The Use of Artificial Intelligence Tools and its Impact on the Learning Outcomes of Computer Science Students
Open accessAbstract
This study focuses on the use of Artificial Intelligence (AI) tools and their impact on the learning outcomes of Computer Science students. AI tools are commonly used by students for coding, research, problem-solving, and completing academic tasks. The study aims to determine how AI tools affect students’ academic performance, learning efficiency, and programming skills. A quantitative research design using survey questionnaires is used to collect data from Computer Science students. The findings are expected to show that AI tools help improve productivity and support learning, but excessive use may affect students’ critical thinking and independent problem-solving skills. The study may help provide guidance on the proper and responsible use of AI tools in education
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
1 of 13 cited sources checked · 0 supported · 0 flagged · accuracy not yet measured
- Interpretationcitation
The absence of significant correlation and its negative direction can be explained by the dependency threshold introduced by Chaudhary et al. (2024) and cognitive offloading discovered by Kosmyna et al. (2025).
“The absence of significant correlation and its negative direction can be explained by dependency threshold introduced by Chaudhary et al. (2024) and cognitive offloading discovered by Kosmyna et al. (2025) with electroencephalography technique.”
Quote in the paperSubstantial claim1 of 2 sources checked- cites Chaudhary et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites Kosmyna et al. (2025) source
The Discussion (p. 145) reports weaker EEG connectivity with LLM assistance, and the cognitive-offloading discussion (p. 129) says AI tools may offload some cognitive processes. The paper supports cognitive offloading as an interpretation of its EEG findings, but does not establish it as a definitive discovery. The absence and negative direction of the citing authors’ correlation are their own finding, not a result attributed to Kosmyna et al. (LLM, conf=0.84)
Experimental — accuracy not yet measured
- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Findingown data
Item 10, using AI tools for completing quizzes, was rated the lowest at 2.85.
“Item 10”
Quote in the paperNot a substantial claim - Findingown data
There was a moderate positive correlation between Frequency of Usage and Logical Reasoning (r = 0.5997, p < 0.001).
“There was a moderate positive correlation with Logical Reasoning (r = 0.5997, p < 0.001)”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Findingown data
Frequency of AI engagement was mainly 1-2 times per day (n = 38; 70.37%).
“The frequency of their engagement was mainly within 1-2 times per day (n = 38; 70.37%), which reflects moderate and regular usage.”
Quote in the paperNot a substantial claim - Findingown data
The correlation between Frequency of Usage and Application Contexts was r = 0.7159 (p < 0.001), a highly positive connection.
“Overall, the correlation between Frequency of Usage and Application Contexts had r = 0.7159 (p < 0.001)”
Quote in the paperNot a substantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Findingown data
ChatGPT was the most popularly adopted AI tool, with 42 out of 54 participants (77.78%) using it.
“From Table 3, ChatGPT emerged as the most popularly adopted AI software among the respondents, with 42 out of 54 participants (77.78%) using it.”
Quote in the paperNot a substantial claim - Findingown data
The average weighted mean for the frequency of AI tool usage was 3.22 (Agree/High).
“The average weighted mean for the frequency of AI tool usage was 3.22 (Agree/High).”
Quote in the paperSubstantial claim - Findingown data
Each of the 54 respondents utilized AI-based tools at least once daily.
“As evident from Table 4, each of the 54 respondents utilized AI-based tools at least once daily.”
Quote in the paperNot a substantial claim - Findingown data
The overall weighted mean score for Application Context of AI tool usage was 3.19 (Agree/High).
“The overall weighted mean score (Agree/High level) of 3.19”
Quote in the paperNot a substantial claim - Findingown data
No correlation was identified between Frequency of Usage and Academic Performance (r = -0.2075, p = 0.1323).
“Finally, it is important to mention the fact that no correlation was identified between Frequency of Usage and Academic Performance (r = -0.2075,p = 0.1323).”
- Also stated in: § 5.1.4 Significant Relationship Between AI Tool Usage and Learning Outcomes · plain
“At the same time, there is no correlation between the use of AI tools and academic performance of students.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Also stated in: § 5.1.4 Significant Relationship Between AI Tool Usage and Learning Outcomes · plain
- Findingown data
There was a very high positive correlation between Frequency of Usage and Problem-Solving Abilities (r = 0.8970, p < 0.001).
“Regarding sub-scales, there was a very high positive correlation between Frequency of Usage and Problem-Solving Abilities (r = 0.8970, p < 0.001)”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Interpretationown data
All three subscales were considered high, suggesting students generally agree that frequent AI tool interaction facilitates their learning in all aspects of programming expertise.
“All three subscales were considered high, suggesting that students generally agree that frequent AI tool interaction facilitates their learning in all aspects of programming expertise.”
Quote in the paperSubstantial claim - Interpretationown data
Students effectively apply AI tools for completing academic tasks such as quizzes, performance tasks, analysing codes, and debugging.
“shows that students effectively apply AI tools for completing academic tasks such as quizzes, performance tasks, analysing codes, and debugging.”
Quote in the paperSubstantial claim - Interpretationown data
The more students use AI technology, the greater benefit they receive in troubleshooting abilities, programming strategies, and applying previously acquired knowledge in new circumstances.
“which means that the more students use the AI technology, the greater benefit they receive in relation to their troubleshooting abilities, programming strategies, and the application of the previously acquired knowledge in new circumstances.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Premisecitation
Grützner & Schreck (2024) found that AI-generated code may increase extraneous cognitive load if the code is too complex for students' current level.
“The study found that”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Grützner & Schreck (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Interpretationreasoning
The respondents comprise mainly first-and second-year college students enrolled in basic computer science programs.
“The above statistics imply that the respondents comprise mainly first-and second-year college students, who are currently enrolled in basic computer science programs where the use of AI tools is most applicable.”
Quote in the paperNot a substantial claim - Interpretationown data
Students make good use of AI tools primarily for logical verification, understanding algorithms, debugging, and project deconstruction.
“Conclusion: It may thus be stated that students make good use of AI tools primarily for the purposes of logical verification, understanding algorithms, debugging, and project deconstruction.”
Quote in the paperSubstantial claim - Interpretationown data
Those who frequently use AI also integrate them more effectively in various academic activities.
“This suggests that those who frequently use AI also integrate them more effectively in various academic activities.”
Quote in the paperNot a substantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
- Interpretationown data
AI tools are an integral part of each respondent's daily study routine.
“In the absence of any respondents who utilized AI-based tools 0 times daily, it can be said that AI tools are an integral part of each respondent's daily study routine.”
Quote in the paperNot a substantial claim - Interpretationcitation
Unless there is guidance, AI may result in shallow knowledge of programming concepts.
“This suggests that unless there is guidance, AI may result in”
Quote in the paperSubstantial claim
- Classified as probably not a substantial claim. 19
- May combine two claims. 18
- The quote only partly states this claim. 1
Probably not a substantial claim: may not pass S3 (load-bearing).
Findingown dataThe largest number of respondents (n = 46, 85.19%) were 19 years old.
“the largest number of the respondents (n = 46, 85.19%) were 19 years old”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown data3-5 times daily was reported by n = 13 (24.07%) and 6-10 times per day by n = 3 (5.56%).
“Additionally, some individuals utilized AI-based tools up to 3-5 times daily (n = 13; 24.07%), whereas others used them 6-10 times per day (n = 3; 5.56%).”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Findingown data20-year-olds were n = 6 (11.11%) and 21-year-olds were n = 2 (3.70%).
“then followed by the 20-yearolds (n = 6, 11.11%), and lastly, the 21-year-olds (n = 2, 3.70%).”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataIn terms of learning outcome variables, Logical Reasoning obtained the highest subscale mean of 3.31, Problem-Solving Ability had 3.19, and Academic Performance had 3.15.
“In terms of the three learning outcome variables, Logical Reasoning obtained the highest subscale mean of 3.31, while Problem-Solving Ability had 3.19 and Academic Performance had 3.15.”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataDola accounted for 16.67% (n = 9), Claude 9.26% (n = 5), and Gemini 7.41% (n = 4) of respondents.
“Dola accounted for 16.67% (n = 9) of the respondents, while 9.26% (n = 5) and 7.41% (n = 4) adopted Claude and Gemini, respectively.”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataMales made up 55.56% (n = 30) and females 44.44% (n = 24) of respondents.
“As can be observed from Table 2 below, males made up the larger percentage of respondents (n=30; 55.56%) whereas females were 44.44% (n=24).”
Quote in the paperNot a substantial claimMay combine two claims.
Findingown dataSubscale means for Application Context were Logical Reasoning 3.31, Problem-Solving Ability 3.22, and Academic Performance 3.03.
“In terms of means for the subscales, Logical Reasoning got the highest mean score of 3.31, while Problem-Solving Ability and Academic Performance obtained 3.22 and 3.03, respectively.”
Quote in the paperNot a substantial claimClassified as probably not a substantial claim.
Study design and conductreasoningPearson correlation is applicable because Likert scale data is quantitative in nature.
“Pearson correlation is applicable since data obtained using the Likert scale is quantitative in nature.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
The quote only partly states this claim.
Interpretationown dataAI-aided code analysis and stepwise confirmation of logical accuracy positively affect reasoning abilities of users.
“thus demonstrating how AI-aided code analysis and stepwise confirmation of its logical accuracy positively affect reasoning abilities of users.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKlingbeil et al. (2024) found a likely decrease in human performance due to trust and reliance, especially in the use of incorrect solutions in logic applications.
“The results obtained from the experiments presented by the authors revealed that there is a likely decrease in human performance due to the”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Klingbeil et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductreasoningSimple random sampling reduces bias and unfairness during the selection process.
“This reduces biasness and unfairness during the selection process.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Study design and conductnone givenSimple random sampling ensures that all Computer Science students have equal chances of participating in the study.
“Simple random sampling ensures that all Computer Science students have equal chances of participating in the study.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Study design and conductnone givenParticipation in the study was voluntary and confidentiality was assured.
“Participation will be voluntary and confidentiality assured.”
Quote in the paperNot a substantial claim- Rests on an assumption: Respondents answered the survey honestly.
Classified as probably not a substantial claim. May combine two claims.
PremisecitationGonzaga et al. (2024) indicated that AI should act as incremental assistance and not replace students' logical reasoning, which is critical for long-term acquisition of programming skills.
“Gonzaga et al.'s research indicated that for the AI to ensure the desired effectiveness, it should act as a vehicle for the students to learn by means of”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Gonzaga et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationAghiomesi et al. (2024) found that generative AI tools improve problem-solving skills, coding accuracy, and academic achievement in tertiary institutions.
“In a regional context, Aghiomesi et al. (2024) evaluated the impact of generative AI tools in tertiary institutions and found improvements in problem-solving skills, coding accuracy, and academic achievement.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Aghiomesi et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationYilmaz and Yilmaz (2024) reported that generative AI tools positively affect students' computational thinking skills, programming self-efficacy, and learning motivation.
“Similarly, Yilmaz and Yilmaz (2024) reported that the use of generative AI tools positively affects students' computational thinking skills, programming self-efficacy, and learning motivation.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Yilmaz and Yilmaz (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
May combine two claims.
Interpretationown dataUsage of AI tools by Computer Science students of SNSU is quite frequent and contributes to logical reasoning and problem-solving ability of students.
“In general, one may conclude that usage of AI tools by Computer Science students of SNSU is quite frequent and contributes to logical reasoning and problem-solving ability of students.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
May combine two claims.
Interpretationown dataThe most notable advantage of AI use is its potential to aid students in following the logic of the program, checking whether the coding process is valid, and justifying the steps in the programming process.
“Based on the findings, the most notable advantage of AI use is its potential to aid students in following the logic of the program, checking whether the coding process is valid, and justifying the steps in the programming process.”
Quote in the paperSubstantial claimClassified as probably not a substantial claim.
PremisecitationAI-powered tools such as code generators, intelligent tutoring systems, and automated debugging assistants are increasingly utilized to support students in mastering complex programming concepts.
“AI-powered tools such as code generators, intelligent tutoring systems, and automated debugging assistants are increasingly utilized to support students in mastering complex programming concepts (Córdova-Esparza et al., 2024).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Córdova-Esparza et al., 2024
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
May combine two claims.
Interpretationown dataStudents in this study use AI tools in a structured way by focusing on learning logic and adopting proper problem-solving strategies.
“Students in this study seem to use AI tools in a structured way by focusing on their ability to learn about the logic and adopt proper problem-solving strategies.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationSwaraj and Chavan (2024) highlighted that AI threatens to standardize human cognitive thinking and creativeness.
“Swaraj and Chavan's study (2024) on the impact of the use of ChatGPT on the cognitive thinking and creativeness of human beings highlighted how AI, although the basis for creating innovation itself, simultaneously threatens to standardize human cognitive thinking and creativeness.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Swaraj and Chavan (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationGonzaga et al. (2024) posited that for AI systems to be effective, it is important to integrate them into the curriculum to encourage independence.
“Finally, Gonzaga et al. (2024) posited that for AI systems to be effective, it was important to integrate them into the curriculum in a bid to encourage independence.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Gonzaga et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
May combine two claims.
Study design and conductown dataThe balanced ratio of male and female respondents guarantees that findings would not be biased by gender preferences, making findings applicable to both sexes.
“The more balanced ratio of male and female respondents guarantees that the research findings would not be biased by gender preferences, making the findings applicable to both sexes in the Computer Science discipline.”
Quote in the paperNot a substantial claimMay combine two claims.
Interpretationown dataThe frequency of AI usage is consistent and moderate, indicating AI assistance is a routine practice rather than an emergency measure.
“This information is particularly relevant since the frequency of AI usage seems consistent and moderate, indicating that AI assistance is a routine practice rather than an emergency measure.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Interpretationown dataThe low rating of using AI for quizzes shows constraints on academic integrity when using AI tools in assessments.
“showing possibly more constraints on academic integrity when using AI tools in assessments.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.
Study design and conductreasoningSimple random sampling increases the accuracy and reliability of the collected data.
“This technique increases the accuracy and reliability of the collected data.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationOribhabor (2023) identified perceived usefulness and institutional support as key determinants of AI adoption.
“While Oribhabor (2023) identified perceived usefulness and institutional support as key determinants of AI adoption”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Oribhabor (2023)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationKhan et al. (2024) stated that pedagogical software tools should be appropriate to the student's mental capacity and not overwhelm it.
“In this context, Khan et al. (2024) highlighted the importance of improving automated feedback instruction in terms of Cognitive Load Theory and stated:”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Khan et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationKlingbeil et al. (2024) revealed that human users are likely to experience delegation bias in evaluating AI output, failing to critique it properly.
“This concern was further supported by another study done by Klingbeil et al. (2024), which revealed that human users, in most cases, are likely to experience a”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Klingbeil et al. (2024)
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).
Interpretationown dataThe relatively low value for Academic Performance compared to others is an early sign that students recognize a difference between AI-assisted skill acquisition and academic performance.
“The relatively low value for the Academic Performance variable compared to others is an early sign that students recognize a difference between AI-assisted skill acquisition and academic performance.”
Quote in the paperNot a substantial claimMay combine two claims.
Interpretationown dataAlthough students report higher problem-solving and reasoning skills with frequent AI usage, they do not have better quiz, test and task completion scores.
“This result deserves particular attention -although the students report higher problem-solving and reasoning skills in cases of frequent AI usage, they do not have better quiz, test and task completion scores.”
Quote in the paperSubstantial claim- Rests on an assumption: The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
The paper relies on this without support.
Programming courses are traditionally characterized by high cognitive demand.
“Because programming courses are traditionally characterized by high cognitive demand, AI tools are frequently viewed as essential resources for improving student understanding and academic performance.”
The paper relies on this without support.
AI tools are frequently viewed as essential resources for improving student understanding and academic performance.
“Because programming courses are traditionally characterized by high cognitive demand, AI tools are frequently viewed as essential resources for improving student understanding and academic performance.”
The paper relies on this without support.
The data satisfy the assumptions required for Pearson correlation, such as normality and linearity.
“Pearson correlation is applicable since data obtained using the Likert scale is quantitative in nature.”
“Spearman Rank Correlation will be used as an alternative where data do not meet the requirements for applying parametric tests such as normality test.”
The paper relies on this without support.
Respondents answered the survey honestly.
“Participation will be voluntary and confidentiality assured.”
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.
Reported outcome measures differ from the planned measures
MinorThe methodology names Study Dedication as one of three measures of the dependent variable, but the results instead name Academic Performance as a subscale. The reported set omits Study Dedication and treats the dependent variable itself as one of its components.
“Three factors will describe it: (1) Study Dedication, (2) Logical Thinking, and (3) Problem-solving Ability.”
“In terms of the three learning outcome variables, Logical Reasoning obtained the highest subscale mean of 3.31, while Problem-Solving Ability had 3.19 and Academic Performance had 3.15.”
Cites
Trust and reliance on AI — An experimental study on the extent and costs of overreliance on AI
Klingbeil, A., Grützner, C., & Schreck, P. (2024) · Computers in Human Behavior · doi:10.1016/j.chb.2024.108352 · Cited by 450 · see citing works
Kosmyna, N., Hauptmann, E., Yuan, Y. T., Situ, J., Liao, X.-H., Beresnitzky, A. V., Braunstein, I., & Maes, P. (2025) · arXiv · doi:10.48550/arxiv.2506.08872 · Cited by 220 · see citing works
Flagged first, then by citations
Research works printed but not linked: 5
- [1]Chaudhary, A. A., Arif, S., Calimlim, R. F., Jr., Khan, S. Z., & Sadia, A. (2024). The impact of AI-powered educational tools on student engagement and learning outcomes at higher education levels is important. International Journal of Contemporary Issues in Social Sciences, 3(2), 2842–2852.
- https://scholar.google.com/scholar?q=Chaudhary+Arif+Calimlim+Khan+Sadia+2024+AI-powered+educational+tools+student+engagement+learning+outcomes
- [2]Gonzaga, J. T., Jiang, Y., & Vassar, A. (2024). Scaffolding independence: The role of AI agents in the introductory programming curriculum. ACM Technical Symposium on Computer Science Education (SIGCSE 2024). https://scholar.google.com/scholar?q=Gonzaga+Jiang+Vassar+2024+Scaffolding+independence+AI+agents+introductory+programming+SIGCSE
- [3]Grützner, C., & Schreck, P. (2024). Student-perceived cognitive load of LLM-generated programming exercises. International Journal of Computer Science Education Research. https://scholar.google.com/scholar?q=Gr%C3%BCtzner+Schreck+2024+student+perceived+cognitive+load+LLM+programming+exercises
- [4]Khan, M. F. A., Feri, L. E., Nguyen, H., & Karimi, H. (2024). Empowering CS1 educators: Enhancing automated feedback instruction with Cognitive Load Theory. Proceedings of the 55th ACM Technical Symposium on Computer Science Education. https://scholar.google.com/scholar?q=Khan+Feri+Nguyen+Karimi+2024+Empowering+CS1+educators+automated+feedback+Cognitive+Load+Theory Klingbeil, A.,
- Authors
- Xyrus L. Bautista, R. N. S., & Jayson V. Sereno, J. R. E. F.
- Affiliations
- Surigao Del Norte State University, Surigao City, Philippines
- Venue
- International Journal of Advanced Research in Science, Communication and Technology
- Publisher
- Naksh Solutions
- Published
- 2026-06-11
- Pages
- 30
- ISSN
- 2581-9429 (ISSN-L)
- OpenAlex
- W7164354827
- Cited by
- 0
- References
- 1
- FWCI
- 0.00
- Type
- journal article
- Open access
- diamond
- doi.org
- Licence
- other-oa
- Locations
- International Journal of Advanced Research in Science Communication and Technology (publishedVersion, other-oa)
- Topics
- Social Sciences › Education › Technology-Enhanced Education Studies
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Innovative Education Methods and Tools
- Social Sciences › Education › E-Learning and COVID-19
- Keywords
- AI tools, computer science students, learning outcomes, academic performance, programming skills, learning efficiency, critical thinking, problem-solving skills
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 2026-10-11.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-11.
Cited sources are checked by 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
