GenAI as a Learning Assistant, an Empirical Study in

Higher Education

Lukas Spirgi

and Sabine Seufert

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

Keywords: Artificial Intelligence (AI), Learning, AI Literacy, Higher Education.

Abstract: This empirical study investigates the use of Generative Artificial Intelligence (GenAI) as a learning assistant

in higher education and explores the impact of AI literacy on its frequency of use. Based on theoretical models

of cognitive and metacognitive learning strategies, the study analyzes how students use GenAI to support

these learning processes. Several usage scenarios were developed to explore how GenAI can be used as a

learning assistant for cognitive and metacognitive learning strategies. The results show that GenAI is

predominantly used to support cognitive learning strategies, such as explaining complex concepts and

summarizing texts, while its use to support metacognitive learning strategies, such as self-regulation and

learning planning, is less frequent. The study is based on an online survey of 266 students from the University

of St.Gallen. Using the AI literacy model of Ng, four dimensions of AI competence (affective, behavioral,

cognitive, and ethical) are measured, with the behavioral dimension identified as a significant predictor of

GenAI usage for learning activities. Developing targeted programs to promote practical AI literacy is

considered necessary to facilitate the integration of GenAI into learning processes and realize its potential

more fully.

1 INTRODUCTION

There are many different ways in which AI can be

used in education. Several studies are exploring the

opportunities and risks of new AI technologies for

educational organizations (Adiguzel et al., 2023;

Michel-Villarreal et al., 2023; Zhang & Aslan, 2021).

The use of AI tools in universities is currently the

subject of heated debate. Some view this type of

technological support as a form of cheating, mainly

when students use AI to assist with tasks like writing

papers or generating content. Others, however,

believe that collaborating with AI tools represents the

future of contemporary education, where such

technologies can enhance learning outcomes and

foster new ways of acquiring and applying knowledge

(Schneider, 2023).

A previous study by Spirgi et al. (2024) shows that

students use generative AI (GenAI) for writing

academic texts, but its potential for supporting other

learning processes remains partly underexplored. AI

technologies have and will have a significant impact

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

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

on the way we learn (Chen et al., 2020). These

changes are particularly important for students, who

need to build up a lot of knowledge in order to

complete their studies successfully.

This study focused on how university students use

GenAI for learning. In its role as a learning assistant

(LA), GenAI thus becomes an integral part of

students' learning strategy.

Research in this area is essential because GenAI

offers students many opportunities for individualized

learning and can, therefore, increase their learning

success (Alshami et al., 2023; Michel-Villarreal et al.,

2023). Early studies have already shown that students

use AI to learn (von Garrel et al., 2023).

2 AI IN ACADEMIC LEARNING

2.1 AI as a Learning Assistant

As highlighted in the introduction, GenAI tools can

be utilized to support individualized learning for

Spirgi, L. and Seufert, S.

GenAI as a Learning Assistant, an Empir ical Study in Higher Education.

DOI: 10.5220/0013199300003932

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

In Proceedings of the 17th International Conference on Computer Supported Education (CSEDU 2025) - Volume 2, pages 27-34

ISBN: 978-989-758-746-7; ISSN: 2184-5026

students. The AI tool becomes a learning assistant

when GenAI becomes an integral part of a student's

learning strategy. There are two types of learning

strategies that AI can be used for.

Learning strategies can be defined as internal

programs for controlling learning processes. Learning

strategies guide intentional learning (Schnotz, 2011).

A distinction can be made between cognitive and

metacognitive learning strategies (Wild & Schiefele,

1994).

Cognitive learning strategies refer to the direct

processes used to acquire, process, and understand

knowledge (Nückles, 2021). These strategies enable

learners to take in, organize, and remember

information actively. They focus on the "what" of

learning - the specific content and how to process it.

Metacognitive learning strategies are strategies

that relate to "thinking about thinking." They involve

planning, monitoring, reflecting, and adapting one's

learning process (Hasselhorn & Andju, 2021). These

strategies help learners to manage and optimize their

cognitive processes. It is about knowing how to learn

and being able to self-regulate learning.

GenAI can function as a learning assistant for both

cognitive and metacognitive strategies. For instance,

it can provide explanations for complex concepts that

were not sufficiently covered in class, helping

students actively acquire and understand knowledge

(cognitive). Additionally, GenAI can assist in

creating a structured learning plan, enabling students

to plan, monitor, and adapt their learning process

effectively (metacognitive). For more details, see

Table 3

2.2 AI Literacy

With the advent of large language models like

ChatGTP, AI became accessible to everyone (Seufert

& Spirgi, 2024). Technological advancements have

led to increased and more complex requirements for

the digital competencies of both students and

employees (Seufert & Guggemos, 2021). To navigate

the opportunities and challenges presented by AI,

individuals must acquire a foundational knowledge of

AI and the skills to utilize and assess AI systems

effectively (Hornberger et al., 2023). These abilities

are commonly known as AI literacy (Long &

Magerko, 2020). AI literacy has been recognized as

an essential digital literacy across various disciplines

and aspects of everyday life (Kandlhofer et al., 2016;

Long & Magerko, 2020; Ng et al., 2021a; Ng et al.,

2021b). The concept of "AI literacy" is understood

and defined in multiple ways. In their study, Ng et al.

(2024) define AI literacy as a comprehensive concept

based on four dimensions: affective dimension,

behavioral dimension, cognitive dimension, and

ethical dimension.

Affective dimension: This dimension refers to

learners' emotional responses and attitudes towards

AI. It includes intrinsic motivation, self-efficacy,

career interest, and confidence when interacting with

AI. The goal is to foster positive emotions and

attitudes to support learning and engagement with AI.

Behavioral dimension: This dimension focuses on

learners' actual behaviors when interacting with AI,

including behavioral intention, engagement, and

collaboration. The aim is to measure behaviors that

indicate active participation and the application of AI

skills in learning contexts.

Cognitive dimension: This captures learners'

knowledge and higher-order thinking skills, ranging

from understanding AI principles to applying and

evaluating AI solutions.

Ethical dimension: This dimension addresses the

moral understanding and responsible use of AI

technologies. It focuses on ethical issues such as

privacy, social responsibility, transparency, and

digital security. The goal is to ensure students use AI

responsibly and reflectively in all stages of their

learning and development.

3 THE PRESENT STUDY

This study aims to assess the prevalence of using

GenAI tools as learning assistants. Additionally, this

study aims to examine how AI literacy influences the

usage of these tools in educational settings. To

explore the role of GenAI in learning strategies, we

pose three research questions:

1. How often do students use GenAI as a

learning assistant for cognitive and

metacognitive learning strategies?

2. Are there any gender differences in using AI

as a learning assistant?

3. How does AI literacy affect the frequency of

using GenAI as a learning assistant?

Ng et al. (2024) concept of AI literacy serves as the

foundation for addressing research question 3 (see

Chapter 2.2). By exploring the role of GenAI in

educational processes, this study deepens our

understanding of how AI can support learning. It

advances the field by offering a comprehensive

perspective on the integration of AI into the learning

process, with a particular focus on both cognitive and

metacognitive learning strategies.

CSEDU 2025 - 17th International Conference on Computer Supported Education

4 METHODS

4.1 Online Survey and Sample

A digital survey was selected for the study to

thoroughly investigate students' experiences with AI.

The survey was conducted online using the

"Qualtrics" platform between September and October

2024. All the questions were single-choice. A total of

266 students from the local university completed the

survey. The average age of the respondents is 19.7

(SD = 1.1) years. The students taking part in the

survey are studying either economics or law. Table 1

shows the composition of the sample.

Table 1: Sample.

Characteristic Absolute Percentage

Female Students 107 40.2 %

Male Students 157 59.0 %

Diverse Students 2 0.8 %

First Semester Students 253 95.1 %

Bachelor Students 8 3.0 %

Master Students 5 1.9 %

4.2 Development of Instrument

The questionnaire consisted of two sections. The first

section assessed the frequency of using GenAI as a

learning assistant. The authors of this article

developed four items covering scenarios in which

GenAI is used to support cognitive learning

strategies. The areas addressed include acquiring

knowledge, summarizing content, applying

knowledge, and explicitly explaining concepts.

Additionally, four items were developed to address

scenarios where GenAI supports metacognitive

learning strategies. These included planning the

learning process, self-monitoring, self-reflection, and

adapting learning strategies. The fully elaborated

items are presented in Table 3. Participants were

asked to indicate the frequency of their usage on a 5-

point Likert scale. The scale included the following

response options: "Never" = no use, "Rarely" = once

per semester, "Sometimes" = once per month,

"Frequently" = multiple times per month, and

"Usually" = once or more per week.

In the second part of the survey, Ng et al. (2024)

validated instrument was used to assess participants'

AI literacy. This instrument measured four

dimensions of AI literacy: the affective dimension,

the behavioral dimension, the cognitive dimension,

and the ethical dimension. Participants were

presented with 5 to 6 items for each of these

dimensions. Respondents were asked to indicate the

extent to which they agreed with each statement using

a 5-point Likert scale with the following response

options: "strongly disagree,” "disagree,” "neutral,”

"agree," and "strongly agree." An example item was:

"To what extent do you agree with the following

statement: Artificial intelligence is relevant to my

everyday life (e.g., personal, work)."

4.3 Statistical Testing

Statistical testing is performed initially using multiple

regression analysis. Multiple regression analysis

evaluates whether a dependent variable could be

predicted based on independent variables (Seber &

Lee, 2012). The significant difference is set at

α = 0.05. Assumptions of the regression, including

linearity, normality of residuals, homoscedasticity,

multicollinearity, and independence of errors, were

tested to ensure the validity of the model. Regression

analysis was performed using R.

The scales were created by calculating the mean

of their corresponding items. A total of 5 different

scales were created for this paper (see Table 2). The

Learning Assistant scale represents the use of AI as a

learning assistant. The other scales represent one

dimension of AI literacy.

The difference between two means is calculated

using an independent two-sample t-test.

5 RESULTS

5.1 Internal Consistency

The internal consistency of the scales was evaluated

using Cronbach's alpha, a measure that indicates how

closely related the items within an index are.

Table 2: Frequency of use index.

Scale Items Cronbach's Alpha

Learning Assistant

α = 0.87

Affective dimension 6 α = 0.80

Behavioral dimension 5 α = 0.73

Cognitive Dimension 5 α = 0.77

Ethical dimension 6 α = 0.74

All indexes showed values above 0.7, reflecting

an acceptable to a good level of internal consistency

(Cronbach, 1951).

GenAI as a Learning Assistant, an Empirical Study in Higher Education

Table 3: AI as Personal Learning Assistant: "I use gen AI…".

Type of use of ChatGPT

(SD)

Never Rarely

Sometime

Frequently Usually

cogitive

- to get explanations for complex

concepts that were not sufficiently

covered in class

2.9

(1.4)

22 % 17 % 24 % 20 % 18 %

- to get concise summaries of longer texts

or articles

2.9

(1.3)

21 % 17 % 25 % 24 % 14 %

- to obtain examples of how to apply

theoretical knowledge in practical

situations.

2.2

(1.2)

39 % 24 % 19 % 12 % 6 %

- to get simplified explanations of

difficult terms or theories.

3.0

(1.4)

22 % 14 % 23 % 21 % 20 %

metacognitive

- to create a structured learning plan.

1.5

(0.9)

71 % 17 % 6 % 5 % 1 %

- to check my understanding of the

learning material by testing myself or

completing exercises

1.7

(1.1)

65 % 14 % 11 % 5 % 5 %

- to receive feedback on my learning

progress and to identify areas for

improvement.

1.5

(1.0)

72 % 14 % 7 % 4 % 3 %

- to find alternative learning methods

when I have difficulty understanding the

material

1.8

(1.1)

60 % 15 % 14 % 7 % 3 %

5.2 Current Frequency of Using GenAI

as a Learning Assistant

Table 3 presents data on the usage frequency of GenAI

as a personal learning assistant across two primary

categories: cognitive and metacognitive learning

strategies. In the cognitive domain, the average

frequency of AI use is higher compared to

metacognitive strategies. The difference is statistically

significant. Overall, GenAI is used infrequently for

learning, as indicated by the generally low mean scores

across both cognitive and metacognitive tasks.

The standard deviation of the means is similar for

all items. The distributions are shown in the table. The

percentages indicate the frequency of answer choices.

The item with the highest average (mean) value is

"to get simplified explanations of difficult terms or

theories", with a mean of 3.0 (SD = 1.4). This

indicates that GenAI supports this cognitive task the

most frequently among the participants in the study.

The item with the lowest average (mean) value is

"to create a structured learning plan" in the

metacognitive category, with a mean of 1.5 (SD =

0.9). This suggests that using GenAI for this specific

metacognitive task is the least frequent.

5.3 Gender Differences

The following figure shows the gender differences in

frequency of use. For the sake of simplicity, the

gender category "Diverse" was not included in the

diagram.

Figure 1: Gender differences.

For both cognitive and metacognitive learning

strategies using GenAI, the differences in mean

scores between women and men were not statistically

significant. This indicates that there is insufficient

statistical evidence to support the existence of

systematic differences between genders on these

scales.

5.4 Frequency of Use and AI Literacy

The regression analysis shows that the model

provides a somewhat limited but sufficient

2,84

1,57

2,73

1,67

cogitive metacognitive

Women Men

CSEDU 2025 - 17th International Conference on Computer Supported Education

explanation of the variance in the dependent variable,

Learning Assistant. The overall model is statistically

significant.

Among the predictors, behavioral dimension

emerged as a significant positive predictor of GenAI

usage as a learning assistant, suggesting that

individuals with stronger behavioral dimension

strategies are more likely to engage with the Learning

Assistant. While not significant, the Affective

dimension showed a trend toward a positive

association. Neither the cognitive dimension nor

ethical dimension were significant predictors in the

model, indicating that these dimensions of AI

Literacy do not substantially influence the use of the

Learning Assistant in this context.

These results highlight the importance of

behavioral strategies in this learning environment,

while other AI Literacy dimensions seem to have less

influence on the use of GenAI as a Learning

Assistant.

Table 4: Regression.

Dependent variable:

Learning Assistant

Estimate P-Value

Intercept 0.0080 (0.9838)

Affective dimension 0.1967

(

0.0512

)

Behavioral dimension 0.3479 (0.0005) ***

nitive Dimension 0.1296

(

0.1439

)

Ethical dimension -0.0402 (0.6460)

Residual Std. Erro

: 0.7904 (df = 261)

: 0.1938

Adjusted R

: 0.1814

F Statistic: 15.68

(

df = 4; 261

)

***

6 DISCUSSION

6.1 Low Frequency of Use of AI as a

Learning Assistant

The results indicate that students currently

underutilize GenAI tools as comprehensive learning

assistants. Their most frequent application is

simplifying and clarifying course materials,

particularly by providing explanations of complex

concepts not adequately covered in class or

summarizing lengthy texts and articles. This result is

consistent with the usage frequencies collected by

von Garrel et al. (2023). This suggests that AI tools

are primarily embedded within cognitive learning

strategies, focusing on acquiring and processing

knowledge. These tasks are usually concrete and less

demanding, which makes the use of GenAI more

intuitive and can lead to greater acceptance among

students.

In contrast, students significantly less frequently

use AI tools to support metacognitive learning

strategies. Activities such as creating structured

learning plans, assessing one's understanding, or

obtaining feedback on learning progress are rarely

supported by AI tools. This limited usage suggests

that students may either be unaware of the capabilities

of AI tools in these areas or may be reluctant to trust

AI in more personal and reflective processes.

One possible explanation for the higher

prevalence of cognitive applications over

metacognitive ones is that cognitive tasks are more

concrete and less complex. Asking for explanations

or summaries is a tangible activity that can be easily

integrated into existing study routines. In contrast,

metacognitive strategies, which involve planning,

monitoring, and evaluating one's learning, require a

higher level of self-awareness and self-regulation. It

is possible that students are either unfamiliar with the

relevant functions of AI tools or hesitate to use them

for tasks that require more intensive personnel.

Previous studies have consistently observed that

men tend to use AI tools more intensively than

women, particularly in contexts where AI is

employed as a writing assistant (Seufert et al., 2024;

Spirgi et al., 2024). However, this gender disparity is

no longer evident when AI is utilized as a learning

assistant. One possible explanation for this shift could

be that the functions associated with AI in educational

settings, such as obtaining explanations or

simplifying complex materials, may appeal equally to

both genders.

6.2 AI Literacy Influences AI Learning

A particularly noteworthy finding from the regression

analysis is the statistically significant impact of the

behavioral dimension on the use of GenAI as a

learning assistant. With a coefficient of 0.3479 and a

highly significant p-value of 0.0005, this dimension

shows a clear positive correlation with the frequency

of GenAI usage (see table 4). This suggests that

students' actual behaviors when interacting with AI,

such as active usage, engagement, and interaction

with AI tools, play a critical role in determining how

frequently GenAI is integrated into learning

processes.

There are several potential reasons for the strong

influence of the behavioral dimension. First, this

indicates that learners who take a proactive approach

GenAI as a Learning Assistant, an Empirical Study in Higher Education

toward using AI are more likely to incorporate these

technologies into their daily learning routines. The

behavior of actively engaging with AI might reflect a

higher degree of confidence in their ability to use AI

tools effectively. Engagement with AI can also be

seen as a marker of greater technological competence,

which could lower the perceived barriers to the

frequent use of AI tools like GenAI.

Additionally, the accessibility and user-friendly

nature of GenAI tools, such as ChatGPT, may

significantly reinforce this behavior. These tools are

designed for ease of use, allowing learners who are

willing to experiment with them to quickly recognize

their benefits in various learning contexts. This ease

of use influences the frequency of technology use

(Davis, 1985). As learners actively engage with these

technologies and experience positive outcomes—

such as improved learning efficiency or enhanced

understanding—this can create a positive feedback

loop, motivating further usage.

Social factors also likely play a role. Observing

peers' successful use of AI tools in collaborative

learning environments can encourage others to follow

suit. According to social learning theory, individuals

are more likely to adopt behaviors that they see

modeled successfully by others (Bandura, 1977),

further reinforcing the behavioral dimension's

influence on AI adoption.

In summary, the significant influence of the

behavioral dimension on the use of GenAI as a

learning assistant can be explained by several factors:

the willingness to adopt new technologies, the

accessibility and feedback from AI tools, and the

social and institutional context. This dimension

highlights that actual behavior and active engagement

with AI are crucial in fully realizing the potential

benefits of AI in educational settings.

7 CONCLUSIONS

7.1 Theoretical Implications

The findings of this study not only highlight the

underutilization of GenAI tools as learning assistants

but also contribute to the theoretical understanding of

their role in educational contexts. Specifically, we

have developed the concept of GenAI as a learning

assistant by distinguishing its application between

cognitive and metacognitive learning strategies.

Cognitive strategies involve processes related to

knowledge acquisition and organization, while

metacognitive strategies focus on self-regulation,

reflection, and strategic planning. Additionally, this

study integrates Ng's AI literacy framework, linking

it to the concept of GenAI as a learning assistant.

While students frequently use AI to support

cognitive learning strategies—such as simplifying

complex concepts or summarizing materials—there

is a notable gap in the use of GenAI for more

complex, metacognitive learning processes, such as

self-regulation, reflection, and strategic planning.

This indicates that the full range of GenAI's

capabilities remains largely untapped in educational

settings. A key takeaway from the regression analysis

is the significant influence of the behavioral

dimension on GenAI usage. The proactive use of AI,

including active engagement and experimentation

with AI tools, is more important than merely

possessing cognitive knowledge about AI. This

suggests that "doing"—actively using and

experimenting with AI tools—is more critical than

"knowing" in terms of effectively integrating AI into

both cognitive and metacognitive learning processes.

7.2 Practical Implications

These findings emphasize the need to foster hands-on

engagement with AI technologies in educational

environments, as behavioral engagement is a crucial

driver of AI adoption and frequent usage. There is a

need to equip students with solid competencies in AI

literacy, particularly in practical and behavioral

aspects. As AI continues to evolve and become more

integrated into various sectors, the ability to use AI

tools effectively will become increasingly critical, not

only for academic success but also for future career

readiness (Baidoo-Anu & Owusu Ansah, 2023;

Laupichler et al., 2022). To address this gap, targeted

training programs should be developed to encourage

students to explore the full spectrum of AI tools and

their applications in learning. These programs should

prioritize active engagement and provide

opportunities for students to experiment with AI

rather than focusing solely on theoretical knowledge.

By doing so, educational institutions can empower

students to leverage AI as a comprehensive learning

assistant, ultimately enhancing both their learning

outcomes and their preparedness for a technology-

driven future. University lecturers play a key role in

this process, as they can provide students with easy

access to AI tools by integrating them into their

courses. In this way, the behavioral dimension of AI

literacy can be directly addressed in the classroom.

7.3 Further Research

While this study highlights important findings

CSEDU 2025 - 17th International Conference on Computer Supported Education

regarding the underutilization of GenAI tools and the

significance of behavioral engagement, several areas

for future research remain. One key focus could be

exploring the specific barriers that prevent students

from fully utilizing GenAI for metacognitive learning

strategies, such as self-regulation and reflection.

Identifying whether these barriers stem from a lack of

awareness, trust issues, or insufficient AI literacy

could help inform the development of targeted

interventions.

7.4 Limitations

This study is subject to several limitations. First, the

sample primarily consisted of first-semester students,

which may limit the generalizability of the findings to

more experienced students. Additionally, the study

was conducted exclusively with students from the

local university, further restricting its scope. Lastly,

all participants were enrolled in economics-related

programs, meaning the results may not be fully

applicable to students from other academic

disciplines.

REFERENCES

Adiguzel, T., Kaya, M. H., & Cansu, F. K. (2023).

Revolutionizing education with AI: Exploring the

transformative potential of ChatGPT. Contemporary

Educational Technology, 15(3), ep429.

https://doi.org/10.30935/cedtech/13152

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

Baidoo-Anu, D., & Owusu Ansah, L. (2023). Education in

the Era of Generative Artificial Intelligence (AI):

Understanding the Potential Benefits of ChatGPT in

Promoting Teaching and Learning. SSRN Electronic

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

10.2139/ssrn.4337484

Bandura, A. (1977). Social learning theory. Englewood

Cliffs.

Chen, X., Xie, H., Di Zou, & Hwang, G.‑J. (2020).

Application and theory gaps during the rise of Artificial

Intelligence in Education. 2666-920X, 1, 100002.

https://doi.org/10.1016/j.caeai.2020.100002

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

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

https://doi.org/10.1007/BF02310555

Davis, F. D. (1985). A technology acceptance model for

empirically testing new end-user information systems:

Theory and results [, Massachusetts Institute of

Technology]. EndNote Tagged Import Format.

https://dspace.mit.edu/bitstream/handle/1721.1/15192/

14927137-mit.pdf

Hasselhorn, M., & Andju, S. L. (2021, November 24).

Metakognitive Lernstrategien. Dorsch Lexikon der

Psychologie. https://dorsch.hogrefe.com/stichwort/

lernstrategien-metakognitive

Hornberger, M., Bewersdorff, A., & Nerdel, C. (2023).

What do university students know about Artificial

Intelligence? Development and validation of an AI

literacy test. 2666-920X, 5, 100165. https://doi.org/10.

1016/j.caeai.2023.100165

Kandlhofer, M., Steinbauer, G., Hirschmugl-Gaisch, S., &

Huber, P. (2016). Artificial intelligence and computer

science in education: From kindergarten to university.

In 2016 IEEE Frontiers in Education Conference (FIE)

(pp. 1–9). IEEE. https://doi.org/10.1109/FIE.2016.

7757570

Laupichler, M. C., Aster, A., Schirch, J., & Raupach, T.

(2022). Artificial intelligence literacy in higher and

adult education: A scoping literature review. 2666-

920X, 3, 100101. https://doi.org/10.1016/j.caeai.

2022.100101

Long, D., & Magerko, B. (2020). What is AI Literacy?

Competencies and Design Considerations. In R.

Bernhaupt, F. '. Mueller, D. Verweij, J. Andres, J.

McGrenere, A. Cockburn, I. Avellino, A. Goguey, P.

Bjørn, S. Zhao, B. P. Samson, & R. Kocielnik (Eds.),

Proceedings of the 2020 CHI Conference on Human

Factors in Computing Systems (pp. 1–16). ACM.

https://doi.org/10.1145/3313831.3376727

Michel-Villarreal, R., Vilalta-Perdomo, E., Salinas-

Navarro, D. E., Thierry-Aguilera, R., & Gerardou, F. S.

(2023). Challenges and Opportunities of Generative AI

for Higher Education as Explained by ChatGPT.

Education Sciences, 13(9), 856. https://doi.org/10.

3390/educsci13090856

Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S.

(2021a). AI Literacy: Definition, Teaching, Evaluation

and Ethical Issues. Proceedings of the Association for

Information Science and Technology, 58(1), 504–509.

https://doi.org/10.1002/pra2.487

Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S.

(2021b). Conceptualizing AI literacy: An exploratory

review. Computers and Education: Artificial

Intelligence, 2, 100041. https://doi.org/10.1016/j.

caeai.2021.100041

Ng, D. T. K., Wu, W., Leung, J. K. L., Chiu, T. K. F., &

Chu, S. K. W. (2024). Design and validation of the AI

literacy questionnaire: The affective, behavioural,

cognitive and ethical approach. British Journal of

Educational Technology, 55(3), 1082–1104.

https://doi.org/10.1111/bjet.13411

Nückles, M. (2021, November 24). Kognitive

Lernstrategien. Dorsch Lexikon der Psychologie.

https://dorsch.hogrefe.com/stichwort/lernstrategien-

kognitive

Schneider, R. U. (2023, November 23). Chat-GPT erobert

die Universitäten: Darf der Computer die Seminararbeit

schreiben? NZZ. https://www.nzz.ch/gesellschaft/ki-

GenAI as a Learning Assistant, an Empirical Study in Higher Education

an-der-uni-wenn-chat-gpt-die-seminararbeit-schreibt-

ld.1766150

Schnotz, W. (2011). Pädagogische Psychologie (2.,

überarb. und erw. Aufl.). Beltz Kompakt. Beltz.

Seber, G. af, & Lee, A. J. (2012). Linear regression

analysis. John Wiley & Sons.

Seufert, S., & Guggemos, J. (2021). Zukunft der Arbeit mit

intelligenten Maschinen:Implikationen der Künstlichen

Intelligenz für die Berufsbildung - Einleitung zum

Beiheft. Steiner. https://www.alexandria.unisg.ch/

265016/

Seufert, S., & Spirgi, L. (2024). Soziotechnische

Systemgestaltung im Kontext generativer KI: eine

Konzeption in der Hochschulbildung. Interner

Arbeitsbericht.

Seufert, S., Spirgi, L., Delcker, J., Heil, J [Joana], &

Ifenthaler, D. (2024). Umgang mit KI-Robotern:

maschinelle Übersetzer, Textgeneratoren, Chatbots &

Co – Eine empirische Studie bei Erstsemester-

Studierenden. In K. Kögler, J. M.-C. Schmidt, & M.

Egloffstein (Eds.), Empirische Pädagogik: 38 (1).

Lehren und Lernen mit und über Künstliche Intelligenz

in der Aus- und Weiterbildung (pp. 47–72).

Verlag Empirische Pädagogik. https://www.vep-

landau.de/produkt/empirische-paedagogik-2024-38-1-

kap-3-digital/

Spirgi, L., Seufert, S., Delcker, J., & Heil, J [Joanna]

(2024). Student Perspectives on Ethical Academic

Writing with ChatGPT: An Empirical Study in Higher

Education. Proceedings of the 16th International

Conference on Computer Supported

Education(Volume 2), 179–186.

von 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_k

i-im-studium.pdf

Wild, K.‑P., & Schiefele, U. (1994). Lernstrategien im

Studium: Ergebnisse zur Faktorenstruktur und

Reliabilität eines neuen Fragebogens (15(4)).

https://psycnet.apa.org/record/1996-85746-001

Zhang, K., & Aslan, A. B. (2021). AI technologies for

education: Recent research & future directions. 2666-

920X, 2, 100025. https://doi.org/10.1016/j.caeai.

2021.100025.

CSEDU 2025 - 17th International Conference on Computer Supported Education