NEYART Magazine  
ISSN: 2992-7161  
AN ASSESSMENT OF CHATGPT USE AMONG FIRST-SEMESTER  
COLLEGE STUDENTS  
EVALUACIÓN DEL USO DE ChatGPT EN ESTUDIANTES  
UNIVERSITARIOS DE PRIMER SEMESTRE  
Velázquez Nava Oscar Guillermo  
TecNM/Chihuahua Institute of Technology II  
https://orcid.org/0009-0006-1175-1303  
oscar.vn@chihuahua2.tecnm.mx  
García Guerrero Braulio René  
TecNM/Chihuahua II Institute of Technology  
https://orcid.org/0000-0003-1285-2924  
braulio.gg@chihuahua2.tecnm.mx  
Martínez Vela Carlos Alberto  
TecNM/Chihuahua II Institute of Technology  
https://orcid.org/0009-0000-4374-2838  
carlos.mv@chihuahua2.tecnm.mx  
Reyes Ledezma Daniela María  
TecNM/Chihuahua II Institute of Technology  
https://orcid.org/0009-0001-1885-6312  
danielamaria.rl@chihuahua2.tecnm.mx  
Villar Laguna Víctor  
Instituto Politécnico Nacional  
https://orcid.org/0000-0001-5449-5238  
vick_villar@hotmail.com  
DOI: https://doi.org/10.61273/neyart.v4i1.209  
| Received: 23/02/2026 | Accepted: 30/04/2026 | Published: 02/06/2026  
This work is licensed under  
an international  
Creative Commons Attribution 4.0.  
Page 368  
NEYART Magazine  
ISSN: 2992-7161  
Abstract--This study evaluated the use, perceptions, and academic impact of ChatGPT among forty-four  
first-semester college students using a positivist quantitative approach. These results showed an overall  
low-to-moderate adoption profile consisting of limited knowledge and mostly expressive rather than  
academic use. The majority of academic uses included text translation and understanding content  
presented in class, indicating that students primarily use the tool for lower-order cognitive tasks.  
However, even with this early use of generative artificial intelligence, we observe significant ethical  
scrutiny through the lens of educational integrity from participating students expressing critical reflection.  
As for the downside, distraction while studying and an unwillingness to do their share were among the  
top issues identified, with low scores for performance decrease. To conclude, the major transition to and  
workload of higher learning lauds a time where adaptability and vulnerability is arguably at its peak.  
Consequently, explicitly pedagogic strategies and digital literacy programs should be a focus in the first  
years of higher education, as they have shown to foster critical skills, above all original thinking and an  
ethical, responsible and educational use of these technologies.  
Keywords-- Academic adaptation, ChatGPT, Higher education, Academic integrity, Artificial  
intelligence.  
Abstract-- Este estudio cuantitativo analizó un total de cuarenta y cuatro estudiantes del primer semestre  
para comprender el uso, las percepciones y el impacto académico de ChatGPT. El análisis reveló un uso  
predominantemente moderado a bajo (conocimiento y uso ocasional), principalmente para fines  
personales. Los casos de uso más citados fueron la traducción y la comprensión, que probablemente sean  
tareas simples de realizar a nivel cognitivo. Se plantearon preocupaciones sobre la integridad académica  
por parte de los participantes. Señalaron la distracción y la disminución del trato personal entre los  
aspectos negativos, mientras que la reducción del rendimiento fue mínima. Finalmente, el ingreso reciente  
a la educación superior contextualiza un período de transición importante y de alto riesgo. Por esta razón,  
es urgente que, en las instituciones de educación superior, que son responsables desde hace algún tiempo  
de forzar la exhibición y el uso de estas herramientas junto con su comunidad, incluyendo metodologías  
pedagógicas específicas desde el primer ciclo escolar para promover un aprendizaje crítico y una  
imaginación original con una enseñanza ética y responsable para un mejor uso con fines educativos.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 369  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
Keywords-- Adaptación académica, ChatGPT, Educación superior, Integridad académica, Inteligencia  
artificial.  
INTRODUCTION  
In recent years, the advent and consolidation of generative artificial intelligence has completely reshaped  
the higher education landscape. In particular, recently released ChatGPT—an OpenAI language model,  
publicly available since November 2022 has had a disruptive impact on university academic environments  
and provoked expectations and questions about its teaching/learning use as well as pedagogical, ethical  
and evaluation ramifications. This tool embodies an opportunity for support with academic adaptation and  
integrit but also a risk to the development of fundamental competencies — critical thinking, academic  
writing, intellectual autonomy for students transitioning into their first semester at college. This research  
assesses the implementation, perceptions and effects of ChatGPT on this particular cohort within the  
university community, providing empirical data to guide evidence-based decision-making at both  
institutional and pedagogical level.  
International Background  
From an international perspective, the use of ChatGPT in university settings has been the subject of  
numerous studies documenting both its benefits and risks. Below, we present the most relevant findings  
identified in the international scientific literature.  
Ngo (2023) conducted pioneering research with 200 university students in Vietnam, combining online  
surveys and semi-structured interviews to explore perceptions regarding the use of ChatGPT in learning.  
The results revealed that, in general, students have a favorable attitude toward the tool, valuing its ability  
to support academic tasks, although they also acknowledge barriers related to the reliability of information  
and ethical implications. This study was one of the first to provide systematic empirical evidence on  
student perceptions of ChatGPT in the university educational setting.  
Ravšelj et al. (2025) conducted one of the most extensive studies to date on students’ perceptions of  
ChatGPT in higher education, with a multinational sample involving universities from multiple countries  
and continents. Their results show that student attitudes towards this technology differ widely according  
to cultural context, discipline, and prior exposure getting experience of artificial intelligence tools so their  
recommendations therefore stress that any strategies or institutional policies around the use of AI need to  
be contextualized.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 370  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
Li et al. (2024) performed a scoping review of ethical issues related to the application of ChatGPT in  
higher education literature written in English, Chinese and Japanese. They found that the key issues were  
with regards to academic integrity, reshaping assessments and data privacy. Their findings highlight the  
need for empirical studies that will help to build institutional policies governing the responsible use of  
generative AI in university classrooms.  
Nemt-allah et al. (2024) developed and validated a scale to measure ChatGPT use among graduate  
students offering a tool with sound psychometric properties for future research in the space. The results  
indicate that the swift commercial uptake of ChatGPT by academic staff leads to valid concerns about  
dependence on AI-generated content, originality, and integrity in scientific knowledge production.  
Dogaru et al. (2025) gap on perception of ChatGPT's influence in academic learning, finding that while  
personalized learning is enabled by ChatGPT, its overuse indicates a reliance on technology and may  
hamper proper academic citation practices. The study calls for institutional policy around AI literacy to  
inspire critical thinking and academic integrity in the context of technology.  
At Australian universities, Sullivan, Kelly and McLaughlan (2023) explored what ChatGPT means for  
academic integrity and student learning. Their results suggest that the tool contradicts age-old ways of  
assessing students, and recommend that universities rethink how they evaluate learning since more  
complex tasks that we engage at university are not ideal for generative AI style completion.  
Latin American Context  
Academic studies on the use of ChatGPT at universities in Latin America have been highly inclusive  
since 2023, taking into account the unique sociocultural, technological and institutional contexts specific  
to the region.  
García Sánchez (2023) researched the experience and perception of higher education students in Mexico  
on the use of ChatGPT. The findings revealed that students utilize the tool as an extra resource in their  
academic tasks — most of which are fact-finding and text-generating in nature — but express uncertainty  
about the accuracy and reliability of generated texts. In this sense, it presupposes the most basic work in  
the Mexican context on how generative AI affects academic practices in university spaces.  
In line with this direction, Morales Caluña and Cervantes Díaz (2024) examined the perception and use  
of ChatGPT among Ecuadorian university students through Multiple Correspondence Analysis (MCA)  
to uncover behavioral patterns. They confirmed strong student interest in employing ChatGPT for  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 371  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
academic purposes, but described a significant hesitancy among students to admit that they were even  
using it. The study demonstrates that the motivation to use this tool transcends demographics including  
gender.  
A recent study by Diaz-Leon and Iraola-Real (2024), investigated the actual use and intention to use  
ChatGPT among Peruvian university students, providing empirical data in a Latin America setting where  
little research has been conducted combined with low prior activity on the issue. The findings indicate  
that the use of ChatGPT has progressively increased within higher education contexts in Peru, and factors  
such as perceived usefulness and ease of use predict significant intention to continue using the tool for  
academic uses.  
In a phenomenological qualitative study, Loayza-Maturrano (2024) explored the use of ChatGPT in  
academic writing by university students in Lima (Peru), collecting data via semi-structured interviews  
with faculty members, students and experts. Their findings show that, although the tool may help improve  
fluency and textual structure in academic writing, unsupervised use undermines the cultivation of  
communication skills that are fundamental to a university education.  
Soto Ortiz and Reyes Flores (2024) conducted research in Mexico aimed at describing university  
students’ perceptions of ChatGPT use, considering both its positive aspects—time optimization, support  
in completing assignments, and perceived improvement in academic performance—and negative aspects,  
notably the risk of dependency and the inhibition of critical and creative thinking.  
Horizontes (2025) published a quasi-experimental study conducted with 236 incoming students at the  
School of Engineering of a public university in Lima, Peru, during the year 2024. The study evaluated  
the impact of ChatGPT use on first-year university students, concluding that, although the freshmen  
already had prior knowledge of the tool, a specific training intervention significantly increased their  
responsible academic use of it and their information literacy skills. It highlights the importance of  
specifically studying first-year students, who represent the group most vulnerable to the unregulated use  
of AI.  
Study Objective  
The overall objective of this study is to evaluate the use, perceptions, and academic impact of ChatGPT  
on first-semester college students, identifying patterns of use, attitudes toward the tool, and its potential  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 372  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
effects on academic performance, integrity in assessments, and the development of competencies specific  
to higher education.  
Rationale for the Study  
The rationale for this study is grounded in four fundamental dimensions: theoretical, practical, social, and  
institutional.  
Theoretically, the research provides new insights into the role of generative artificial intelligence  
technologies in higher education - a very dynamic field with many empirical gaps to fill and particularly  
for students early in their university career. The literature review indicates that most of the published  
studies to date have focused on advanced-level or graduate students, and failed to attend to the experiences  
of first-year students, who are at a stage in their academic careers where they are forming crucial academic  
habits..  
From a practical standpoint, the findings of this research will offer empirical evidence to higher education  
institutions on policies for the responsible use of ChatGPT, new assessment practices that reinforce critical  
thinking skills and artificial intelligence literacy programs focused on first year students.  
At the same time, on a social front, ChatGPT access and proficiency among adolescence demand  
immediate consideration around equity in education, as students utilizing this tool differently may widen  
or narrow performance gaps. In order to provide a top-notch, equitable university experience, it is  
important to understand how first-semester college students (a time of high vulnerability and adjustment)  
use such technology.  
Third, from the institutional perspective, universities must update their regulatory and pedestrian  
frameworks to be able to catch up with an emergence of generative AI. This study gives local evidence  
that can help higher education institutes find a balance between readiness for technological change,  
preservation of academic integrity standards and enhancement of core competencies in universities.  
Study Limitations  
Like any research effort, this study acknowledges the existence of limitations that frame the scope and  
interpretation of its results:  
• Social desirability bias. Given that the use of ChatGPT may be perceived as questionable in terms  
of academic integrity, there is a possibility that participants may tend to underestimate the  
frequency or actual purposes of its use, affecting the validity of self-reports.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 373  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
• Temporal scope. The academic period which this study goes with is limited. As ChatGPT rapidly  
continues to evolve technologically, the implications for institutional and student attitudes may  
lead to this being a short-lived finding.  
• Contextual specificity. The findings generated cannot possibly generalise to other higher  
education Institutions with different socio-economic, cultural or pedagogical profiles from the one  
in which this research took place.  
DEVELOPMENT  
Research Paradigm  
This study is framed within the positivist paradigm, which conceives of reality as objective, measurable,  
and capable of being understood through empirical methods and systematic procedures. From this  
perspective, scientific knowledge is constructed through controlled observation, quantitative  
measurement, and hypothesis testing, separating the researcher’s subjective judgments from the  
phenomenon under study (Ricoy, 2006). This paradigm is relevant to the present study given that it seeks  
to identify patterns of ChatGPT use in a defined student population, based on quantifiable and  
generalizable data.  
Creswell (2014) notes that the positivist paradigm—or postpositivist in its contemporary version—  
grounds quantitative studies in deterministic assumptions, oriented toward variable reduction, the  
formulation of specific questions and hypotheses, and the use of standardized measurement instruments.  
Under this approach, the results obtained from representative samples allow for the identification of trends  
and comparisons in educational contexts  
Research Approach  
The approach adopted is quantitative, which, according to Hernández-Sampieri et al. (2014), is  
characterized by data collection using standardized instruments, analysis through statistical methods, and  
the search for patterns that allow for the description or explanation of the phenomenon under study. This  
represents an appropriate way of assessing ChatGPT usage by first-year collegiate students such that the  
frequency, modes, and participant justifications could be quantified in a systematic and replicable  
manner.  
Consistent with the above, García Sánchez (2023) conducted a descriptive-quantitative study on the use  
and perception of ChatGPT in higher education students, using questionnaires on a Likert scale as the  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 374  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
primary research tool. This methodological strategy is consistent with the objectives of the present study,  
as it allows for the collection of comparable and representative data from the student population under  
analysis.  
Participants  
The sample consisted of 44 first-semester students from a higher education institution, of whom 29 were  
men (65.9%) and 15 were women (34.1%), aged between 17 and 20 years. Participants were selected  
using non-probabilistic convenience sampling, taking into account the group’s accessibility and  
homogeneity in academic level (first semester). All students present at the time of the survey  
administration who voluntarily agreed to participate, following informed consent, were included. Those  
who did not complete the entire survey were excluded  
Instrument  
The study described in this document employed a structured set of instruments to collect  
sociodemographic and technical information on the use of artificial intelligence. A detailed overview of  
the instrument used is presented below:  
Instrument Technical Specifications  
• Name: Survey on the Use of AI Tools in Higher Education.  
• Structure: It consists of a personal data form and a main questionnaire with 25 items.  
• Validation: The instrument was validated by the Ethics and Research Committee of the Mazatlán  
School of Nursing, and its internal consistency was measured using Cronbach’s alpha.  
Sections of the Instrument  
The questionnaire is organized into four key categories to measure students’ interaction with ChatGPT:  
1. Sociodemographic Form (7 items): Collects variables such as age, gender, marital status,  
occupation, monthly income, and grade point average.  
2. Knowledge and Use (4 items): Measure of familiarity with the tool; response options; not at all  
familiar, somewhat familiar, very familiar  
3. ChatGPT Usage (5 questions): Type of use (personal, professional or academic) + Morality and  
Job los.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 375  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
4. Academic Use (11 items): Explores if the technology helps students track what they are learning,  
recall key concepts, and organize their ideas – things we should be concerned with when assessing  
a tech tool  
5. Negative effects (5 items): Assesses distraction, academic performance decline and poor sleep  
routine.  
Measurement Scale  
For the categories of management, academic use, and effects, a 5-point Likert scale was used:  
• (5) Always.  
• (4) Often.  
• (3) Sometimes.  
• (2) Almost never.  
• Never.  
This methodology allowed the researchers to perform descriptive and inferential statistical analyses using  
SPSS software version 25.00.  
Procedure  
The study was conducted in three phases.  
In the first phase (instrument design and validation), a self-developed questionnaire was created,  
structured into two sections: the first containing items on sociodemographic data (gender, age, major)  
and the second containing items on a five-point Likert scale (1 = Never, 5 = Always) to explore the  
frequency of ChatGPT use, the academic purposes for which it is used, perceptions of its usefulness, and  
concerns associated with its use. The instrument underwent expert review (by three faculty researchers  
in the field of educational technology), and Cronbach’s alpha was subsequently calculated to verify its  
internal consistency prior to final administration.  
In phase 2 (instrument administration), participants were informed about the purpose of the study,  
confidentiality of their responses and other issues before agreeing to participate. Throughout normal class  
hours, teacher-directed settings were utilized to administer the questionnaire in person (same formulation  
and conditions for how it was administered) to each participant with instructor approval.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 376  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
During phase three (processing of the data), the data was inputted into a digital base for later statistical  
processing. Review of the integrity of each questionnaire was carried out to remove missing or conflicting  
responses in all questionnaires completed before proceeding with analysis.  
Data Analysis  
Data analysis was performed using the Statistical Package for the Social Sciences (SPSS), version 25.  
Firstly was performed a descriptive analysis, with absolute and relative frequencies, measures of central  
tendency (arithmetic mean) and measures of dispersion (standard deviation) for each item of the  
questionnaire and for the sociodemographic variables. This enabled us to describe the profile of the  
participants and characterize general patterns of ChatGPT use.  
Data Analysis  
Sociodemographic Characteristics of the Sample  
The sample consisted of a total of 44 first-semester college students. Regarding the distribution by sex,  
29 participants were male (65.9%) and 15 were female (34.1%), as shown in Table 1.  
Table 1. Distribution of participants by gender.  
Gender  
Male  
Frequency  
Percentage (%)  
29  
15  
44  
65.9  
34.1  
Female  
Total  
100.0  
Source: Author’s own calculations.  
In terms of age, the most common ages were 18 – 19 years which reflects that majority profile of students  
entering higher education. The majority of participants were single, without financial livestock  
dependents and had low to middle monthly animal family incomes. The reported grade point average was  
7.0 to 8.4 in the first year, equivalent to an intermediate level of student achievement among this  
population of new college entrants.  
Knowledge and Use of ChatGPT  
Regarding the level of knowledge and familiarity with ChatGPT, the results reveal a picture of nascent  
use. As shown in Table 2, 47.7% of students reported having “little” knowledge of the tool, 29.5%  
indicated they knew “nothing” about it, and only 22.7% stated they knew “a lot” about it. These data  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 377  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
suggest that, as these are first-year students, formal exposure to artificial intelligence tools within the  
academic setting is still limited.  
Table 2. Level of knowledge about ChatGPT.  
A lot  
Little  
None  
Total  
10  
21  
13  
44  
22.7  
47.7  
29.5  
100.0  
Source: Author’s own work.  
Regarding frequency of use, 52.3% of students indicated that they use the tool occasionally, 27.3% have  
never used it, and only 20.5% reported frequent or regular use. These findings are consistent with the  
profile of students who are in the process of adapting to university life and have not yet systematically  
integrated AI tools into their educational process.  
Use of ChatGPT  
Regarding the type of use students make of ChatGPT, personal use predominated over academic and  
professional use. The overall average for this category was 2.8 out of 5 points on the Likert scale,  
indicating a low-to-intermediate level of proficiency, as shown in Table 3.  
Table 3. Using ChatGPT.  
Item  
Mean  
3.4  
DE  
Use for personal activities (entertainment, general information search)  
Use for academic purposes (homework, research, summaries)  
Use for professional or work purposes  
0.91  
1.02  
0.87  
1.14  
1.08  
2.7  
1.8  
Perception that ChatGPT could replace human labor  
Ethical concerns about the use of AI in education  
Overall average for the category  
2.9  
3.1  
2.8  
Source: Author’s own calculations.  
It is noteworthy that ethical concerns regarding the use of AI in education received an average score of  
3.1, indicating that students, even with limited use, are already reflecting on the implications of these  
technologies for their academic education.  
Academic Use of ChatGPT  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 378  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
The academic use category received an overall average of 2.6 points, placing it at a low-to-middle level  
on the scale. The items with the highest scores corresponded to support in text translation and  
understanding class topics, while items related to content retention and exam preparation received the  
lowest scores, as shown in Table 4.  
Table 4. Academic use of ChatGPT.  
Item  
Average  
3.0  
SD  
Comprehension of topics explained in class  
Organization of ideas for written assignments  
Searching for bibliographic references  
Answering questions outside of class hours  
Preparing summaries of long texts  
Exam preparation  
1.05  
1.10  
0.98  
1.03  
1.07  
1.01  
0.95  
1.12  
0.99  
1.04  
1.00  
2.9  
2.7  
2.8  
2.9  
2.3  
Retention of course content  
2.1  
Translation of texts into another language  
Participation in online activities or educational platforms  
Support with writing and proofreading  
Support in learning new subjects or disciplines  
Overall average for the category  
Source: Author’s own analysis.  
3.2  
2.4  
2.6  
2.5  
2.6  
The item with the highest score was the one related to text translation (M=3.2), which could indicate that  
students rely on the tool with greater confidence for specific linguistic tasks than for more complex  
cognitive processes associated with deep learning.  
Perceived Negative Effects  
In the negative effects dimension, the overall mean was 2.5 points, reflecting that, although adverse effects  
are not perceived with high frequency, there is a moderate presence of some of them in the study  
population, as shown in Table 5.  
Table 5. Perceived negative effects of using ChatGPT.  
Item  
Mean  
SD  
Distraction during academic activities  
Decline in academic performance  
2.8  
1.09  
0.97  
2.2  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 379  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
Impact on sleep habits  
2.6  
2.7  
2.3  
2.5  
1.11  
1.05  
1.00  
Reduced effort in tasks  
Feeling of dependence on the tool  
Overall category average  
Source: Author’s own calculations.  
The most relevant point on distraction while doing academic activities was rated highest (M=2.8),  
indicating that when the use of it occurs, it usually disrupts the concentration of the student. When it came  
to the decline in academic performance, that was rated the lowest of this category (M=2.2), confirming  
once again that usage is low.  
In total, the results reveal a relatively low-to-moderate profile of ChatGPT utilization among first-  
semester college students. This trend is largely due to where a majority of the respondents still are in their  
educational pathways: newly entering tertiary education comes with a period for adjusting and learning,  
which means knowledge of sophisticated technology tools takes time to implement. Nevertheless, the  
ethical reflections and the perception of effects such as distraction and less individual effort indicate key  
aspects to be addressed in pedagogical strategies that favor the use of artificial intelligence appropriately  
within the university context.  
Discussion  
The findings of this study reveal a low-to-moderate profile of ChatGPT use among first-semester  
university students, with an overall average of 2.6 out of 5 points for academic use and 47.7% of  
participants reporting “little” familiarity with the tool. These results are consistent with those reported by  
Horizontes (2025), who, working with 236 freshmen at a Peruvian public university, found that first-year  
students, while possessing prior knowledge of ChatGPT, have not yet developed a systematic and  
responsible academic use of it, highlighting precisely that this group is the most vulnerable to the  
unregulated use of AI.  
As for the main type of use, study participants made more use of the tool for personal activities (M=3.4)  
rather than strictly academic activities (M=2.7), which reflects a trend previously documented by García  
Sánchez (2023) in the Mexican context, identifying that university students were using ChatGPT mainly  
as supplementary information and writing tools, but with caution about their validity. Likewise, Morales  
Caluña and Cervantes Díaz (2024) in Ecuador observed that, while there is widespread motivation to use  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 380  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
the tool, students express some hesitation in openly acknowledging its academic use, which could explain  
the social desirability bias also present in the self-reports of the analyzed sample.  
Regarding negative effects, distraction during academic activities was the item with the highest score  
(M=2.8), a finding that aligns with Dogaru et al. (2025), who found that the use of ChatGPT in STEM  
settings correlates with technological dependence, and with Soto Ortiz and Reyes Flores (2024), who  
documented in Mexico that among the most frequent negative perceptions are the risk of dependence and  
the inhibition of critical thinking. However, the low score for decreased academic performance (M=2.2)  
is consistent with the still- r incipient use reported and with the observations of Nemt-allah et al. (2024),  
who note that the consequences on original intellectual production manifest primarily in more advanced  
phases of use.  
Finally, ethical concerns received an average score of 3.1, suggesting that, even with limited use, students  
are already reflecting on the implications of AI for their education. This is consistent with Li et al. (2024),  
whose comprehensive review identifies academic integrity as the central concern regarding the use of  
ChatGPT in higher education, and with Sullivan, Kelly, and McLaughlan (2023), who note that this tool  
challenges traditional models of university assessment, making it urgent to reformulate pedagogical  
strategies that promote the responsible use of AI from the very first semesters.  
Conclusions  
This study evaluated the use, perceptions, and academic impact of ChatGPT among 44 first-semester  
college students, revealing a low-to-moderate adoption profile consistent with the characteristics of this  
stage of education. The conclusions derived from each research objective are presented below.  
Concerning the frequency and for what purposes ChatGPT is being used, it is concluded that most  
students use ChatGPT from time to time and mainly privately; therefore, its academic usage has not been  
established yet. The most common academic uses involved text translation and understanding class  
content while exam preparation and content retention were the least likely suggesting that such tools will  
be used for broader yet lower order cognitive tasks.  
With regard to perceived benefits and risks, students moderately acknowledge the practicality of  
ChatGPT for academic purposes; however, they also express ethical concerns on its usage, which is a  
reflective and non-uncritical attitude towards this tool–in support of fostering responsible use practices  
at the start of their university trajectory.  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 381  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
As for perceived negative effects, although these occur at a low-to-moderate level, distraction and reduced  
personal effort were the most frequently cited effects, pointing to priority areas for attention in orientation  
and digital literacy programs aimed at first-year students.  
Finally, it is concluded that higher education institutions should incorporate explicit pedagogical  
strategies regarding the ethical and critical use of artificial intelligence starting in the first semester, before  
patterns of dependency or misuse become entrenched. The findings of this study provide local empirical  
evidence that can guide the design of context-specific institutional policies for the responsible use of  
ChatGPT in higher education.  
References  
Creswell, J. W. (2014). Research design: Qualitative, quantitative, and mixed methods approaches (4th  
ed.). SAGE Publications.  
Diaz-Leon, I., & Iraola-Real, I. (2024). The use of ChatGPT by Peruvian university students: An  
exploration of actual use and intention to use. Revista Ibérica de Sistemas e Tecnologias de  
Informação, E69, 89–101.  
Dogaru, M., Pisică, O., Popa, C.-Ș., Răgman, A.-A., & Tololoi, I.-R. (2025). The perceived impact of  
artificial intelligence on academic learning. Frontiers in Artificial Intelligence, 8, Article 1611183.  
https://doi.org/10.3389/frai.2025.1611183  
García Sánchez, O. V. (2023). Use and perception of ChatGPT in higher education. Journal of Research  
in Information Technologies, 11(23), 98–107. https://doi.org/10.36825/RITI.11.23.009  
Hernández-Sampieri, R., Fernández-Collado, C., & Baptista-Lucio, M. P. (2014). Research Methodology  
(6th ed.). McGraw-Hill Interamericana.  
Li, M., Enkhtur, A., Cheng, F., & Yamamoto, B. A. (2024). Ethical implications of ChatGPT in higher  
education: A scoping review. Journal of Interdisciplinary Studies in Education, 13(1), 55–68.  
https://doi.org/10.48550/arXiv.2311.14378  
Loayza-Maturrano, E. F. (2024). University students’ perceptions of the use of ChatGPT in academic  
writing. Educare et Comunicare, 12(2), 28–38. https://doi.org/10.35383/educare.v12i2.1195  
Morales Caluña, E. R., & Cervantes Díaz, D. J. (2024). University students’ perceptions and use of  
ChatGPT.  
Ciencia  
Latina  
Revista  
Científica  
Multidisciplinar,  
8(5),  
6872–6886.  
https://doi.org/10.37811/cl_rcm.v8i5.14101  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 382  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
Nemt-allah, M., Khalifa, W., Badawy, M., Elbably, Y., & Ibrahim, A. (2024). Validating the ChatGPT  
Usage Scale: Psychometric properties and factor structures among postgraduate students. BMC  
Psychology, 12(1), Article 576. https://doi.org/10.1186/s40359-024-01983-4  
Ngo, T. T. A. (2023). The perception by university students of the use of ChatGPT in education.  
International Journal of Emerging Technologies in Learning (iJET), 18(17), 4–19.  
https://doi.org/10.3991/ijet.v18i17.39019  
Ravšelj, D., Keržič, D., Tomaževič, N., Umek, L., Brezovar, N., Iahad, N. A., ... Aristovnik, A. (2025).  
Higher education students' perceptions of ChatGPT: A global study of early reactions. PLOS ONE,  
20(2), Article e0315011. https://doi.org/10.1371/journal.pone.0315011  
Ricoy, C. (2006). Contribution on research paradigms. Educação. Revista do Centro de Educação, 31(1),  
11–22. https://www.redalyc.org/pdf/1171/117117257002.pdf  
Soto Ortiz, J. L., & Reyes Flores, I. A. (2024). University students’ views on the use of ChatGPT. REPED.  
Paraguayan Journal of Distance Education, 5(2), 56–65. https://doi.org/10.56152/reped2024-  
dossierIA1-art5  
Sullivan, M., Kelly, A., & McLaughlan, P. (2023). ChatGPT in higher education: Considerations for  
academic integrity and student learning. Journal of Applied Learning and Teaching, 6(1), 31–40.  
Tello-Gamarra, J., & Torres-Sifuentes, M. (2025). Generative artificial intelligence among incoming  
students at a public university [Generative AI and incoming students]. Horizontes. Journal of  
Research in Education Sciences, 9(35)  
Collaborative Work Table  
Role  
Author(s)  
Conceptualization  
Methodology  
Oscar Guillermo Velázquez Nava  
García Guerrero Braulio René, Reyes Ledezma Daniela  
María  
Software  
Martínez Vela Carlos Alberto, Villar Laguna Víctor  
Oscar Guillermo Velázquez Nava  
García Guerrero Braulio René, Reyes Ledezma Daniela  
María  
Validation  
Formal Analysis  
Research  
Martínez Vela Carlos Alberto, Villar Laguna Víctor  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 383  
Original Research Article  
NEYART Magazine  
ISSN: 2992-7161  
Resources  
Oscar Guillermo Velázquez Nava  
Data Curation  
García Guerrero Braulio René, Reyes Ledezma Daniela  
María  
Writing - Preparation of the original  
draft  
Martínez Vela Carlos Alberto, Villar Laguna Víctor  
Writing - Review and editing  
Visualization  
Oscar Guillermo Velázquez Nava  
García Guerrero Braulio René, Reyes Ledezma Daniela  
María  
Supervision  
Martínez Vela Carlos Alberto, Villar Laguna Víctor  
Oscar Guillermo Velázquez Nava  
García Guerrero Braulio René, Reyes Ledezma Daniela  
María  
Project Management  
Fundraising  
Issue 4 | Vol. 4 – No. 1 | January – June 2026 |  
Page 384  
Original Research Article