Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study
Open accessPublished version of a preprint
The DOI matches its Crossref record; 43 of the 43 references it cites were found in a registry or an index; 50 of the 67 sources its claims cite were checked for support.
Abstract
Background: University students are increasingly reporting common mental health problems, such as stress, anxiety, and depression, and they frequently face barriers to seeking psychological support because of stigma, cost, and availability of mental health services. This issue is even more critical in the challenging time of the COVID-19 pandemic. Digital mental health interventions, such as those delivered via chatbots on mobile devices, offer the potential to achieve scalability of healthy-coping interventions by lowering cost and supporting prevention. Objective: The goal of this study was to conduct a proof-of-concept evaluation measuring the engagement and effectiveness of Atena, a psychoeducational chatbot supporting healthy coping with stress and anxiety, among a population of university students. Methods: In a proof-of-concept study, 71 university students were recruited during the COVID-19 pandemic; 68% (48/71) were female, they were all in their first year of university, and their mean age was 20.6 years (SD 2.4). Enrolled students were asked to use the Atena psychoeducational chatbot for 4 weeks (eight sessions; two per week), which provided healthy-coping strategies based on cognitive behavioral therapy, positive psychology, and mindfulness techniques. The intervention program consisted of conversations combined with audiovisual clips delivered via the Atena chatbot. Participants were asked to complete web-based versions of the 7-item Generalized Anxiety Disorder scale (GAD-7), the 10-item Perceived Stress Scale (PSS-10), and the Five-Facet Mindfulness Questionnaire (FFMQ) at baseline and postintervention to assess effectiveness. They were also asked to complete the User Engagement Scale-Short Form at week 2 to assess engagement with the chatbot and to provide qualitative comments on their overall experience with Atena postintervention. Results: Participants engaged with the Atena chatbot an average of 78 (SD 24.8) times over the study period. A total of 61 out of 71 (86%) participants completed the first 2 weeks of the intervention and provided data on engagement (10/71, 14% attrition). A total of 41 participants out of 71 (58%) completed the full intervention and the postintervention questionnaires (30/71, 42% attrition). Results from the completer analysis showed a significant decrease in anxiety symptoms for participants in more extreme GAD-7 score ranges (t39=0.94; P=.009) and a decrease in stress symptoms as measured by the PSS-10 (t39=2.00; P=.05) for all participants postintervention. Participants also improved significantly in the describing and nonjudging facets, based on their FFMQ subscale scores, and asked for some improvements in the user experience with the chatbot. Conclusions: This study shows the benefit of deploying a digital healthy-coping intervention via a chatbot to support university students experiencing higher levels of distress. While findings collected during the COVID-19 pandemic show promise, further research is required to confirm conclusions. © 2021 JMIR Publications. All rights reserved.
Overview
CoChat Summary
TL;DR
This proof-of-concept mixed-methods study found that a 4-week psychoeducational chatbot (Atena) was engaging for many university students and was associated with reductions in anxiety and stress, especially among students with more severe baseline symptoms. The authors conclude the results are promising but preliminary, with no control group and limited generalizability.
Key findings
- The study recruited 71 university students; 68% (48/71) were female, all were in their first year, and mean age was 20.6 years (SD 2.4).
- Engagement was moderate: participants interacted with Atena an average of 78 times (SD 24.8) over 4 weeks; 86% (61/71) completed the week-2 engagement measure, but only 58% (41/71) completed postintervention questionnaires.
- In completer analyses, anxiety symptoms decreased significantly for participants in the more extreme GAD-7 range (t39=0.94; P=.009), and stress decreased for all participants on the PSS-10 (t39=2.00; P=.05).
- Mindfulness outcomes improved: the describing and nonjudging facets of the FFMQ increased significantly, and participants’ feedback highlighted the value of videos and mindfulness content, while also requesting UX improvements.
Methods
- Mixed-methods proof-of-concept study aligned with ORBIT Phase I and Phase IIa; intervention tested over 4 weeks with 8 sessions (2 per week), each about 10 minutes.
- Sample: 71 first-year university students from the University of Trento, Italy, recruited during the COVID-19 pandemic; all were Italian speakers in northeastern Italy and used Telegram-accessed Atena voluntarily.
- Main analysis: descriptive statistics, Shapiro tests for normality, paired-samples t tests for pre/post GAD-7, PSS-10, and FFMQ; engagement measured with the UES-SF at week 2; qualitative comments analyzed thematically.
Limitations
- The paper states that findings are limited by being a preliminary proof-of-concept, involving a homogeneous university-student sample and no control group, which limits generalizability.
- The authors note that because log data were deidentified, it was difficult to assess how deeply users engaged with each session’s content; they say more objective behavioral monitoring would be preferable.
- Evident from the text: there was substantial attrition (42% at postintervention), and the intervention was run during the second wave of COVID-19 in Italy, which may have affected participants’ mental well-being and the intervention’s observed effects.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures



Verification
Claims the paper makes
19 Valid · 2 Not valid · 6 Not checked · 68 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The integration of human coaching and support in digital mental health interventions can help improve adherence and behavior change outcomes [17,18]”
§ Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study > Introduction
What the sources say
- Partly supportedcites [17] sourceExperimental — accuracy not yet measured
In “How important is guidance in guided ICBT?” (page not provided), the review defines support as human contact and reports that greater support is associated with larger treatment effects. The “Key issues” section (page not provided) also reports an association between therapist behaviors and module completion, which bears on adherence. The paper does not establish improved behavior-change outcomes as distinct from treatment effects, and its evidence concerns internet-based psychological treatment rather than digital mental health interventions generally. (LLM, conf=0.84)
- Partly supportedcites [18] sourceExperimental — accuracy not yet measured
Section 3.2 reports that guided Internet-based mental health interventions had higher completion rates, more completed modules, and lower symptom severity than unguided interventions. This supports improved adherence and mental health outcomes, but the review does not establish an improvement in behavior-change outcomes as a distinct measure. (LLM, conf=0.84)
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 support only part of this claim — read each source’s verdict below.
In the paper
“as well as their effectiveness in reducing perceived stress [16-25]”
§ Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study > Introduction
What the sources say
- Partly supportedcites [16] sourceExperimental — accuracy not yet measured
The Results (Completer analysis; page unavailable) report significantly lower perceived stress among participants who adhered to the chatbot intervention. But the Results (Outcome measures; page unavailable) found no significant effect on perceived stress in the full sample. The unqualified effectiveness claim therefore omits an important limitation of this pilot study. (LLM, conf=0.88)
- Could not verifycites [17] sourceExperimental — accuracy not yet measured
The present review discusses controlled trials of internet-delivered treatments for depression, not chatbot studies. Searches for chatbots and perceived-stress reduction found no passage in our copy addressing the attributed claim. (LLM, conf=0.76)
- Could not verifycites [18] sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) mentions synchronous chat as a mode of human support, not chatbots. The communication-mode results concern symptom severity in a social-phobia intervention. Searches for chatbots, stress management, and perceived-stress outcomes found no passage reporting that chatbots reduce perceived stress. (LLM, conf=0.76)
- Partly supportedcites [19] sourceExperimental — accuracy not yet measured
The Methods (Measures; page not provided) identify K10 as a measure of psychological distress, and the Results (Table 1; page not provided) show lower K10 scores after the conversational-agent intervention. The paper does not report a perceived-stress measure, so it supports a related reduction in distress rather than the specific claim about perceived stress. (LLM, conf=0.84)
- Not checked — no full text of the sourcecites [20] source
- Partly supportedcites [21] sourceExperimental — accuracy not yet measured
The Results > Psychological Outcomes section (page not provided) reports that five controlled studies found reductions in psychological distress compared with inactive controls. It does not establish reductions in perceived stress specifically, and it also reports no significant advantage in several studies with active controls. (LLM, conf=0.86)
- Partly supportedcites [22] sourceExperimental — accuracy not yet measured
The Discussion (Principal Findings; page not provided) says recent studies of chatbot- or videogame-based interventions report reduced stress. It does not specifically establish that chatbots reduce perceived stress, and the Conclusions say this pilot intervention still needs an efficacy evaluation. (LLM, conf=0.84)
- Could not verifycites [23] sourceExperimental — accuracy not yet measured
The Discussion’s Principal Findings report reduced anxiety and depression symptoms among students using Tess, but not reduced perceived stress. The Methods’ Measures section lists depression, anxiety, and positive and negative affect measures; a search for “perceived stress” found no assessment or finding for that outcome. (LLM, conf=0.76)
- Partly supportedcites [24] sourceExperimental — accuracy not yet measured
The Discussion’s Comparison With Prior Work reports chatbot studies with effects on depression or psychological well-being; one used the Perceived Stress Scale. It does not state that Wysa reduced perceived stress or specify a stress-reduction result from the comparison study. No page number is available in our copy. (LLM, conf=0.76)
- Supportedcites [25] source
The Introduction says recent studies found that engaging with a conversational agent can reduce perceived stress. The Results (Symptom Change) also report a substantial reduction in stress scores among users of this paper’s chatbot, although the authors describe their own findings as preliminary. (LLM, conf=0.90)
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.
- Partly supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“The baseline levels of stress and anxiety among our participants were significantly higher than those found in the same population of university students by previous research [42-44], as well as in the general population [45].”
§ Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study > Discussion > Principal Findings
What the sources say
- Not checked yetcites [42] doi:10.4172/2167-1044.1000330
- Supportedcites [43] source
The Methods describe a study of 192 Saudi university male students, and the Results (Table 1) report both anxiety (GAD-7) and perceived-stress (PSS) scores. The citation supports the previous-research values used for the citing authors’ comparison; the comparison with their own participants is the citing authors’ conclusion. (LLM, conf=0.88)
- Supportedcites [44] source
The Results (Participants’ Characteristics; page not provided) report mean perceived-stress and anxiety scores for university students, so this paper supplies a previous-study benchmark for the citing authors’ comparison. The assertion that the citing authors’ participants scored significantly higher is their own comparison; the general-population comparison is attached to a different citation [45]. (LLM, conf=0.88)
- Supportedcites [45] source
Citation [45] supports the general-population comparison, not the separate comparison with university-student studies [42–44]. The Methodology (Study design and study population; page unavailable) says the survey recruited the general public, and the Results (page unavailable) report stress and anxiety scores for its 4,479 respondents. The claim that the citing authors’ baseline was significantly higher is their comparison, not a finding attributed to this source. (LLM, conf=0.84)
plain- 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.
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Previous studies analyzing the mental health of university students found lower levels of anxiety and stress symptoms, evaluated through the GAD-7 and the PSS-10 questionnaires, compared to those of our sample of university students [42-44].”
§ Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study > Discussion > Principal Findings
What the sources say
- Not checked yetcites [42] doi:10.4172/2167-1044.1000330
- Supportedcites [43] source
The Results, ‘Participants’ Characteristics’ (Table 1; page not provided), report mean GAD-7 and PSS scores for Saudi university students. The Methods, ‘Measures’ (page not provided), identify these as the anxiety and 10-item perceived-stress questionnaires. The comparison with the citing authors’ own sample is their comparison, not a finding attributed to this source. (LLM, conf=0.88)
- Supportedcites [44] source
The Results (Participants’ Characteristics) report mean PSS-10 and GAD-7 scores of 16.16 and 5.27 among Saudi university students. The Methods (Participants and Study Design) specify that the participants were healthy male students. The comparison with the citing authors’ sample is the citing authors’ own comparison, not a finding attributed to this source. (LLM, conf=0.88)
plain- 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.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Our results are in line with recent studies targeting the same population, showing that online stress management interventions are more effective for students with higher levels of stress, anxiety, and depression [40,41].”
§ Engagement and Effectiveness of a Healthy-Coping Intervention via Chatbot for University Students During the COVID-19 Pandemic: Mixed Methods Proof-of-Concept Study > Discussion > Principal Findings
What the sources say
- Not checked — no full text of the sourcecites [40] source
- Partly supportedcites [41] sourceExperimental — accuracy not yet measured
The Results and Clinical scenario sections report that patients with more severe depression at baseline had greater benefit from low-intensity interventions, though the additional benefit was modest. The Inclusion criteria describe community and primary-care patients, not specifically students, and the paper does not establish the same finding for higher stress or anxiety or specifically for online stress-management interventions. Its comparison of internet-based and written interventions found no statistically significant difference in the severity effect. (LLM, conf=0.89)
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.
25 more claims
Create a free account to see all 95 claims and to verify the 3 references still open.
- Probably not a substantial claim. 39
- It may state two separate claims as one. 35
- The quote only partly states this claim. 1
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
“Almost all of interventions were digital, with the majority being delivered through a web-based mobile health app-based platform [45-47,49-56,58-68,70-73,76,79-82].”
The cited source says what the claim attributes to it.
In the citing paper
“Several studies were of mindfulness-based interventions or included mindfulness components [43,44,46,50-52,56,57,60,66-69].”
The cited source says what the claim attributes to it.
In the citing paper
“the 4-week pretest-posttest study by Gabrielli et al [52] (N=71) combined mindfulness with behavioral coping strategies that assisted with emotional regulation.”
The cited source says what the claim attributes to it.
In the citing paper
“Of the studies that demonstrated effectiveness, one was a multicomponent positive psychological intervention [53], several were mindfulness-based interventions [43,52,64,66,70,72,82] as well as CBT based [54,55,58,59,63,64,73,74,77].”
The cited source says what the claim attributes to it.
In the citing paper
“For example, only the chatbot study showed significant effects compared with the web-based mindfulness one [52,62].”
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Using thematic analysis in psychology
Braun, V., & Clarke, V. (2006) · Qualitative Research in Psychology · doi:10.1191/1478088706qp063oa · Cited by 194,414
A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7
Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Löwe, B. (2006) · Archives of Internal Medicine · doi:10.1001/archinte.166.10.1092 · Cited by 31,765 · cited 1×
“The 7-item Generalized Anxiety Disorder scale (GAD-7) [34] is a 7-item self-report scale based on a 4-point Likert scale, ranging from 0 (not at all) to 3 (nearly every day).”
MethodUsing Self-Report Assessment Methods to Explore Facets of Mindfulness
Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L. (2006) · Assessment · doi:10.1177/1073191105283504 · Cited by 7,511 · cited 11×
“Moreover, the mean scores of the subscales describing and nonjudging, as well as the mean total score of the FFMQ, showed significant increases (P .05) between pre- and postintervention.”
MethodInfluentialAge of onset of mental disorders: a review of recent literature
Kessler, R. C., Amminger, G. P., Aguilar-Gaxiola, S., Alonso, J., Lee, S., & ??st??n, T. B. (2007) · Current Opinion in Psychiatry · doi:10.1097/yco.0b013e32816ebc8c · Cited by 3,457 · cited 1×
“In addition, up to 74% of adults experience their first onset of a mental health diagnosis before the age of 24 years [5].”
BackgroundFitzpatrick, K. K., Darcy, A., & Vierhile, M. (2017) · JMIR Mental Health · doi:10.2196/mental.7785 · Cited by 2,730 · cited 4×
“In recent years, wider access to digital technology and mobile phones has presented new opportunities for overcoming these barriers, by offering the possibility of delivering digital mental health interventions in a more scalable and convenient way [10,11].”
BackgroundResultInfluentialMental health problems and help-seeking behavior among college students
Hunt, J., & Eisenberg, D. (2010) · Journal of Adolescent Health · doi:10.1016/j.jadohealth.2009.08.008 · Cited by 1,949 · cited 1×
“However, about three-quarters of college students who are in need of clinical services do not access them [6]; this is because they have low mental health literacy and do not recognize a need for treatment [7], but also because of the high cost of treatment, low availability, or attitudinal barriers, such as perceived…”
BackgroundBarriers to mental health treatment: results from the WHO World Mental Health surveys
Andrade, L. H., Alonso, J., Mneimneh, Z., Wells, J. E., Al-Hamzawi, A., Borges, G., Bromet, E., Bruffaerts, R., de Girolamo, G., de Graaf, R., et al. (2014) · Psychological Medicine · doi:10.1017/s0033291713001943 · Cited by 1,289 · cited 1×
“However, about three-quarters of college students who are in need of clinical services do not access them [6]; this is because they have low mental health literacy and do not recognize a need for treatment [7], but also because of the high cost of treatment, low availability, or attitudinal barriers, such as perceived…”
BackgroundFrom ideas to efficacy: The ORBIT model for developing behavioral treatments for chronic diseases
Czajkowski, S. M., Powell, L. H., Adler, N., Naar-King, S., Reynolds, K. D., Hunter, C. M., Laraia, B., Olster, D. H., Perna, F. M., Peterson, J. C., et al. (2015) · Health Psychology · doi:10.1037/hea0000161 · Cited by 1,161 · cited 1×
“The study design was based on a mixed methods approach and encompassed two phases of the ORBIT (Obesity-Related Behavioral Intervention Trials) framework [30] for intervention design (Phase I) and preliminary testing (Phase IIa).”
MethodHelp-Seeking and Access to Mental Health Care in a University Student Population
Eisenberg, D., Golberstein, E., & Gollust, S. E. (2007) · Medical Care · doi:10.1097/mlr.0b013e31803bb4c1 · Cited by 1,131 · cited 1×
“However, about three-quarters of college students who are in need of clinical services do not access them [6]; this is because they have low mental health literacy and do not recognize a need for treatment [7], but also because of the high cost of treatment, low availability, or attitudinal barriers, such as perceived…”
BackgroundInkster, B., Sarda, S., & Subramanian, V. (2018) · doi:10.2196/preprints.12106 · Cited by 1,063 · cited 5×
“The total scores are separated into the following four categories to determine the anxiety symptom levels: none (0-4), mild (5-9), moderate (10-14), and severe (≥15) [23,24].”
BackgroundResultInfluential
Research works printed but not linked: 2
- Beck, JS, 2011, Cognitive Behavior Therapy: Basics and Beyond. 2nd edition
- 10.4172/2167-1044.1000330 · doi:10.4172/2167-1044.1000330
Cited by
Artificially intelligent chatbots in digital mental health interventions: a review
Boucher, E. M., Harake, N. R., Ward, H. E., Stoeckl, S. E., Vargas, J., Minkel, J., Parks, A. C., & Zilca, R. (2021) · Expert Review of Medical Devices · doi:10.1080/17434440.2021.2013200 · Cited by 440
Engagement and retention in digital mental health interventions: a narrative review
Boucher, E. M., & Raiker, J. S. (2024) · BMC Digital Health · doi:10.1186/s44247-024-00105-9 · Cited by 181
“While some research suggests that more severe mental health symptoms may be a barrier to engagement [29, 44, 47], other studies have found better levels of adherence among people with more severe mental health symptoms including depression [34], anxiety, and perceived stress [48].”
BackgroundAkpan, I. J., Kobara, Y. M., Owolabi, J., Akpan, A. A., & Offodile, O. F. (2025) · International Transactions in Operational Research · doi:10.1111/itor.13522 · Cited by 137
Hua, Y., Siddals, S., Ma, Z., Galatzer‐Levy, I., Xia, W., Hau, C., Na, H., Flathers, M., Linardon, J., Ayubcha, C., et al. (2025) · World Psychiatry · doi:10.1002/wps.21352 · Cited by 117
Mitsea, E., Drigas, A., & Skianis, C. (2023) · Behavioral Sciences · doi:10.3390/bs13121008 · Cited by 101
Metadata
- Authors
- Gabrielli, S., Rizzi, S., Bassi, G., Carbone, S., Maimone, R., Marchesoni, M., & Forti, S.
- Affiliations
- Fondazione Bruno Kessler (IT)
- Corresponding author
- Gabrielli, Silvia
- ORCID
- Venue
- JMIR mHealth and uHealth
- Publisher
- JMIR Publications Inc.
- Published
- 2021-05-28
- Volume
- 9
- Issue
- 5
- Pages
- e27965
- ISSN
- 2291-5222 (ISSN-L)
- PMC
- PMC8166265
- OpenAlex
- W3159362230
- Cited by
- 125
- References
- 40
- Citations by year
- 2026: 17 · 2025: 27 · 2024: 29 · 2023: 33 · 2022: 9 · 2021: 10
- FWCI
- 17.98
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- JMIR mhealth and uhealth (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Padua Research Archive (University of Padova) (submittedVersion, cc-by)
- Topics
- Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
- Social Sciences › Psychology › Clinical Psychology › COVID-19 and Mental Health
- Social Sciences › Psychology › Developmental and Educational Psychology › Student Stress and Coping
- Keywords
- chatbots, psychoeducational intervention, university students, COVID-19 pandemic, anxiety, stress, mindfulness, cognitive behavioral therapy, positive psychology, digital mental health, user engagement, attrition, ATENA, mobile health, mental health stigma, first-year students, intervention effectiveness, GAD-7, PSS-10
- MeSH terms
- COVID-19, SARS-CoV-2, Adaptation, Psychological, Adult, Female, Humans, Male, Students, Universities, Young Adult, Pandemics
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 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (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.