Student Perspectives on Ethical Academic Writing with ChatGPT:

An Empirical Study in Higher Education

Lukas Spirgi

, Sabine Seufert

, Jan Delcker

and Joana Heil

Institute for Educational Management and Technologies, University of St. Gallen, Switzerland

Chair for Economic and Business Education – Learning, Design & Technology, University of Mannheim, Germany

Keywords: Human-AI-Collaboration, AI Ethics, ChatGPT, Academic Writing.

Abstract: The emergence of ChatGPT has significantly reshaped the landscape of higher education, sparking concerns

about its potential misuse for academic plagiarism (Cotton et al., 2023). This study examines the use of

ChatGPT in academic writing among students at the University of Mannheim in Germany and St. Gallen in

Switzerland, using a proposed Human-AI collaboration framework with six levels of AI-enabled text

generation (Boyd-Graber et al., 2023). The survey of 699 students reveals varied ChatGPT usage across all

six levels, with Level 3 (Literature Search) being slightly more utilized. Students expressed mixed opinions

on ethical issues, such as the declaration of ChatGPT-generated content in academic work and the extent to

which ChatGPT is allowed at their university. The results of the study highlight students' concerns about

negative effects on grades, a lack of clarity about university policies on ChatGPT, and fears that hard work

will not be rewarded. Despite these issues, most students support open access to ChatGPT. The findings

suggest the need for clear ethical guidelines in academia regarding AI use and highlight the potential

stigmatization of AI, which could hinder technology acceptance and AI-related skills development.

1 INTRODUCTION

The emergence of ChatGPT indicates the

incorporation of Artificial Intelligence (AI) in

professional and educational settings. AI appears to

be having an escalating impact on people's lives due

to greater interactions between humans and robots

(Kim, 2022). AI in Higher Education has been used to

provide personalized feedback on academic writing

(Knight et al., 2020). The developments in the field

of generative AI (such as ChatGPT) are accelerating

the transformation in the area of knowledge work

(Dell'Acqua et al., 2023). Generative AI can be

defined, according to (Lim et al., 2023, p. 2), 'as a

technology that (i) leverages deep learning models to

(ii) generate human-like content (e.g., images, words)

in response to (iii) complex and varied prompts (e.g.,

languages, instructions, questions)'.

The effectiveness of this AI has led to widespread

apprehensions in higher education, especially

pertaining to the potential misuse by students for

https://orcid.org/0000-0002-3807-6460

https://orcid.org/0009-0003-7182-949X

https://orcid.org/0000-0002-0113-4970

https://orcid.org/0000-0001-5069-0781

plagiarism through the utilization of AI-generated

content in unmonitored academic tasks (Lo, 2023).

Consequently, discussions in the public domain

frequently emphasize the viewpoints of educators and

university administrations. To date, there is a

restricted amount of research on the application of AI

in higher education (Garrel et al., 2023; Kim, 2022;

Lim et al., 2023).

2 AI IN ACADEMIC WRITING

2.1 Framework for

Human-AI-Collaboration in

Academic Writing

Artificial intelligence tools for academic writing can

be described as human-like robots. Initially, the term

robot referred to appearance in the sense of physical

presence, but it is increasingly used to describe

Spirgi, L., Seufert, S., Delcker, J. and Heil, J.

Student Perspectives on Ethical Academic Writing with ChatGPT: An Empirical Study in Higher Education.

DOI: 10.5220/0012555700003693

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 2, pages 179-186

ISBN: 978-989-758-697-2; ISSN: 2184-5026

179

human-like performance (Murphy, 2019).

Human-like capabilities can be characterized by

the breadth and complexity of their functionalities

(Dang & Liu, 2022). In the field of AI-based text

generation in academic contexts, the model proposed

by Boyd-Graber et al. (2023) can serve as a reference

framework. The Association for Computational

Linguistics, an international research community

focused on language models such as ChatGPT, has

published guidelines for the ethical use of AI-based

writing tools (Boyd-Graber et al., 2023). Within these

guidelines, different levels can be defined that

indicate increasing levels of AI performance in text

generation, which affects the novelty of the content

generated.

Level 1: Assistance Purely With the Language

of the Paper. The AI assistant performs the task of

paraphrasing and refining the author's initial content.

The human carries out the final correction.

Level 2: Short-Form Input Assistance. The AI

assistant serves as a writing aid for brief texts, while

the human is accountable for examining the produced

text.

Level 3: Literature Search. The AI assistant acts

as a search tool, guiding the user while the human is

responsible for searching, reading, and discussing

references, similar to a typical search engine

(Alshami et al., 2023).

Level 4: Low-Novelty Text. The virtual assistant

is accountable for producing text that describes

widely accepted concepts or presents an automated

literature review summary. Subsequently, the human

reviewer is responsible for ensuring precision and

discerning whether to employ the generated text.

Level 5: New ideas. The AI assistant generates

research ideas and model results, while the human

develops these further by formulating theses for

discussion and defining the research problem.

Moreover, humans are tasked with searching for

reliable sources to support these ideas.

Level 6 New Ideas + New Text: The AI assistant

plays a dual role in generating and executing text,

whilst the human is responsible for verifying

accuracy and deciding whether to adopt the generated

text. In addition, the human is tasked with further

development, including formulating discussion

theses and defining research problems, as well as

searching for well-established sources to support

these ideas.

2.2 Ethical Guidelines for AI

Advancements in artificial intelligence present

significant opportunities and substantial challenges,

necessitating the ethical and responsible application

of AI. In their work, Bao et al. (2022) devised an

index to evaluate AI's potential advantages and risks.

The ethical application of AI is evidently of

paramount significance.

The ethical use of AI has led to the development

of various guidelines (Floridi & Cowls, 2019). Jobin

et al. (2019) conducted a meta-study examining and

comparing existing ethics guidelines for AI. They

created an overview of current principles and

guidelines for ethical AI to assess whether there is

global convergence in the principles of ethical AI and

the requirements for its implementation. Their

analysis revealed global alignment on five ethical

principles: 1) Transparency, 2) Justice and Fairness,

3) Data Protection and Privacy, 4) No Harm and

Solidarity and 5) responsible AI development.

The ethical guidelines emphasize the significance

of customizing them for specific AI systems and

application domains, as suggested by Jobin et al.

(2019). To tackle these issues competently, adopting

a particular perspective that resonates with the

respective stakeholder group is essential.

3 THE PRESENT STUDY

The objective of this investigation is to assess the

prevalence of using artificial intelligence tools for

academic writing. Additionally, this study aims to

scrutinize the ethical standards which are deemed

crucial by students. To explore the usage and ethics

of academic writing when employing ChatGPT, we

pose two research questions:

1. How frequently do students use ChatGPT for

the different levels according to the Human-AI-

collaboration framework in academic writing?

2. How do students perceive ethical guidelines for

the use of ChatGPT regarding transparency and

fairness?

The ethical principles developed by Jobin et al.

(2019) for the use of AI were applied in this study and

specifically adapted for higher education. The

perspective of students is relevant. Therefore, we

focused on two aspects:

Transparency: The passages created with these

tools are clearly marked as such. The declaration of

originality at the end of a written work is adjusted and

specifies the use of such tools (with the aim of

acknowledging the human's contribution to AI).

These are often new ethical standards at Higher

Education Institutions. Consequently, we asked about

whether students are afraid of lowered grades for

declaring the use of ChatGPT. The consequence of

CSEDU 2024 - 16th International Conference on Computer Supported Education

180

the ethical aspect might lead to unfair evaluation from

the student's perspective. Furthermore, we analyzed

the awareness of the extent to which the use of

ChatGPT is permitted at the university and how clear

the communication is for students.

Justice, Fairness, and Equality: Free access for all

learners to avoid social inequalities using AI is an

issue many Higher Education Institutions are thinking

about whether ChatGPT 4.0 should be offered. As a

possible consequence, we wanted to know from

students whether the use of ChatGPT at university

means that hard work is no longer rewarded.

Furthermore, the other way around, we asked how

students perceive if teachers correct with ChatGPT in

terms of unfair or fair evaluation.

4 METHODS

4.1 Online Survey and Sample

An online survey was chosen for the study to

comprehensively explore students' experiences with

AI. The survey was conducted digitally using the

'Qualtrics' platform from September to October 2023.

All questions were single-choice. In total, 699

students from the University of St. Gallen and the

University of Mannheim participated. The mean age

of the students surveyed was 21.4 years (SD = 2.94).

Students from different disciplines were surveyed at

both universities.

Table 1: Sample.

Characteristic Absolute Percentage

Female Students 348 49.8%

Male Students 341 48.8%

Diverse Students 10 1.4%

First Semester Students 317 45.4%

Bachelor Students 285 40.8%

Master Students 97 13.8%

University of St. Gallen 274 39.2%

University of Mannheim 425 60.8%

4.2 Development of Instrument

The questionnaire comprised two parts. In the first

part, two specific questions were formulated for each

level established in the theoretical framework

(Human-AI-collaboration) to assess usage intensity.

Respondents were prompted to rate their responses on

a seven-point scale, ranging from 1 (never) to 7

(always). The questionnaire explained in detail the

frequency of each choice. 'Never' (1) means that

ChatGPT is never used this way. 'Rarely' (2)

represents a use once per semester. 'Occasionally' (3)

means sporadic use, i.e. several times per semester.

'Sometimes' (4) means a of use about once a month.

'Frequently' (5) means using ChatGPT several times

a month in the defined way. 'Usually' (6) means once

a week. 'Always' (7) means constant use (several

times a week).

The study emphasized ethical considerations in

the second part, specifically transparency and

fairness. The choice to concentrate on these facets

arises from their pivotal importance for students.

Responses to ethical considerations were gauged

employing a five-point Likert scale, spanning from

'strongly disagree' (1) and 'disagree' (2) to 'neutral'

(3), 'agree' (4), and 'strongly agree' (5).

A total of 18 items were analyzed for this study.

5 RESULTS

5.1 Internal Consistency

The internal consistency of the constructed indices,

designed to assess the frequency of usage at each

level, was evaluated using Cronbach's Alpha

(Cronbach, 1951). Two questions were combined at

each level (1 – 6) to form an index, capturing the

nuances of usage patterns among university students.

Cronbach's Alpha is a measure of internal

consistency, reflecting the extent to which the items

within an index are correlated. The values obtained

for each index are all above 0.7, indicating an

acceptable to good level of internal consistency

(Cronbach, 1951). This suggests that the selected

items within each index reliably measure the intended

construct of usage frequency among university

students

Table 2 presents the computed indices for all six

levels, with the Cronbach's Alpha value. The index

calculated reflects the average usage by students.

Table 2: Frequency of use index.

Leve1 Index (SD) Cronbach's Alpha

2.56 (1.79)

α = 0.75

2 2.58 (1.77) α = 0.83

3 2.65 (1.79) α = 0.78

4 2.59 (1.74) α = 0.70

5 2.21 (1.58) α = 0.81

6 2.45 (1.69) α = 0.89

Student Perspectives on Ethical Academic Writing with ChatGPT: An Empirical Study in Higher Education

181

The index values have been calculated to range

from 2.21 to 2.65, signifying a frequency of use

between 'rarely' and 'occasionally'. The highest index

value is observed at Level 3, indicating that ChatGPT

is most used for literature searches.

Although the general average usage of ChatGPT

across all levels is low, the data suggests that a

significant number of respondents frequently use

ChatGPT for academic writing.

Table 3 illustrates the frequency of usage for

various scenarios, categorized from levels 1 to 6. The

table displays the percentage of respondents who

utilize ChatGPT in the described manner for each

defined type of use. The original 7-point scale has

been condensed into a 4-point to ensure clarity in the

table. Respondents who selected 'never' (1) in the

usage questionnaire are also represented as 'never' in

Table 3. 'Rarely' (2) and 'occasionally' (3) have been

merged into 'sporadically', indicating that ChatGPT is

used in this manner once or several times per

semester. Similarly, 'sometimes' (4) and 'frequently'

(5) are combined as 'often', indicating that ChatGPT

is used once to several times monthly. Those who

indicated 'usually' (6) and 'always' (7) are grouped as

'very often', indicating ChatGPT usage once or

several times a week. The most significant proportion

of students for all types of use is 'never', but there are

always at least 20% who 'usually' or 'mostly' use

ChatGPT in the ways described.

The two categories, 'generating keywords for

literature searches (brainstorming)' and 'using AI to

define terms and explain concepts', have the highest

proportion of students who say they use ChatGPT

often (28%) or very often (13%). This means that

these students use ChatGPT in the way described at

least once a week. The categories 'using AI for

concept development and design' and 'integrating AI-

generated concepts seamlessly into your text' have the

highest proportion of students who say they 'never'

use these methods (54% and 58%).

Table 3: Frequency of use of ChatGPT (N = 699).

Lev. Type of use of ChatGPT Never Sporadically Often Very often

- Spell and grammar check

39% 27% 23% 11%

- Translate text

46% 27% 19% 7%

- Develop coherent text based on provided keywords

34% 29% 25% 12%

- Apply the AI-corrected text directly in one's writing

53% 25% 16% 6%

- Generate keywords for literature searches (brainstorming)

34% 25% 28% 13%

- Identify pertinent literature sources with AI

51% 26% 18% 5%

- Utilise AI to define terms and explain concepts 33% 26% 28% 13%

- Incorporate AI-generated concepts seamlessly into your text

54% 26% 15% 5%

- Use AI for concept development and design 58% 23% 16% 4%

- Use AI for data analysis to generate new ideas

46% 29% 20% 5%

- Use AI to draft comprehensively on given topics and goals 42% 31% 19% 7%

- Enhance AI-generated drats with more precise prompts

46% 27% 21% 6%

Table 4: Ethical Aspects Transparency and Fairness (N = 699).

Lev. Item

Strongly

disa

ree

disagree neutral agree

Strongly

ree

Transparency

- In my opinion, ChatGPT should only be allowed if

the generated passages are marked as such.

6% 19% 34% 31% 11%

- I am afraid that the teachers will lower my work if I

declare that I use ChatGPT.

4% 11% 21% 42% 22%

- I am not currently aware of the extent to which the

use of ChatGPT is permitted at university

4% 15% 27% 39% 15%

Fairness

- In my opinion, open access to ChatGPT for all

learners is essential

4% 10% 29% 39% 17%

- In my opinion, using ChatGPT at university means

that hard work is no longer rewarded.

19% 34% 21% 20% 5%

- I would find it unfair if teachers corrected my work

using ChatGPT.

6% 19% 26% 30% 19%

CSEDU 2024 - 16th International Conference on Computer Supported Education

182

5.2 Ethical Aspects

Table 4 presents the students' views on specific

ethical aspects. The table shows the proportion of

students who agree or disagree with the statement.

The statements are divided into the criteria'

transparency' and 'fairness'.

All statements relating to the ethical aspect of

transparency are approved. This means that the

proportion of students who agree or strongly agree

with the statement is greater than the proportion of

students who disagree or strongly disagree with it.

The statement 'I am afraid that the teachers will lower

my work if I declare that I use ChatGPT' has the

highest agreement rate (agree = 42% and strongly

agree = 22%).

In the Fairnaiss category, two statements are more

likely to be agreed and one statement is more likely

to be disagreed. The statement 'In my opinion, the use

of ChatGPT at university means that hard work is no

longer rewarded' is the one that most strongly

disagrees with. 53% of respondents tend to disagree,

and only 5% strongly agree.

6 DISCUSSION

Students' average usage of ChatGPT is currently quite

diverse; most students follow low or medium

frequency. On the one hand, a subgroup of students

consistently refrain from using ChatGPT at every

level (between 20% to 41%). This may be attributed

to the relatively restrictive regulations imposed by

universities. Furthermore, students might be opting

for alternative AI tools like Deepl Write for assistance

in writing, which corresponds to a specific level of

usage (Level 2) in our framework. On the other hand,

a small segment of students (about 4-13%)

consistently utilize ChatGPT across all levels,

including the most advanced level, where ChatGPT

functions similarly to a co-author by generating new

ideas and text. This indicates a high frequency, almost

to the point of being a regular pattern or habit.

On average, students use ChatGPT the most at

Level 3 (Index 2.65). At this level, ChatGPT is

predominantly employed for keyword searches in

literature research. ChatGPT is particularly suitable

for brainstorming, as the factual accuracy of the

output is less critical than, for instance, when

explaining theories. There are no significant

differences in the frequency of ChatGPT usage across

the various levels of the Human-AI Collaboration

Framework.

When analyzing the data, it is noticeable that

some students (15%) do not use ChatGPT in any of

the usage scenarios described. This means that

despite the considerable hype surrounding AI text

generators, some students do not yet have confidence

in this new technology and do not use it.

The following three topics focus on the ethical

guideline 'transparency' and possible consequences

for students following this issue:

Marking ChatGPT Passages: This survey data

reflects a range of opinions on whether ChatGPT-

generated passages should be marked. While a

significant portion of respondents are neutral, there is

a notable presence of both agreement and

disagreement, suggesting a nuanced and mixed

viewpoint on this issue. Further research and context

may be needed to understand the reasons behind these

opinions and their potential implications. Some

students might believe marking is essential for

transparency and accountability, helping users

distinguish between human and AI-generated

content. Furthermore, marking could allow accurate

assessment of a student's own understanding.

Marking might empower users to make informed

decisions about engaging with AI-generated content.

On the other hand, opponents argue that marking

restricts creative freedom and experimentation with

AI tools. Concerns about grading or assessment

biases against AI-generated content may influence

opinions (see item: Influence on the grade). The

approving position is the most substantial group.

Concerns about the transparency of academic

accomplishments may arise, as it may become

difficult to distinguish between work produced solely

by students and work assisted by AI.

In discussions about technology and ethics,

neutrality can often be seen as a balanced and

cautious approach (Green, 2021). Respondents in the

neutral group may be taking a middle-ground

position, considering both the potential benefits and

concerns associated with marking AI-generated

content.

Fear of Lowered Grades for Declaring the Use

of ChatGPT: A substantial majority of respondents

express concerns about their work being negatively

affected by declaring the use of ChatGPT. This group

constitutes 65% of the total respondents (agree and

strongly agree) and is the highest value of all six

ethical topics. Some students may worry that using

ChatGPT could be viewed as a form of cheating or

academic dishonesty, which could result in penalties

or lower grades. Furthermore, students might be

concerned that teachers or evaluators could have

biases against AI-generated work, leading to unfair

Student Perspectives on Ethical Academic Writing with ChatGPT: An Empirical Study in Higher Education

183

assessment or grading. Worries about how disclosing

ChatGPT usage might affect teachers' perceptions of

students' capabilities and dedication to their work.

Educational systems often place high expectations on

students to excel. The fear of potentially lower grades

could add to the pressure students already feel.

Awareness of the Extent to Which the Use of

ChatGPT is Permitted at University: A significant

proportion of respondents (19%) indicate that they are

not aware of the extent to which the use of ChatGPT

is permitted at their university. Both universities

provide guidelines to the students on how ChatGPT

could be used for academic writing. However,

University policies on AI tool usage could be

complex and challenging to understand fully. The

policies have recently been introduced, giving

students insufficient time to become aware.

Furthermore, it could be an indicator that more than

communication is needed. Students should be

provided with training on responsible AI tool usage.

The ethical guideline 'fairness' is discussed with

the following three aspects:

Importance of Open Access to ChatGPT for all

Students: The data suggests a notable level of

support and a significant neutral stance towards

permitting ChatGPT usage at universities. While

there is some opposition, it is not the dominant

viewpoint. Optimistic respondents might view

ChatGPT as a valuable tool in academic studies.

ChatGPT can be tailored to individual needs,

allowing students to receive personalized assistance

and support in their coursework. Some students might

appreciate ChatGPT's ability to assist in improving

writing skills and generating content for assignments.

Supportive respondents may believe that exposure to

AI technology is essential for students to be prepared

for future career opportunities as AI becomes

increasingly prevalent in many professions.

ChatGPT Impact on Rewarding Hard Work:

About a quarter of the students (25%) express

agreement with the idea that ChatGPT usage may

reduce the rewards for hard work. This group believes

that technology may make it easier to achieve

academic success without putting in as much effort.

Some respondents may worry that using ChatGPT

could be seen as a form of academic dishonesty or

cheating, which could undermine the value of their

hard work. Concerns may arise about the fairness of

evaluating students when some have access to AI

tools that can generate high-quality content,

potentially giving them an advantage over those who

do not use such tools. Students who put significant

effort and time into their coursework may feel that the

availability of AI-generated content devalues their

hard work and dedication. There could be concerns

that AI-generated work might disrupt the meritocratic

nature of education, where success is traditionally

based on individual effort and abilities. Some

students may worry that relying on AI tools for

assignments could hinder the development of critical

thinking and problem-solving skills, which are

essential aspects of the learning process. There may

be concerns that students feel pressured to use AI

tools like ChatGPT to keep up with their peers, even

if they prefer not to. Additionally, some students may

worry that using AI tools could conflict with the

educational values of effort, learning, and personal

growth.

Unfairness if Teachers Correct With ChatGPT:

The data shows a wide range of opinions on whether

using ChatGPT to correct work is considered unfair.

This indicates that the topic of AI tool usage in

educational assessment is complex, and opinions vary

widely among respondents. The 'Agree' and 'Strongly

Agree' categories collectively make up 49% of

respondents, indicating that almost half of the

respondents find it unfair if teachers rely on ChatGPT

to correct their work.

Some students may believe that using ChatGPT

for corrections could lead to generic, automated

feedback lacking the personal touch and tailored

guidance teachers can provide. Concerns about the

accuracy of AI tools like ChatGPT in assessing and

correcting complex or subjective assignments may

lead to perceptions of unfairness. Students might

worry about AI bias in assessments, as AI systems

may not account for diverse perspectives, cultural

nuances, or individual learning styles (Jobin et al.,

2019). Concerns that AI-generated corrections might

inadvertently introduce bias or reinforce existing

biases in evaluations. Students may feel that relying

on ChatGPT for corrections undermines the expertise

and knowledge of teachers, potentially diminishing

the value of their education. Worries that students

may not learn as effectively if AI tools are used for

corrections, as they might not receive explanations or

insights into their mistakes. There may be concerns

that students' engagement and motivation to improve

their work could decrease if they receive automated

corrections without the opportunity for meaningful

interaction with teachers. Overall, it might reduce the

teacher-student connection and the potential for

mentorship and guidance.

CSEDU 2024 - 16th International Conference on Computer Supported Education

184

7 CONCLUSIONS

Balancing fears (e.g., fear that using AI tools may be

perceived as academic dishonesty, leading to lower

grades, unfair grading by AI-based correction tools)

and potential positive effects (e.g., free use of

ChatGPT as a powerful tool for academic studies) is

essential for responsible AI integration in education.

The overarching ethical aspect 'transparency' is

crucial in addressing these concerns and ensuring

responsible AI integration in education. Additionally,

the ethical principle of 'fairness' is central to

discussions about equal access, the impact on hard

work, and the potential biases associated with AI

tools. To alleviate concerns and promote responsible

AI usage in education, universities should provide

clear guidelines, educational resources, and open

discussions to empower students to make informed

decisions and navigate the evolving landscape of AI

in academia.

Limited communication or education around the

ethical and practical use of AI tools in education can

contribute to these concerns. Students may feel that

they lack guidance on how to navigate this issue

responsibly.

Developing norms and guidelines for the ethical

use of generative AI for academic writing currently

presents a significant and complex challenge for

universities. The requirement to label AI-generated

content in academic work can contribute to

strengthening and upholding ethical, academic, and

pedagogical standards. Clear marking helps preserve

academic integrity by distinguishing between

students' own work and machine-generated content

(Boyd-Graber et al., 2023). It aids in adhering to

ethical standards in academic work. Teachers can

better assess the quality of AI-generated content and

evaluate how well students use and understand these

AI systems. This measure could also promote

students' awareness of responsible AI use and its

impact on their learning processes.

However, the results of our studies reveal

substantial arguments against labelling AI-generated

passages in academic work. Labelling could

stigmatize the use of AI in academic work, implying

that its use is inherently less valuable or legitimate.

Mandatory labelling could discourage students from

exploring and using new technologies, inhibiting

technology acceptance and the development of

necessary AI-related competencies. Regarding

human contribution, defining precisely what

constitutes AI-generated content may be challenging,

especially when students heavily edit and customize

AI outputs. Demanding labelling could be interpreted

as distrust in students' ability to handle AI

independently and responsibly. From students'

perspective, there is also a valid concern that open

communication about using AI in their work might

lead to less favourable evaluations or a loss of trust

on the part of teachers.

A significant dilemma appears between

establishing ethical academic integrity standards by

declaring ChatGPT-generated outputs and nurturing

students' AI competencies to learn how to utilize AI

tools effectively. In further research efforts, we aim

to delve deeper into this student perspective to

explore solutions that enable AI's ethical and

responsible use in higher education while

simultaneously supporting the development of

necessary AI competencies rather than hindering

them.

REFERENCES

Alshami, A., Elsayed, M., Ali, E., Eltoukhy, A. E. E., &

Zayed, T. (2023). Harnessing the Power of ChatGPT

for Automating Systematic Review Process:

Methodology, Case Study, Limitations, and Future

Directions. Systems, 11(7), 351. https://doi.org/10.

3390/systems11070351

Bao, L., Krause, N. M., Calice, M. N., Scheufele, D. A.,

Wirz, C. D., Brossard, D., Newman, T. P., &

Xenos, M. A. (2022). Whose AI? How different publics

think about AI and its social impacts. Computers in

Human Behavior, 130, 107182. https://doi.org/10.

1016/j.chb.2022.107182

Boyd-Graber, J., Okazaki, N., & Rogers, A. (2023). ACL

2023 policy on AI writing assistance. https://2023.

aclweb.org/blog/ACL-2023-policy/

Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2023).

Chatting and cheating: Ensuring academic integrity in

the era of ChatGPT. Innovations in Education and

Teaching International, 1–12. https://doi.org/10.

1080/14703297.2023.2190148

Cronbach, L. J. (1951). Coefficient alpha and the internal

structure of tests. Psychometrika, 16(3), 297–334.

https://doi.org/10.1007/BF02310555

Dang, J., & Liu, L. (2022). Implicit theories of the human

mind predict competitive and cooperative responses to

AI robots. Computers in Human Behavior, 134, 107300.

https://doi.org/10.1016/j.chb.2022.107300

Dell'Acqua, F., McFowland, E., Mollick, E. R., Lifshitz-

Assaf, H., Kellogg, K., Rajendran, S., Krayer, L.,

Candelon, F., & Lakhani, K. R. (2023). Navigating the

Jagged Technological Frontier: Field Experimental

Evidence of the Effects of AI on Knowledge Worker

Productivity and Quality. SSRN Electronic Journal.

Advance online publication. https://doi.org/10.

2139/ssrn.4573321

Student Perspectives on Ethical Academic Writing with ChatGPT: An Empirical Study in Higher Education

185

Floridi, L., & Cowls, J. (2019). A Unified Framework of

Five Principles for AI in Society. Harvard Data Science

Review. Advance online publication. https://doi.org/

10.1162/99608f92.8cd550d1

Garrel, J. von, Mayer, J., & Mühlfeld, M. (2023).

Künstliche Intelligenz im Studium Eine quantitative

Befragung von Studierenden zur Nutzung von ChatGPT

& Co. https://opus4.kobv.de/opus4-h-da/frontdoor/

deliver/index/docId/395/file/befragung_ki-im-studium.

pdf

Green, B. (2021). The Contestation of Tech Ethics: A

Sociotechnical Approach to Technology Ethics in

Practice. Journal of Social Computing, 2(3), 209–225.

https://doi.org/10.23919/JSC.2021.0018

Jobin, A., Ienca, M., & Vayena, E. (2019). The global

landscape of AI ethics guidelines. Nature Machine

Intelligence, 1(9), 389–399. https://doi.org/10.1038/s

42256-019-0088-2

Kim, S. (2022). Working With Robots: Human Resource

Development Considerations in Human–Robot

Interaction. Human Resource Development Review,

21(1), 48–74. https://doi.org/10.1177/15344843211

068810

Knight, S., Shibani, A., Abel, S., Gibson, A., & Ryan, P.

(2020). AcaWriter: A Learning Analytics Tool for

Formative Feedback on Academic Writing. Journal of

Writing Research, 12(vol. 12 issue 1), 141–186.

https://doi.org/10.17239/jowr-2020.12.01.06

Lim, W. M., Gunasekara, A., Pallant, J. L., Pallant, J. I., &

Pechenkina, E. (2023). Generative AI and the future of

education: Ragnarök or reformation? A paradoxical

perspective from management educators. The

International Journal of Management Education, 21(2),

100790. https://doi.org/10.1016/j.ijme.2023.100790

Lo, C. K. (2023). What Is the Impact of ChatGPT on

Education? A Rapid Review of the Literature.

Education Sciences, 13(4), 410. https://doi.org/

10.3390/educsci13040410

Murphy, R. (2019). Introduction to AI robotics (Second

edition). Intelligent robotics and autonomous agents.

The MIT Press.

CSEDU 2024 - 16th International Conference on Computer Supported Education

186