CoChat
Sign Up
Cited by 127Artificial Intelligence in EducationTechnology Adoption and User BehaviourTechnology-Enhanced Education Studies

Artificial intelligence and communication technologies in academia: faculty perceptions and the adoption of generative AI

Open access
Shata, Aya, Hartley, Kendall·2025·International Journal of Educational Technology in Higher Education
DOI 10.1186/s41239-025-00511-7
Open this paper in CoChat — free account

Abstract

Artificial intelligence (AI) is ushering in an era of potential transformation in various fields, especially in educational communication technologies, with tools like ChatGPT and other generative AI (GenAI) applications. This rapid proliferation and adoption of GenAI tools have sparked significant interest and concern among college professors, who are dealing with evolving dynamics in digital communication within the classroom. Yet, the effect and implications of GenAI in education remain understudied. Therefore, this study employs the Technology Acceptance Model (TAM) and the Social Cognitive Theory (SCT) as theoretical frameworks to explore higher education faculty’s perceptions, attitudes, usage, and motivations, as the underlying factors that influence their adoption or rejection of GenAI tools. A survey was conducted among full-time higher education faculty members (N = 294) recruited from two mid-size public universities in the US. Results found that college professors’ perceived usefulness of AI predicted their attitudes and intention to use and adopt the technology, more than their perceived ease of use. Trust and social reinforcement strongly influenced college professors’ GenAI adoption decisions and acted as significant mediators to better understand the relationship between TAM and SCT. Findings emphasized the power of social dynamics in shaping professors’ self-efficacy, attitudes, and use of GenAI. Trust enhances peer influence and affects how perceived usefulness shapes users’ willingness to adopt technology, whereas self-efficacy has a minimal impact. This research provides valuable insights that inform higher education policies aimed at improving the educational experience for college students in an AI-driven workforce.

AI snapshot

Written from the paper’s full text.

TL;DR

This survey of 294 full-time social sciences/humanities faculty at two U.S. public universities applies TAM and SCT to generative AI adoption. Perceived usefulness, trust, and social reinforcement were central predictors, while perceived ease of use and self-efficacy had limited direct effects.

Key findings

  • Perceived usefulness (PU) and perceived ease of use (PEU) predicted attitudes toward GenAI [F(2, 188) = 108.476, p < 0.001, R² = 0.536], with PU stronger (β = 0.655) than PEU (β = 0.163); PU and attitude predicted behavioral intention [R² = 0.579], but PEU was not significant (p = 0.645).
  • Trust significantly influenced PU, PEU, attitude, and behavioral intention (e.g., attitude b = 0.630, p < 0.001; intention b = 0.503, p < 0.001), and moderated the PU→intention relationship: the effect was stronger at low trust (B = 0.85) than high trust (B = 0.48).
  • Social reinforcement significantly predicted PU, PEU, attitude, and self-efficacy; it mediated PEU→intention (B = 0.05, 95% CI 0.00–0.12) but not PU→intention. Trust mediated social reinforcement→intention (B = 0.11, 95% CI 0.04–0.19).
  • Self-efficacy significantly predicted PEU (b = 0.131, p < 0.001) but not attitude (p = 0.113) or behavioral intention (p = 0.571). 66.4% of professors reported having used GenAI before.

Methods

  • Online Qualtrics survey of full-time higher education faculty in social sciences and humanities at two mid-size public U.S. universities, one East Coast and one Southwest; N = 294 completed; data collected January–March 2024.
  • Used TAM and SCT constructs: behavioral intention, PU, PEU, AI self-efficacy, attitudes, social reinforcement, trust, comfort with technology, and AI concerns, measured with adapted multi-item scales.
  • Analyses included multiple regression, multivariate multiple regression, and SPSS PROCESS macro mediation/moderation with 5,000 bootstrap iterations.

Limitations

  • Stated: targeted non-STEM sampling limits generalizability; two mid-sized research-intensive public U.S. universities may not generalize to smaller, private, teaching-intensive, or international settings; self-report on a controversial topic may produce socially desirable responses; single time point cannot capture evolving attitudes; self-selection bias; unmeasured factors include institutional support/culture, training, and discipline.
  • Evident: main analyses used subsets of GenAI users (df around 188–193) rather than all 294, and non-users were routed to different questions and terminated; the cross-sectional survey cannot establish causal direction.

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

Integrity signals

Registry

Crossref · 10.1186/s41239-025-00511-7

Cites

Statistics

Claims the paper makes

0 of 15 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Results

    The model predicting attitudes from perceived usefulness and perceived ease of use explained 54% of the variance.

    “Results found that both variables statistically significantly predicted the attitude to use GenAI [F (2, 188) = 108.476, p < 0.001, R2 = 0.536], accounting for 54% of the variance.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Discussion

    Social reinforcement has no direct effect on behavioral intention to use GenAI.

    “The mediation analysis showed there was no direct effect from social reinforcement in behavioral intention, but there was a significant indirect effect of trust mediating the relationship between social reinforcement and behavioral intention.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Discussion

    For perceived ease of use, there was a significant mediation effect on behavioral intention to use GenAI.

    “On the other hand, for perceived ease of use, there was a significant mediation effect on behavioral intention to use GenAI.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results

    GenAI self-efficacy did not significantly predict behavioral intention to use GenAI (p = 0.571).

    “Results found that GenAI Self-efficacy statistically significantly predicted PEU (b = 0.131, p < 0.001, 95% CI [0.090, 0.172]), but neither predicted attitude (p = 0.113), nor behavioral intention to use GenAI (p = 0.571).”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results

    Social reinforcement mediated the relationship between perceived ease of use and behavioral intention to use GenAI (B = 0.05, 95% CI 0.00 to 0.12).

    “For PEU, results showed that the model was significant [F(2, 189) = 15.19, p < 0.01, R2 = 0.14] and that social reinforcement mediated the relation between PEU and behavior intention (B = 0.05, 95% CI of 0.00 to 0.12)”
    Quote in the paperSubstantial claim
  • May combine two claims. 16
  • Classified as probably not a substantial claim. 11

  • May combine two claims.

    Findingown data
    § Discussion

    Of the total effect of perceived ease of use on behavioral intention (B = 0.498), the indirect effect through attitudes was B = 0.413 and the direct effect was B = 0.085.

    “mediation analysis revealed that of the total effect (B = 0.498), the indirect effect accounted for (B = 0.413), whereas the direct effect was relatively minor, at (B = 0.085).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Discussion

    Social reinforcement has a direct and positive influence on perceptions of usefulness and ease of use, fostering favorable attitudes and encouraging the overall decision to adopt GenAI technology.

    “results found that social reinforcement has a direct and positive influence on perceptions of usefulness and ease of use, fostering favorable attitudes and encouraging the overall decision to adopt the GenAI technology.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    In the mediation model of social reinforcement on behavioral intention through trust, only the indirect effect was significant; the direct effect was not significant (p = 0.079).

    “It is worth noting that only the indirect effect was significant, while the direct effect was not significant (p = 0.079).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Discussion

    Trust is the most significant predictor influencing all factors (perceived usefulness, perceived ease of use, social reinforcement, self-efficacy, attitudes) related to the decision to adopt GenAI, closely followed by social reinforcement.

    “Overall, trust emerges as the most significant predictor influencing all factors (PU, PEU, SR, SE, Att) related to the decision to adopt GenAI technology, closely followed by social reinforcement.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    The simple slopes of perceived usefulness on behavioral intention are positive at all levels of trust, but the magnitude of this positive effect decreases as trust increases.

    “Analysis revealed that the simple slopes at different levels of Trust (e.g., low, medium, and high levels) are all positive, indicating that the PU has a positive effect on the behavior intention at all levels of Trust. However, the magnitude of this positive effect decreases as the Trust increases.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    The Technology Acceptance Model (TAM) holds that technology acceptance is strongly influenced by perceived ease of use (PEU) and perceived usefulness (PU).

    “The Technology Acceptance Model (TAM) suggests that technology acceptance is strongly influenced by its perceived ease of use (PEU) and perceived usefulness (PU) (Davis, 1989; Granić & Marangunić, 2019; Venkatesh et al., 2003).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Davis, 1989

      Not checked in the free run.

    • cites Granić & Marangunić, 2019 source

      Not checked in the free run.

    • cites Venkatesh et al., 2003 source

      Not checked in the free run.

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

    Study design and conductnone given
    § Acknowledgements

    The study was reviewed and approved by the University of Nevada, Las Vegas Office of Research Integrity.

    “The study was reviewed and approved by the University of Nevada, Las Vegas Office of Research Integrity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § Discussion

    Trust can foster a sense of simplicity of the technology, which in turn boosts professors' self-efficacy.

    “It can also foster a sense of simplicity, which, in turn, boosts their self-efficacy—their confidence in their ability to effectively utilize the technology.”
    Quote in the paperSubstantial claim

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

    Interpretationreasoning
    § Discussion

    A possible explanation for the limited significance of self-efficacy is that faculty may question GenAI's relevance or effectiveness.

    “Another possible explanation for the limited significance of self-efficacy, despite faculty feeling capable of using AI technology (i.e., high self-efficacy), they may question its relevance or effectiveness.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given
    § Methods

    All theoretical constructs in the study used existing measures adapted to the AI context and multi-item scales validated in previous research.

    “All the theoretical constructs in this study used existing measures, adapted to fit the AI context and measured using multi-item scales validated in previous research.”
    Quote in the paperNot a substantial claim

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

    Premisecitation
    § Introduction

    A recent meta-analysis found that the social influence association with behavior was non-significant but had an overall significant impact on attitudes.

    “While a recent meta-analysis concluded that the social influence association with behavior was non-significant, it did find an overall significant impact on attitudes (Marikyan et al., 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Marikyan et al., 2023 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § Discussion

    Increased trust can lead to more positive attitudes and stronger intentions to adopt GenAI technology.

    “Increased trust can lead to more positive attitudes and stronger intentions to adopt GenAI technology.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Abstract

    Trust and social reinforcement strongly influenced college professors' GenAI adoption decisions and acted as significant mediators between TAM and SCT.

    “Trust and social reinforcement strongly influenced college professors’ GenAI adoption decisions and acted as significant mediators to better understand the relationship between TAM and SCT.”
    Quote in the paperSubstantial claim

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

    Premisenone given
    § Introduction

    Behavioral intention to use technology is directly related to attitude toward the technology, which is in turn influenced by perceived ease of use and perceived usefulness.

    “In brief, the behavioral intention to use technology will be directly related to the attitude towards the technology, which is in turn influenced by the perceived ease of use (PEU) and the perceived usefulness (PU) of the technology.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Study design and conductnone given
    § Conclusion

    Self-selection bias can influence the findings and limit the generalizability of the results.

    “Another potential limitation is self-selection bias can influence the findings and limit the generalizability of the results.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § Conclusion

    Perceived ease of use alone is insufficient to encourage GenAI adoption, but when coupled with positive attitudes and social circles it becomes influential in encouraging adoption.

    “However, perceived ease alone is insufficient to encourage adoption, but when coupled with positive attitudes and social circles, it becomes influential in encouraging adoption.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given
    § Conclusion

    Reliance on self-report data regarding potentially controversial topics may result in responses that are viewed as more favorable at the expense of accuracy.

    “The reliance on self-report data regarding potentially controversial topics may result in responses that might be viewed as more favorable at the expense of accuracy.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given
    § Methods

    Attention-check questions were added to the survey to ensure respondents were paying sufficient attention and to avoid speeding or random answers that could harm data quality.

    “Also, a couple of attention-check questions were added to the survey to ensure respondents were paying sufficient attention and avoid speeding through the survey or giving random answers that could harm data quality.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § Abstract

    Trust enhances peer influence and affects how perceived usefulness shapes users' willingness to adopt technology, whereas self-efficacy has a minimal impact.

    “Trust enhances peer influence and affects how perceived usefulness shapes users’ willingness to adopt technology, whereas self-efficacy has a minimal impact.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    One's attitude towards technology can be influenced by its perceived ease of use and perceived usefulness.

    “One’s attitude towards technology can be influenced by its perceived ease of use and its perceived usefulness (Davis et al., 1989; Granić & Marangunić, 2019).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Davis et al., 1989

      Not checked in the free run.

    • cites Granić & Marangunić, 2019 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Introduction

    Users' attitudes towards technological advancements are crucial in determining whether they will embrace innovations.

    “Users’ attitudes towards technological advancements are crucial in determining whether or not they will embrace innovations (Davis, 1989; Granić & Marangunić, 2019).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Davis, 1989

      Not checked in the free run.

    • cites Granić & Marangunić, 2019 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § Introduction

    Trust has a significant impact on perceived ease of use, perceived usefulness, and attitudes toward GenAI.

    “These findings underscore that trust has a significant impact on perceived ease of use, perceived usefulness, and attitudes toward GenAI.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Choung et al., 2022 source

      Not checked in the free run.

    • cites Tanantong and Wongras, 2024 source

      Not checked in the free run.

22 of 22

  • The paper relies on this without support.

    The effect and implications of GenAI in education remain understudied.

    “Yet, the effect and implications of GenAI in education remain understudied.”

  • The paper relies on this without support.

    The use of GenAI in social sciences and humanities differs fundamentally from its application in STEM disciplines.

    “The sample represented all faculty members in the social sciences and humanities disciplines as they use GenAI in a similar manner that differs fundamentally from its application in STEM disciplines.”
2 of 2

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?

  • Using the technology acceptance framework to design faculty development primarily predicts adoption intention rather than use and performance outcomes needed in actual educational practice.

    “Using the technology acceptance framework to design faculty development primarily predicts adoption intention [6, 7] rather than use and performance outcomes needed in actual educational practice [1].”

    From Integrating Task-Technology Fit and Community of Practice Theories to Enhance Medical Educator Skill Confidence in Generative AI: Design, Implementation, and Pilot Outcomes (2025) · cites this paper as [6]

    The Methods and Discussion (pages not provided) show that the study measured intention to adopt GenAI and found TAM factors associated with attitudes and behavioral intentions. It did not evaluate faculty-development designs or establish that actual use and performance outcomes are needed in educational practice. Its performance measure was perceived usefulness, not observed performance. (LLM, conf=0.82)

    Experimental — accuracy not yet measured

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.

  • ConsistentF(2, 188) = 108.476reported p < .001 · recomputed p < .001

    “F (2, 188) = 108.476, p < .001”

  • Consistentt(188) = 12.221reported p < .001 · recomputed p < .001

    “t(188) = 12.221, p < .001”

  • Consistentt(188) = 3.04reported p < .01 · recomputed p = .003

    “t(188) = 3.040, p < .01”

  • ConsistentF(3, 187) = 85.84reported p < .001 · recomputed p < .001

    “F (3, 187) = 85.84, p < .001”

  • Consistentt(187) = 5.07reported p < .001 · recomputed p < .001

    “t(187) = 5.07, p < .001”

  • Consistentt(187) = 6.6307reported p < .001 · recomputed p < .001

    “t(187) = 6.6307, p < .001”

  • ConsistentF(1, 193) = 30.112reported p < .001 · recomputed p < .001

    “F (1, 193) = 30.112, p < .001”

  • Consistentt(193) = 5.487reported p < .001 · recomputed p < .001

    “t(193) = 5.487, p < .001”

  • ConsistentF(2, 188) = 129.2reported p < .01 · recomputed p < .001

    “F(2, 188) = 129.20, p < .01”

  • ConsistentF(2, 188) = 102.49reported p < .01 · recomputed p < .001

    “F(2, 188) = 102.49, p < .01”

  • ConsistentF(2, 189) = 15.19reported p < .01 · recomputed p < .001

    “F(2, 189) = 15.19, p < .01”

  • ConsistentF(3, 191) = 68.41reported p < .01 · recomputed p < .001

    “F(3, 191) = 68.41, p < .01”

  • Consistentt(191) = 11.21reported p < .001 · recomputed p < .001

    “t(191) = 11.21, p < .001”

  • Consistentt(191) = 5.24reported p < .001 · recomputed p < .001

    “t(191) = 5.24, p < .001”

  • ConsistentF(2, 189) = 24.2reported p < .01 · recomputed p < .001

    “F(2, 189) = 24.20, p < .01”

15 of 15

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.

  • TAM hypotheses described as all significant despite a nonsignificant result

    Minor

    The results report that perceived ease of use did not significantly predict behavioral intention, so not all TAM hypotheses were statistically significant.

  • Social reinforcement scale described as both seven-item and five-statement

    Minor

    The measure is described in the same sentence as having seven items and being composed of five statements.

2 of 2

The citation lists haven’t been fetched yet

References 35 · Cited by 127

Research works printed but not linked: 2

  1. Mollick, E. (2024). Co-Intelligence: Living and Working with AI. Penguin Publishing Group. http://ebookcentral.proquest.com/lib/unlv/detail.action?docID=30848006
  2. Venkatesh, V., Thong, J. Y. L., & Xu, X. (2016). Unified Theory of Acceptance and Use of Technology: A Synthesis and the Road Ahead (SSRN Scholarly Paper 2800121). https://papers.ssrn.com/abstract=2800121

Authors
Shata, A., & Hartley, K.
Affiliations
University of Nevada, Las Vegas (US)
Corresponding author
Shata, Aya
ORCID
Venue
International Journal of Educational Technology in Higher Education
Publisher
Springer Science and Business Media LLC
Published
2025-03-14
Volume
22
Issue
1
ISSN
2365-9440 (ISSN-L)
DOI
10.1186/s41239-025-00511-7
OpenAlex
W4408391877
Semantic Scholar
b0c5871ea0d431e3d473745e065e645f6a046ce6
Cited by
127
References
35
Citations by year
2026: 89 · 2025: 38
FWCI
130.63
Type
journal article
Language
en
Open access
Licence
Locations
  • International Journal of Educational Technology in Higher Education (publishedVersion, cc-by)
  • Digital Scholarship - UNLV (University of Nevada Reno) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
  • Social Sciences › Decision Sciences › Information Systems and Management › Technology Adoption and User Behaviour
  • Social Sciences › Education › Technology-Enhanced Education Studies
Keywords
generative AI, AI adoption in higher education, teacher perceptions, technology acceptance, social cognitive theory, trust in AI, peer influence, perceived usefulness, perceived ease of use, self-efficacy, AI in the classroom, ChatGPT

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 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

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