
In the citing paper
“Second, it facilitates input for users with low literacy or with visual [24], intellectual [25], motor, linguistic, and cognitive disabilities [26] and can support more natural health routine tasks when in-person health care is not possible [19,27].”
From Voice-Based Conversational Agents for the Prevention and Management of Chronic and Mental Health Conditions: Systematic Literature Review (2021) · cites this paper as [26]
In the citing paper
“These studies suggest that voice assistants can be used as an assistive technology and increase participation in everyday life.”
From Use and acceptance of voice assistants among people with aphasia in Germany (2023) · cites this paper as Masina et al., 2020
In the citing paper
“This may also suggest a high potential of voice assistants as an assistive technology, since people with aphasia can make use of such system in spite and not because of the language impairment, which makes the technology less stigmatizing than other assistive technologies (Masina et al., 2020).”
From Use and acceptance of voice assistants among people with aphasia in Germany (2023) · cites this paper as Masina et al., 2020
Citations
Cites
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975) · Journal of Psychiatric Research · doi:10.1016/0022-3956(75)90026-6 · Cited by 88,282
Information Theory and an Extension of the Maximum Likelihood Principle
Akaike, H. (1998) · Springer Series in Statistics · doi:10.1007/978-1-4612-1694-0_15 · Cited by 17,781
The Mini‐Mental State Examination: A Comprehensive Review
Tombaugh, T. N., & McIntyre, N. J. (1992) · Journal of the American Geriatrics Society · doi:10.1111/j.1532-5415.1992.tb01992.x · Cited by 5,152 · cited 1×
“Users with severe cognitive impairment (MMSE score <18) [36] may still be able to use these systems effectively (performance index >50”
BackgroundThe FAB: A frontal assessment battery at bedside
Dubois, B., Slachevsky, A., Litvan, I., & Pillon, B. (2000) · Neurology · doi:10.1212/wnl.55.11.1621 · Cited by 3,913 · cited 2×
“More specifically, the neuropsychological functions were assessed with the Addenbrooke’s Cognitive Examination–Revised (ACE-R) [27] and the Frontal Assessment Battery (FAB) [28].”
BackgroundMethodAIC model selection using Akaike weights
Wagenmakers, E.-J., & Farrell, S. (2004) · Psychonomic Bulletin & Review · doi:10.3758/bf03206482 · Cited by 2,684
Mioshi, E., Dawson, K., Mitchell, J., Arnold, R., & Hodges, J. R. (2006) · International Journal of Geriatric Psychiatry · doi:10.1002/gps.1610 · Cited by 2,016 · cited 2×
“More specifically, the neuropsychological functions were assessed with the Addenbrooke’s Cognitive Examination–Revised (ACE-R) [27] and the Frontal Assessment Battery (FAB) [28].”
MethodAdvances in natural language processing
Hirschberg, J., & Manning, C. D. (2015) · Science · doi:10.1126/science.aaa8685 · Cited by 1,712 · cited 1×
“Although most voice assistants exploit machine learning algorithms to adapt to the user and increase their speech recognition accuracy over time [15,16], these systems are still designed and developed for people with clear and intelligible speech.”
BackgroundAlexa, Siri, Cortana, and more: An introduction to voice assistants.
Hoy, M. B. (2018) · Medical Reference Services Quarterly · doi:10.1080/02763869.2018.1404391 · Cited by 965 · cited 1×
“Although most voice assistants exploit machine learning algorithms to adapt to the user and increase their speech recognition accuracy over time [15,16], these systems are still designed and developed for people with clear and intelligible speech.”
BackgroundOlder people, assistive technologies, and the barriers to adoption: A systematic review
Yusif, S., Soar, J., & Hafeez-Baig, A. (2016) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2016.07.004 · Cited by 548
Pradhan, A., Mehta, K., & Findlater, L. (2018) · Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3173574.3174033 · Cited by 448
The TORGO database of acoustic and articulatory speech from speakers with dysarthria
Rudzicz, F., Namasivayam, A. K., & Wolff, T. (2012) · Language Resources and Evaluation · doi:10.1007/s10579-011-9145-0 · Cited by 425 · cited 1×
“More specifically, accuracy in speech processing was tested using sentences extracted from the TORGO database, which includes the recordings of 8 English speakers with dysarthria [18].”
MethodWiley StatsRef: Statistics Reference Online
(2015) · Choice Reviews Online · doi:10.5860/choice.190902 · Cited by 363
Bickmore, T. W., Trinh, H., Olafsson, S., O'Leary, T. K., Asadi, R., Rickles, N. M., & Cruz, R. (2018) · Journal of Medical Internet Research · doi:10.2196/11510 · Cited by 308 · cited 1×
“Voice-activated technologies are becoming pervasive in our everyday life [1,2].”
BackgroundTalk to me: Exploring user interactions with the Amazon Alexa
Lopatovska, I., Rink, K., Knight, I., Raines, K., Cosenza, K., Williams, H., Sorsche, P., Hirsch, D., Li, Q., & Martinez, A. (2019) · Journal of Librarianship and Information Science · doi:10.1177/0961000618759414 · Cited by 298 · cited 1×
“Indeed, the studies investigating user experience and usability of voice assistants mainly involved able-bodied users [3,8-11], thereby neglecting a broad community of users with disabilities.”
BackgroundLópez, G., Quesada, L., & Guerrero, L. A. (2018) · Advances in Intelligent Systems and Computing · doi:10.1007/978-3-319-60366-7_23 · Cited by 277
Executive Functions and the Frontal Lobes
(2010) · doi:10.4324/9780203837863 · Cited by 158
Corbett, E., & Weber, A. (2016) · Proceedings of the 18th International Conference on Human-Computer Interaction with Mobile Devices and Services · doi:10.1145/2935334.2935386 · Cited by 148
Older Adults and Voice Interaction: A Pilot Study with Google Home
Kowalski, J., Jaskulska, A., Skorupska, K., Abramczuk, K., Biele, C., Kopeć, W., & Marasek, K. (2019) · Extended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290607.3312973 · Cited by 125
Pyae, A., & Joelsson, T. N. (2018) · Proceedings of the 20th International Conference on Human-Computer Interaction with Mobile Devices and Services Adjunct · doi:10.1145/3236112.3236130 · Cited by 99 · cited 3×
“Indeed, the studies investigating user experience and usability of voice assistants mainly involved able-bodied users [3,8-11], thereby neglecting a broad community of users with disabilities.”
BackgroundVoice assistants and older people: some open issues
Sayago, S., Neves, B. B., & Cowan, B. R. (2019) · Proceedings of the 1st International Conference on Conversational User Interfaces · doi:10.1145/3342775.3342803 · Cited by 97
Rudzicz, F., Wang, R., Begum, M., & Mihailidis, A. (2015) · ACM Transactions on Accessible Computing · doi:10.1145/2744206 · Cited by 83 · cited 1×
“Furthermore, patients with Alzheimer disease experience difficulties interacting with a voice-controlled robot because of the timing imposed by the device [21].”
BackgroundBranham, S. M., & Mukkath Roy, A. R. (2019) · Proceedings of the 21st International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3308561.3353797 · Cited by 75
Kocaballi, A. B., Quiroz, J. C., Rezazadegan, D., Berkovsky, S., Magrabi, F., Coiera, E., & Laranjo, L. (2020) · Journal of Medical Internet Research · doi:10.2196/15823 · Cited by 73 · cited 1×
“Voice-activated technologies are becoming pervasive in our everyday life [1,2].”
BackgroundReducing working memory load in spoken dialogue systems
Wolters, M., Georgila, K., Moore, J. D., Logie, R. H., MacPherson, S. E., & Watson, M. (2009) · Interacting with Computers · doi:10.1016/j.intcom.2009.05.009 · Cited by 66 · cited 1×
“These abilities involve memory functions, specifically long-term memory and working memory, both crucial when interacting with voice interfaces [25].”
BackgroundOn the impact of dysarthric speech on contemporary ASR cloud platforms
Russis, L. D., & Corno, F. (2019) · Journal of Reliable Intelligent Environments · doi:10.1007/s40860-019-00085-y · Cited by 65
Manual and Automatic Transcriptions in Dementia Detection from Speech
Weiner, J., Engelbart, M., & Schultz, T. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-112 · Cited by 64 · cited 1×
“Cognitive skills were proven to affect the ability to operate a voice-controlled device by Weiner and colleagues [20].”
BackgroundUse of voice activated interfaces by people with intellectual disability
Balasuriya, S. S., Sitbon, L., Bayor, A. A., Hoogstrate, M., & Brereton, M. (2018) · Proceedings of the 30th Australian Conference on Computer-Human Interaction · doi:10.1145/3292147.3292161 · Cited by 63 · cited 2×
“Moreover, users must be able to correctly articulate words, especially multisyllable words (eg, temperature) or specific words that may require more effort to be articulated [24].”
BackgroundGoogle Home: smart speaker as environmental control unit
Noda, K. (2017) · Disability and Rehabilitation Assistive Technology · doi:10.1080/17483107.2017.1369589 · Cited by 52 · cited 1×
“One of the most prominent application domains is the domestic environment, where voice assistants, a branch of voice-activated technologies, allow the user to control and interact with several home appliances in a natural way by uttering voice commands [6,7].”
BackgroundKobayashi, M., Kosugi, A., Takagi, H., Nemoto, M., Nemoto, K., Arai, T., & Yamada, Y. (2019) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-29390-1_4 · Cited by 45 · cited 1×
“For instance, long hesitations or pauses can be misinterpreted by the system as a sentence delimiter [23], causing an alteration of speech segmentation.”
BackgroundUser interactions with “Alexa” in public academic space
Lopatovska, I., & Oropeza, H. (2018) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.2018.14505501034 · Cited by 41
Assessing Virtual Assistant Capabilities with Italian Dysarthric Speech
Ballati, F., Corno, F., & De Russis, L. (2018) · Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3234695.3236354 · Cited by 39 · cited 4×
“First, the actual speech abilities of the users were not assessed because they were either self-reported [7] or not reported at all [17,19].”
BackgroundResultInfluentialAdvances in Human Factors and System Interactions
Nunes, I. L. (2016) · Advances in intelligent systems and computing · doi:10.1007/978-3-319-41956-5 · Cited by 13
Accessibility and acceptability of voice assistants for people with acquired brain injury
Malapaschas, A. (2020) · ACM SIGACCESS Accessibility and Computing · doi:10.1145/3386402.3386406 · Cited by 4 · cited 1×
“Speech and cognitive skills play a significant role in the ability to effectively control voice assistants [17,19,22].”
Background
Research works printed but not linked: 5
- Duffy, J, 2013, Motor speech disorders e-book: substrates, differential diagnosis, and management
- Ballati, F, 2018, Intell Environ, 1
- Fussi, F, 1999, Profilo di valutazione della disartria: adattamento italiano del test di Robertson, raccolta di dati normativi e linee di trattamento
- Boersma, P, 2001, Glot Int, 5, 341
- Myers, R, 1990, Classical and Modern Regression With Applications, 2nd Edition
Cited by
Bérubé, C., Schachner, T., Keller, R., Fleisch, E., v Wangenheim, F., Barata, F., & Kowatsch, T. (2021) · Journal of Medical Internet Research · doi:10.2196/25933 · Cited by 133
“Second, it facilitates input for users with low literacy or with visual [24], intellectual [25], motor, linguistic, and cognitive disabilities [26] and can support more natural health routine tasks when in-person health care is not possible [19,27].”
Harrington, C. N., Garg, R., Woodward, A., & Williams, D. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3501995 · Cited by 98
“not remembering how to activate Google Home or communication breakdowns such as misinterpreted commands might be attributed to older adults’ lack of familiarity with newer devices such as voice assistants [18, 69, 82, 83].”
BackgroundTesla-Rapture: A Lightweight Gesture Recognition System From mmWave Radar Sparse Point Clouds
Salami, D., Hasibi, R., Palipana, S., Popovski, P., Michoel, T., & Sigg, S. (2022) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2022.3153717 · Cited by 98
Jovanovic, M., Mitrov, G., Zdravevski, E., Lameski, P., Colantonio, S., Kampel, M., Tellioğlu, H., & Flórez‐Revuelta, F. (2022) · Journal of Medical Internet Research · doi:10.2196/36553 · Cited by 106
“Off-the-shelf voice assistants were used by persons with motor, linguistic, and cognitive disabilities [155].”
The Impact of Voice Assistant Home Devices on People with Disabilities: A Longitudinal Study
Vieira, A. D., Leite, H., & Volochtchuk, A. V. L. (2021) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2022.121961 · Cited by 71
Metadata
- Authors
- Masina, F., Orso, V., Pluchino, P., Dainese, G., Volpato, S., Nelini, C., Mapelli, D., Spagnolli, A., & Gamberini, L.
- Affiliations
- University of Padua (IT)
- Corresponding author
- Gamberini, Luciano
- ORCID
- Venue
- Journal of Medical Internet Research
- Publisher
- JMIR Publications Inc.
- Published
- 2020-09-25
- Volume
- 22
- Issue
- 9
- Pages
- e18431
- ISSN
- 1438-8871 (ISSN-L)
- PMC
- PMC7547392
- OpenAlex
- W3045900875
- Cited by
- 105
- References
- 32
- Citations by year
- 2026: 14 · 2025: 18 · 2024: 26 · 2023: 15 · 2022: 19 · 2021: 13
- FWCI
- 20.90
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- Journal of Medical Internet Research (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Padua Research Archive (University of Padova) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- European Regional Development Fund (FESR 2014-2020, 2014-2020’, POR FESR 2014-2020)
- Topics
- Health Sciences › Health Professions › Occupational Therapy › Assistive Technology in Communication and Mobility
- Social Sciences › Human Factors and Ergonomics › Digital Accessibility for Disabilities
- Physical Sciences › Computer Science › Human-Computer Interaction › Usability and User Interface Design
- Keywords
- voice assistants, accessibility, disabled users, motor impairment, cognitive impairment, speech disorders, Google Home, home automation, speech-based interaction, Mini-Mental State Examination, living labs
- MeSH terms
- Adult, Female, Persons with Disabilities, Humans, Language, Male, Middle Aged, Voice, Young Adult, Cognitive Dysfunction
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
