Acceptance of Artificial Intelligence in English Learning: Perceptions
of A2-B1 Level University Students at the University of Guayaquil
Aceptación
de la inteligencia artificial en el aprendizaje del inglés: percepciones de
estudiantes universitarios de nivel A2-B1 de la Universidad de Guayaquil
Gabriela Geovanna Guevara Enríquez*
Pablo Fernando Ordoñez Ordoñez*
Introduction
Artificial
intelligence (AI) has established itself as an emerging tool in teaching and
learning processes, particularly in the teaching of English as a foreign
language, where its application enables the creation of personalized
activities, provides immediate feedback, and facilitates self-directed learning
(Peña-Acuña & Corga Fernandes Durão, 2024;
Moradi, 2025). In the context of higher education, these technologies have been
associated with improvements in motivation, academic performance, and engagement
with linguistic content (Ekizer, 2025; Guzmán
Alvarado & Naranjo Andrade, 2025).
Various studies have
indicated that students’ acceptance of AI does not depend solely on
technological availability, but on perceptual factors related to its
usefulness, ease of use, and trust in the tool (Annamalai et al., 2025; Ursavaş et al., 2025). In this regard, the Technology
Acceptance Model (TAM), proposed by Davis (1989), has been widely used to
explain the adoption of educational technologies, establishing that the
intention to use is determined primarily by perceived usefulness (PU) and
perceived ease of use (PEOU).
Although the TAM has
been criticized for its parsimony and for failing to incorporate complex social
or contextual variables, recent research continues to validate its relevance in
exploratory studies on emerging technologies in higher education, especially
when analyzing initial perceptions of adoption (Kanont et al., 2024; Mittal et
al., 2025). In the field of AI-mediated English learning, the model has proven
useful for identifying patterns of acceptance and willingness to use among
university students (Huang & Mizumoto, 2024; Mehrvarz et al., 2025).
Internationally, most
empirical studies on AI acceptance in English language learning focus on Asian
and Middle Eastern contexts (Alotaibi et al., 2025; Moradi, 2025), while
evidence from Latin American public universities remains limited. In Ecuador,
existing research has focused primarily on descriptive analyses of the impact
of AI or on conceptual reviews of its challenges and opportunities (Jara
Alcívar, 2024; Michilena et al., 2025), without delving into explanatory models
that allow for an understanding of the factors influencing usage intention from
the student’s perspective.
In this context, the
present study aimed to analyze the perceptions of A2 and B1 level university
students regarding the acceptance of artificial intelligence in English
language learning, based on the three dimensions of the TAM model: perceived
usefulness, perceived ease of use, and intention to use. This analysis seeks to
provide empirical evidence that helps understand how students value AI as an
educational resource and how these perceptions relate to one another within the
context of Ecuadorian public higher education.
Materials
and methods
The study adopted a
quantitative, non-experimental, cross-sectional, and correlational approach,
with the aim of analyzing the relationship between perceived usefulness,
perceived ease of use, and the intention to use artificial intelligence in
English language learning.
The research was
conducted at the University of Guayaquil, in the language department, where
English courses are taught to students from various faculties. This setting was
chosen due to the institution’s interest in strengthening the teaching of
English as a foreign language. Despite being a public university, the
institution has invested in external platforms such as the Buckingham Center
and English Discoveries, demonstrating its commitment to incorporating
technological resources into teaching and learning processes.
The study population
consisted of university students at levels A2 and B1, according to the Common
European Framework of Reference for Languages (CEFR). Non-probabilistic
convenience sampling was used, taking into account the availability and
accessibility of the study groups. The final sample consisted of 296 students,
mostly women (67.9%) and young people between the ages of 18 and 19. 83.4% were
at level A2 and 16.6% at level B1. Regarding the use of artificial
intelligence, the majority reported using it occasionally (57.8%), with ChatGPT
being the most widely used tool (67.6%), followed by Grammarly and DeepL. These data suggest a moderate and gradual adoption
of the technology by students.
The data collection
instrument was a structured questionnaire using a five-point Likert scale (1 =
strongly disagree to 5 = strongly agree), designed based on Davis’s (1989)
Technology Acceptance Model (TAM) and adapted to the context of learning
English through AI tools. The questionnaire included 15 items distributed
across three dimensions: perceived usefulness (6 items), perceived ease of use
(5 items), and intention to use (4 items). The instrument underwent a content
validation process through expert judgment. Three university professors
specializing in English language teaching evaluated the appropriateness,
clarity, and relevance of each item using a four-point scale. The observations
made were minor and allowed for adjustments to wording details without altering
the structure of the instrument. Subsequently, a pilot study was conducted with
30 students whose characteristics were similar to those of the final sample.
This pilot study confirmed understanding of the items and provided preliminary
evidence of reliability. The Cronbach’s alpha values obtained in this phase
ranged from .814 to .871, indicating adequate internal consistency prior to the
final application.
To ensure
methodological transparency and the study’s replicability, representative
examples of the wording used in each dimension of the instrument are presented.
In the perceived usefulness (PU) dimension, one of the items was: “The use of
artificial intelligence tools improves the quality of my English assignments.”
For perceived ease of use (PEOU), the statement “Interacting with AI platforms
to learn vocabulary is clear and understandable to me” was included. Finally,
in the intention to use (BI) dimension, an example was: “I intend to continue
using artificial intelligence in my future academic work in English.” These
items were adapted to the context of English language learning while
maintaining the conceptual structure originally proposed by Davis (1989).
The questionnaire was
administered virtually via Google Forms, ensuring the anonymous and voluntary
participation of respondents. The data obtained were processed using SPSS
software, applying descriptive statistics (mean, standard deviation, minimum
and maximum values) and Pearson’s correlation coefficient to analyze the
relationship between the variables of the TAM model.
Results
To ensure the validity
of subsequent analyses, the internal consistency of the instrument was first
assessed. The Cronbach’s alpha values obtained indicated high reliability
across the three dimensions of the TAM model, with coefficients exceeding .87,
confirming the internal consistency of the scales used.
Descriptively, the
results show a favorable perception of artificial intelligence in English
language learning. The perceived ease of use (PEOU) dimension had the highest
mean (M = 3.47; SD = 0.86), followed by perceived usefulness (PU) (M = 3.34; SD
= 0.87) and intention to use (BI) (M = 3.26; SD = 0.90). Although all three
dimensions fall above the midpoint of the Likert scale (3), a slight descending
hierarchy is observed from the perception of ease to the final behavioral
intention.
Correlational analysis
using Pearson’s correlation coefficient revealed positive and statistically
significant associations among the model’s variables. The strongest
relationship was found between perceived utility and intention to use (r =
.767, p < .001), indicating that the higher the perceived utility, the
greater the willingness to use artificial intelligence in academic activities.
Likewise, ease of use showed a significant correlation with usage intention (r
= .709, p < .001), while the relationship between perceived usefulness and
ease of use was also strong (r = .694, p < .001).
Taken together, these
results confirm the theoretical structure of the Technology Acceptance Model in
the context studied, demonstrating that perceptions of utility and ease of use
operate as interrelated factors that influence students’ stated intention to
use.
Descriptive
statistics, reliability statistics, and bivariate correlations are presented in
an integrated manner in the following table:
Table 1. General matrix of
empirical results
|
Dimensión |
α |
Media |
DE |
PU |
PEOU |
BI |
|
Utilidad percibida (PU) |
.912 |
3.34 |
0.87 |
— |
.694** |
.767** |
|
Facilidad de uso percibida (PEOU) |
.879 |
3.47 |
0.86 |
.694** |
— |
.709** |
|
Intención de uso (BI) |
.909 |
3.26 |
0.90 |
.767** |
.709** |
— |
Figure 1
Radial representation
of the average dimensions of the TAM model
Figure 1 illustrates
the slight hierarchy among the dimensions, showing a greater emphasis on
perceived ease of use, followed by perceived usefulness, and finally, intention
to use.
The results obtained
provide a clearer understanding of how A2- and B1-level university students
perceive artificial intelligence in their English learning. First, the high
reliability of the instrument and the high scores on perceived usefulness, ease
of use, and intention to use indicate that the TAM model functions
appropriately in this context. These results align with those reported in
recent international research, where students demonstrate a favorable attitude
toward AI when they perceive it as facilitating their tasks and offering
immediate support (Hwang et al., 2025; Shahzad et al., 2024; Wu et al., 2024).
Furthermore, the high
and statistically significant correlations found between perceived usefulness,
ease of use, and behavioral intention reinforce the structure proposed by the
Technology Acceptance Model, according to which these dimensions maintain direct
and predictive relationships with one another (Davis, 1989). This structural
pattern confirms that, in the context of AI-mediated English learning, the
explanatory logic of the TAM holds empirical consistency.
An important finding
of this study is that perceived usefulness scored the highest average. From a
teaching perspective, this makes sense because students tend to value tools
that “directly help them” with immediate tasks: correcting a text, generating examples,
or practicing vocabulary. AI offers visible benefits in a short time, which
explains why this dimension ranks as the primary predictor of usage intention.
This pattern has also been observed in international studies, where utility has
proven to be the most decisive factor in technology adoption (Hwang et al.,
2025).
However, although the
TAM demonstrates explanatory power in this study, it is important to note that
it was initially developed to analyze the acceptance of traditional
technological systems (Davis, 1989). In the case of generative artificial
intelligence, student-technology interaction may involve additional dynamics,
such as the perceived accuracy of responses, trust in the generated
information, and the pedagogical support necessary for its academic use. In
this regard, while the core dimensions of the model proved predictive, future
research could expand the theoretical framework by incorporating complementary
variables that deepen our understanding of the phenomenon in AI-mediated
educational settings.
On the other hand,
ease of use also yielded high scores, indicating that students do not perceive
significant technical difficulties. This perception may be related to the
demographic profile of the sample: mostly young people who already interact
with technology on a daily basis. However, the use of AI is still occasional
for many of them. This behavior reveals something that, as a teacher, can be
frequently observed: students recognize the potential of these tools but have
not yet managed to integrate them systematically into their learning. In this
regard, ease of use alone does not guarantee consistent use; it requires
support, usage models, and spaces for guided practice.
In particular, given
that this is a public university, factors such as access to stable
connectivity, availability of personal devices, and the conditions of computer
labs may indirectly influence the perception of ease of use (PEOU). Although
this study did not specifically measure these variables, the previously cited
studies on technology integration in higher education (Jara Alcívar, 2024;
Torres et al., 2024; Michilena et al., 2025) suggest that infrastructure and
institutional support influence the student’s technological experience.
Therefore, the observed ease of use cannot be understood solely as an intrinsic
characteristic of the tool, but also as the result of the educational
environment in which it is implemented.
Intention to use was
also high, indicating that students are willing to continue using AI. However,
this willingness does not necessarily imply that they will integrate it into
their academic routine in a critical or responsible manner. Here an important
nuance emerges: AI is perceived as useful and easy; however, its adoption
depends on other factors not accounted for by the TAM model. For example,
digital self-efficacy, intellectual curiosity, the level of trust in
AI-generated information, or even the teacher’s perception of its use. These
elements can influence how and to what extent students decide to incorporate AI
into their learning. Authors such as Ursavaş et al. (2025) and Wu et al. (2024)
have highlighted the importance of these additional variables, which could
deepen our understanding of the phenomenon.
Another relevant
aspect is the Ecuadorian context, where the integration of advanced
technologies into higher education is in the process of consolidation. Although
institutional initiatives exist, their pedagogical use has not yet been fully
incorporated (Jara Alcívar, 2024; Torres et al., 2024; Michilena et al., 2025).
In this context, this study demonstrates that Ecuadorian students are open to
this innovation, which represents an opportunity for universities seeking to
strengthen English language instruction with digital resources. However, this
openness must be accompanied by teacher guidance, because students do not
always know how to distinguish between appropriate use and excessive
dependence.
Nevertheless, the
results must be interpreted with certain methodological limitations in mind.
First, non-probabilistic convenience sampling was used, which limits the
generalizability of the findings to other university populations. Second, the
research was conducted at a single institution of higher education and with
students at A2 and B1 levels, which limits its extrapolation to other academic
contexts or levels of language proficiency. Furthermore, the cross-sectional
and correlational design allowed for the identification of significant
associations between variables but did not establish causal relationships among
them. These limitations do not invalidate the results obtained, but they do
suggest the need to expand the sample and methodological scope in future
research.
Finally, the analysis
of the results suggests that artificial intelligence has the potential to
become a valuable ally in English language learning, provided that its
integration is carried out responsibly. To achieve this, institutions must
promote training that not only teaches how to use these tools but also fosters
critical thinking, digital ethics, and the mindful use of technology. This will
enable students not only to accept AI but also to learn how to use it as a
meaningful support in their educational journey.
Conclusions
This study
demonstrated that university students at the A2 and B1 levels at the University
of Guayaquil hold a positive perception of the use of artificial intelligence
as a tool to support English language learning. The dimensions evaluated using
the technology acceptance model—namely, perceived usefulness, ease of use, and
intention to use—showed high averages and significant correlations with one
another, supporting the model’s validity in Ecuadorian higher education
contexts.
Among the most notable
findings is that perceived usefulness was the highest-rated dimension. This
suggests that students recognize the practical benefit of artificial
intelligence tools, especially when these allow them to efficiently complete
tasks such as writing, proofreading, or practicing vocabulary. Although they
also consider these tools to be accessible and easy to use, their use remains
occasional, indicating that the adoption of these technologies requires more
systematic guidance.
The results also
highlight that while the intention to use is high, it may be influenced by
factors not included in the TAM model. These include trust in information
generated by artificial intelligence, the level of digital literacy,
technological self-efficacy, and the teacher’s attitude toward the use of these
tools. Incorporating these variables into future studies would allow for a more
comprehensive understanding of the processes of technological acceptance in
educational settings.
Ultimately, artificial
intelligence represents a valuable opportunity to strengthen English language
learning, provided that its implementation is accompanied by institutional
policies that promote teacher training, critical digital literacy, and the pedagogical
and ethical use of technology. Promoting these conditions will enable students
not only to adopt these tools but also to integrate them consciously and
strategically into their academic education.
..........................................................................................................
References
Alotaibi, H. M.,
Sonbul, S. S., & El-Dakhs, D. A. (2025). Factors influencing the
acceptance and use of ChatGPT among English as a foreign language learners in
Saudi Arabia. Humanities and Social Sciences Communications, 12(1),
628. https://doi.org/10.1057/s41599-025-04945-2
Annamalai, N., Eltahir, M. E., Zyoud, S. H., Soundrarajan, D., Zakarneh, B.,
& Al Salhi, N. R. (2023). Exploring English language learning via Chabot: A
case study from a self determination theory
perspective. Computers and Education: Artificial Intelligence, 5,
100148. https://doi.org/10.1016/j.caeai.2023.100148
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and
user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
Ekizer, F. N. (2025). Exploring the impact of
artificial intelligence on English language teaching: A meta-analysis. Acta Psychologica, 260, 105649. https://doi.org/10.1016/j.actpsy.2025.105649
Guzmán Alvarado, M.
V., & Naranjo Andrade, S. S. (2025). The Impact of Artificial
Intelligence on English Language Learning: A Systematic Review of Tools,
Methods, and Outcomes in Language Skills. Runas. Journal of Education and
Culture, 6(12), e250287. https://doi.org/10.46652/runas.v6i12.287
Huang, J., & Mizumoto, A. (2024). Examining the relationship between
the L2 motivational self system and technology
acceptance model post ChatGPT introduction and utilization. Computers and
Education: Artificial Intelligence, 7, 100302. https://doi.org/10.1016/j.caeai.2024.100302
Jara Alcivar, C. W.
(2024). Aplicaciones de inteligencia artificial (IA) en el contexto educativo
ecuatoriano: Retos y desafíos. Ciencia Latina: Revista Multidisciplinar,
8(3), 7046-7060. https://dialnet.unirioja.es/servlet/articulo?codigo=9787164
Kanont, K., Pingmuang, P., Simasathien, T., Wisnuwong, S., Wiwatsiripong, B., Poonpirome,
K., Songkram, N., & Khlaisang,
J. (2024). Generative-AI, a Learning Assistant? Factors Influencing Higher-Ed
Students’ Technology Acceptance. Electronic Journal of E-Learning, 22(6),
18-33. https://doi.org/10.34190/ejel.22.6.3196
Mehrvarz, M., Salimi, G., Abdoli, S., & McLaren, B. M. (2025). How
does students’ perception of ChatGPT shape online learning engagement and
performance? Computers and Education: Artificial Intelligence, 9,
100459. https://doi.org/10.1016/j.caeai.2025.100459
Michilena, J.,
Jaramillo, M., Arguello, K. D. L. Á., Arteaga, L., & Saritama, J. (2025).
Inteligencia artificial en la educación ecuatoriana: Oportunidades y desafíos
para los docentes : Artificial intelligence in ecuadorian
education: opportunities and challenges for teachers. Revista
Multidisciplinar de Estudios Generales, 4(4), 75-90. https://doi.org/10.70577/reg.v4i4.297
Mittal, N., Batra,
G., & Sijariya, R. (2025). Understanding AI Adoption in Higher Education: A
Systematic Review of Technology Acceptance Model, Technology Readiness Index,
and the Integrated Technology Readiness and Acceptance Model. Asian Journal
of Research in Computer Science, 18(7), 186-209. https://doi.org/10.9734/ajrcos/2025/v18i7729
Moradi, H. (2025). Integrating AI in higher education: Factors
influencing ChatGPT acceptance among Chinese university EFL students. International
Journal of Educational Technology in Higher Education, 22(1), 30. https://doi.org/10.1186/s41239-025-00530-4
Peña-Acuña, B., &
Corga Fernandes Durão, R. (2024). Learning English as a second language with
artificial intelligence for prospective teachers: A systematic review. Frontiers
in Education, Volume 9-2024. https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2024.1490067
Torres, L. M. G.,
Castro, A. E. P., Pita, A. R. L., & Castro, M. M. P. (2024). Innovación
educativa: El impacto de la inteligencia artificial en el aprendizaje en la
educación en Ecuador.: Educational innovation: the impact of artificial
intelligence on learning in Ecuadorian education. Revista Científica
Multidisciplinar G-nerando, 5(2), ág. 2172-2188. https://doi.org/10.60100/rcmg.v5i2.357
Ursavaş, Ö. F.,
Yalçın, Y., İslamoğlu, H., Bakır-Yalçın, E., & Cukurova, M. (2025). Rethinking the
importance of social norms in generative AI adoption: Investigating the
acceptance and use of generative AI among higher education students. International
Journal of Educational Technology in Higher Education, 22(1), 38. https://doi.org/10.1186/s41239-025-00535-z