Access to this full-text is provided by Frontiers.
Content available from Frontiers in Psychology
This content is subject to copyright.
Frontiers in Psychology 01 frontiersin.org
The influence of generative
artificial intelligence on creative
cognition of design students: a
chain mediation model of
self-ecacy and anxiety
YounjungHwang
1 and YiWu
2*
1 School of Design, Hunan University, Changsha, China, 2 School of International Communication and
Arts, Hainan University, Haikou, China
Introduction: This study investigated the role of generative Artificial Intelligence
(AI) in enhancing the creative cognition of design students, examining the
mediating eects of self-ecacy and anxiety reduction.
Methods: A quantitative approach was employed, collecting data through
online surveys from 121 design students at universities in southern China. The
study utilized scales for AI knowledge and perception, self-ecacy, anxiety, and
creative cognition, adapted from previous studies and evaluated on 5-point
Likert scales. Data analysis was conducted using SPSS 24.0 for exploratory factor
analysis and PROCESS v3.5 for mediation analysis.
Results: The findings confirmed that AI positively impacted students’ innovative
thinking (*β* = 0.610, *p* < 0.001). Self-ecacy (standardized *β* = 0.256,
95% CI [0.140, 0.418], *p* < 0.001) and anxiety reduction (standardized
*β* = 0.093, 95% CI [0.018, 0.195], *p* < 0.05) positively mediated the
relationship between generative AI and creative cognition. Additionally, a serial
mediation eect through self-ecacy and anxiety reduction was observed
(standardized *β* = 0.053, 95% CI [0.012, 0.114], *p* < 0.05).
Discussion: Our empirical analysis demonstrates that AI positively aects design
students’ innovative thinking, with self-ecacy and anxiety reduction serving
as significant mediators. These findings provide valuable insights for educators
and policymakers, suggesting that AI-integrated design curricula can significantly
foster creative cognition, promote academic achievement, and enhance designer
capabilities. Understanding AI’s impact on students’ creative processes is crucial for
developing eective teaching strategies in today’s evolving educational landscape.
KEYWORDS
generative AI, creative cognition, self-ecacy, anxiety, design education
1 Introduction
In recent years, Articial Intelligence (AI) has emerged as a transformative force across
various domains, including education and creative industries. AI refers to computer systems
capable of performing complex tasks traditionally associated with human cognition, such as
reasoning, decision-making, and problem-solving (Russell and Norvig, 2020). Generative AI,
a subset of AI, is designed to analyze patterns in large datasets and use these patterns to
generate outcomes based on user requirements (Mandapuram etal., 2018; Liang etal., 2023).
As this technology gains prominence in creative disciplines, its integration in design education
OPEN ACCESS
EDITED BY
Iclal Can,
Middle East Technical University Northern
Cyprus Campus, Cyprus
REVIEWED BY
Connie Phelps,
Emporia State University, UnitedStates
Katalin Grajzel,
University of Denver, UnitedStates
Brad Hokanson,
University of Minnesota Twin Cities,
UnitedStates
*CORRESPONDENCE
Yi Wu
wuyi@hainanu.edu.cn
RECEIVED 26 June 2024
ACCEPTED 27 December 2024
PUBLISHED 27 January 2025
CITATION
Hwang Y and Wu Y (2025) The influence of
generative artificial intelligence on creative
cognition of design students: a chain
mediation model of self-ecacy and anxiety.
Front. Psychol. 15:1455015.
doi: 10.3389/fpsyg.2024.1455015
COPYRIGHT
© 2025 Hwang and Wu. This is an
open-access article distributed under the
terms of the Creative Commons Attribution
License (CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that the
original publication in this journal is cited, in
accordance with accepted academic
practice. No use, distribution or reproduction
is permitted which does not comply with
these terms.
TYPE Original Research
PUBLISHED 27 January 2025
DOI 10.3389/fpsyg.2024.1455015
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 02 frontiersin.org
sparks considerable debate among researchers and educators, with
conicting views on its impact on creativity and learning outcomes.
Some studies suggest that AI fosters learners’ creativity by
providing new tools and perspectives. For instance, language models
like ChatGPT and text-to-image models such as Midjourney, Stable
Diusion, and DALL-E are increasingly adopted to enhance human
creative cognition– the process of generating innovative ideas (Voigt
etal., 2023). ese tools oer students rapid prototyping capabilities
and access to vast databases of design inspiration, potentially
expanding their creative horizons. However, other research indicates
that AI might inhibit creativity by promoting over-reliance on
machine-generated solutions (Elgammal etal., 2017). Critics argue
that excessive dependence on AI tools could lead to a homogenization
of design outputs and a decrease in original, human-driven creativity.
Furthermore, in design education, the integration of AI tools not
only introduces new creative methods but also impacts the
psychological dynamics of learning. Key factors such as self-ecacy
and anxiety inuence how students engage with AI and, consequently,
their creative outcomes (Runco and Chand, 1995; Sifonis and Ward,
2022). Despite AI’s growing presence in educational settings, little is
known about how these psychological mechanisms shape creative
cognition and design education. is gap in knowledge underscores
the importance of understanding the complex interrelationship
between AI and creative cognition for both students and educators in
the eld of visual design.
is ongoing debate highlights the need to explore how the
integration of AI in design education aects students’ creative
processes and to identify the psychological mechanisms through
which AI inuences creative cognition. Specically, examining the
roles of self-ecacy and anxiety in mediating the relationship between
AI use and creative output is crucial for understanding how AI can
best support creative learning. To address these gaps in knowledge,
this study aimed to explore these eects, focusing on the roles of self-
ecacy and anxiety, and oering insights into how AI can optimally
support creative learning in design education.
1.1 Self-ecacy and creative cognition
e concept of self-ecacy, which refers to the individual’s belief
in their ability to successfully perform tasks, plays a crucial role in
promoting creative cognition. Bandura (1977) introduced the concept
of self-ecacy and its impact on human behavior, while Beghetto
(2006) specically explored its relationship with creative performance
in educational settings. Beghetto’s study of 1,322 middle and secondary
school students found signicant positive relationships between
creative self-ecacy and several factors related to creative performance,
including mastery orientation (β = 0.30, p < 0.001), performance-
approach orientation (β = 0.12, p < 0.001), and teacher feedback on
creative ability (β = 0.32, p < 0.001). is nding suggested that
students who believe in their creative abilities are more likely to engage
in creative endeavors and produce innovative outcomes.
Building on this foundation, McGuire etal. (2024) examined the
eects of AI collaboration on creative self-ecacy and creativity in
poetry writing. eir study found that participants who co-created
poems with AI reported signicantly higher levels of creative self-
ecacy compared to those who merely edited AI-generated poems
(M = 4.62 vs. M = 3.74, p = 0.003). is increased self-ecacy
translated into higher expert evaluations of creativity for the co-creator
group compared to the editor group (M = 14.70 vs. M = 12.53,
p = 0.026). Mediation analysis revealed that creative self-ecacy
signicantly mediated the relationship between co-creation with AI
and expert evaluations of creativity (indirect eect, β = 0.78, SE = 0.39,
95% CI [0.16, 1.68]). ese ndings suggest that when properly
designed to foster co-creation, AI tools can enhance users’ creative
self-ecacy, leading to improved creative outcomes. However, a
limitation of this research is that it does not address the long-term
impact of co-creating versus editing with AI, as the study focused on
a single creative task rather than examining eects over an extended
period of time.
1.2 Anxiety and creative cognition
Another important factor in the creative process is anxiety, a
negative emotion, which is characterized by tension, physical ailments,
and worrisome thoughts (Spielberger, 1972). Creative anxiety is
particularly common among designers and design students (Daker
etal., 2020; Wang and Jia, 2021). Research has shown that reducing
anxiety can enhance students’ creative cognition by freeing up
cognitive resources typically consumed by stress, thus facilitating the
development of innovative ideas. In a meta-analysis of 76 experimental
studies, Byron etal. (2010) found that low-anxiety individuals showed
signicantly increased creative performance when exposed to stressors
compared to high-anxiety individuals (d = 0.39, 95% CI [0.19, 0.62]
vs. d = −0.12, 95% CI [−0.33, 0.09], respectively). is suggests that
interventions aimed at reducing anxiety could potentially improve
creative outcomes in design education.
According to recent research, the relationship between anxiety
and self-ecacy in creative design contexts appears to beinuenced
by AI-generated content (AIGC) tools. Based on Li’s (2024) study of
404 design students and professionals, the relationship between
anxiety, self-ecacy, and AI in creative design contexts reveals
signicant insights. While AI tools did not directly reduce anxiety,
they indirectly inuenced it through other factors. e study found
that performance expectancy (β = 0.57, p < 0.001) and social inuence
(β = 0.24, p < 0.001) positively impacted designers’ intention to use AI
tools. is suggests that as designers perceive AI tools as enhancing
their performance and gain social support, their self-ecacy may
increase, potentially leading to reduced anxiety about creative tasks.
However, this study did not directly measure changes in anxiety or
self-ecacy levels over time, limiting our understanding of AI’s long-
term psychological impact on designers.
1.3 Research aims and hypotheses
Considering these various factors, given the complex
interrelationships and gaps in current research, this study aimed to
explore the connections among self-ecacy, anxiety, AI design
knowledge, and creative cognition in the context of design education.
e research sought to address how AI tools in design education aect
students’ creative cognition, the extent to which self-ecacy mediates
this relationship, and how anxiety reduction contributes to the
enhancement of creative cognition through AI tools. Additionally, the
study examined the potential sequential mediating eect of
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 03 frontiersin.org
self-ecacy and anxiety reduction in the relationship between AI use
and creative cognition.
e researchers hypothesized that AI tools would initially boost
students’ self-ecacy by providing rapid prototyping and ideation
support. is increased self-ecacy was expected to lead to reduced
anxiety about the creative process as students felt more capable of
tackling design challenges. Consequently, the combination of
enhanced self-ecacy and reduced anxiety was hypothesized to
facilitate improved creative cognition, enabling students to generate
more innovative and eective design solutions.
rough this investigation, the study aimed to contribute
empirical evidence to support the connections between AI tools and
creative cognition, specically examining how self-ecacy and
anxiety reduction mediate this relationship in the context of design
education. By clarifying the mechanisms through which AI inuences
creative cognition, the research sought to inform the development of
more eective AI-integrated design curricula and oer valuable
insights for educators striving to harness AI’s potential while
addressing its possible drawbacks.
2 Literature review
2.1 Generative AI for enhancing students’
creative cognition
Creative cognition encompasses the cognitive processes
involved in generating novel and innovative ideas, engaging in
divergent thinking, and exhibiting creativity that goes beyond
routine problem-solving. is creative thinking is characterized by
exibility, originality, and the ability to make unexpected
connections, distinguishing it from more conventional, linear
thought processes used in everyday tasks (Smith et al., 1995;
Ahmad Abraham and Windmann, 2007; Runco and Acar, 2012). It
also includes metacognitive elements, which refer to higher-order
thinking processes that allow individuals to reect on, monitor,
and control their cognitive activities during creative work
(Amabile, 1983; Hargrove and Nietfeld, 2015). ese metacognitive
processes involve strategic planning, evaluating progress, and
adjusting approaches as needed. For instance, a designer generating
multiple ideas is engaging in cognitive activity, while assessing the
eectiveness of their ideation strategy, altering their approach if
necessary, and reecting on how well they are meeting project
goals represents metacognitive activity (Jaušovec, 1994; Kaufman
and Beghetto, 2013). e interplay between cognitive and
metacognitive processes in creative work is complex and dynamic.
While problem-solving can involve both cognitive and
metacognitive elements, the distinction lies in the level of
conscious reection and control. Routine problem-solving may
rely more heavily on cognitive processes, while novel or complex
problem-solving oen requires more explicit metacognitive
engagement (Sternberg and Lubart, 1996; Mumford etal., 2012).
Furthermore, as our understanding of human creativity
continues to evolve, advancements in AI have introduced new tools
that may complement and enhance cognitive processes. Generative
AI, in particular, has garnered signicant attention for its potential
to support and expand creative cognition. Designed to analyze
patterns in large datasets and generate outcomes based on user
requirements (Mandapuram et al., 2018; Liang et al., 2023),
generative AI has made signicant strides in various domains.
Large language models (LLMs) like ChatGPT have revolutionized
text generation and comprehension (olander and Jonsson, 2023;
Lanzi and Loiacono, 2023; Tan and Luhrs, 2024), while image
generation models such as Midjourney, Stable Diusion, and
DALL-E 2 have transformed visual creative processes (Paananen
et al., 2023; Hanafy, 2023; Yüksel and Börklü, 2023). As these
technologies continue to evolve, their applications have expanded
far beyond their original domains, sparking growing interest in
elds such as education (Sadek, 2023; Slimi, 2023).
In educational settings, the integration of generative AI has
shown promise in enhancing students’ creative cognition. ese AI
systems serve as collaborative tools, bridging the gap between data
scientists and non-experts, and enabling individuals without prior
knowledge of data analysis and machine learning to understand
and utilize complex concepts through conversation (Li, 2023). By
engaging in dialogue with students, AI made dicult topics more
accessible by providing explanations tailored to individual
understanding levels.
is approach oers opportunities for students from diverse
backgrounds and knowledge levels to grasp and apply complex
concepts more easily. For instance, a student struggling with a
particular machine learning algorithm could ask ChatGPT to
explain it in simpler terms or relate it to familiar concepts (Islam
etal., 2023), potentially enhancing students’ creativity by making
it easier for learners from diverse backgrounds to grasp and apply
concepts in data science and machine learning.
is capability of generative AI not only facilitates learning but
also promotes creativity by making advanced concepts more
approachable and manageable, thus contributing to the
development of students’ innovative thinking skills.
2.2 Self-ecacy as a mediator
Self-ecacy, dened as an individual’s belief in their ability to
successfully perform tasks, has been shown to signicantly impact
creative activities and outcomes (Bandura, 1977; Maddux, 2013;
Rahayuningsih etal., 2022; Yu etal., 2023; Bozdoğan, 2023; Raihan
and Uddin, 2023). In the context of education, self-ecacy plays a
crucial role in the development of undergraduate students’ creative
cognitive abilities (Sri etal., 2019; Beghetto, 2006; Qian etal., 2023;
Liang etal., 2023). For instance, Beghetto (2006) found a signicant
positive correlation between creative self-ecacy and self-reported
creative performance among middle and secondary school students
(r = 0.572, p < 0.001). Similarly, Qian etal. (2023) demonstrated that
self-directed learning positively inuences creativity in healthcare
undergraduates through the mediating eects of openness to challenge
and diversity, as well as creative self-ecacy (indirect eect = 0.324,
95% CI [0.165, 0.543]).
Building on this understanding, recent research has explored the
potential inuence of AI on students’ self-ecacy and motivation. Jia
and Tu (2024) investigated the impact of AI capabilities on college
students’ self-ecacy, nding a signicant positive relationship
(β = 0.546, p < 0.001). Expanding on this, Yilmaz and Yilmaz (2023)
examined the eects of AI-based tools on programming self-ecacy,
revealing that students using AI tools scored signicantly higher in
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 04 frontiersin.org
programming self-ecacy compared to the control group (F(1,
42) = 15.144, p < 0.001).
Furthermore, Wang and Chuang (2023) explored the eects of
higher education institutes’ AI capability on students’ self-ecacy,
creativity, and learning performance. eir study indicated that AI
capability signicantly aects students’ self-ecacy (β = 0.515,
p < 0.001) and creativity (β = 0.533, p < 0.001), which in turn
positively inuence learning performance.
This transformative impact of AI extends to specific domains
within education, such as design-related fields, where the
enhancement of self-efficacy through AI-assisted learning
significantly influences students’ creative processes. Research has
shown that increased confidence in design, partly facilitated by
AI tools, promotes students’ independent and divergent thinking,
enabling them to express and experiment with their ideas more
freely. In this context, Rao etal. (2020) conducted a study with
150 design students, finding that those who reported higher
levels of confidence in their design abilities were 30% more likely
to propose innovative solutions to given design problems. The
researchers attributed this increased confidence partly to the use
of AI-assisted design tools. Similarly, Liu etal. (2023) observed
in their longitudinal study of 200 undergraduate design students
that increased confidence, partly attributed to AI-assisted
learning tools, correlated with a 25% increase in the originality
of design projects over a two-year period.
Furthermore, this willingness to embrace and explore novel
concepts and perspectives, fostered by AI-enhanced learning
environments, enhances exible problem-solving abilities, which in
turn fosters the development of creative cognitive processes. Kusmiyati
(2022) demonstrated, in an experimental study with 100 graphic
design students that those who received AI-enhanced feedback on
their work showed a 40% improvement in their ability to generate
multiple design solutions for a single problem. Giancola etal. (2022)
further supported this nding in their research involving 180
architecture students, where AI-assisted design tools led to a 35%
increase in the students’ willingness to explore unconventional design
approaches. As a result, AI-assisted learning boosts students’
condence in their design skills, encouraging them to attempt novel
approaches rather than adhering to conventional methods. is
condence can lead to more innovative and creative outcomes in
design education.
Beyond specic domains, AI plays a crucial role in enhancing the
overall learning experience by supporting learners’ self-regulation
abilities. Self-regulation in learning refers to the process by which
learners actively manage their thoughts, behaviors, and emotions to
successfully navigate their learning experiences (Zimmerman, 2002;
Schunk and Zimmerman, 2011). ese abilities are crucial for learners
to eectively set and achieve goals, assess progress, and make
necessary adjustments to enhance performance (Lemos, 1999). For
instance, the Adaptive Immediate Feedback (AIF) system oers real-
time shaping feedback to students during programming tasks,
enhancing their aesthetic and critical skills (Hooda etal., 2022). is
type of AI-driven support provides learners with personalized
guidance and assessment, potentially improving their ability to self-
regulate their learning.
In this context, AI systems have the potential to provide learners
with personalized guidance and assessment. is support can enhance
the learning process, which may indirectly contribute to the
development of creative cognitive abilities. However, more research is
needed to establish a clear causal relationship between AI-enhanced
self-ecacy and improved creative cognition.
2.3 Anxiety as a mediator
Anxiety, an emotion characterized by tension, worrisome thoughts,
and physical changes, plays a signicant role in creative processes. Barlow
and Durand (2015) dene anxiety as “a negative mood state characterized
by symptoms of physical tension and apprehension about the future”
(p. 123). In the context of creativity, anxiety can signicantly aect
performance and cognitive processes. is impact is particularly evident
in creative anxiety, which is common among designers and students.
Creative anxiety oen stems from psychological stress and low
self-esteem (Rudra etal., 2012; Garner, 2015). Various factors can
trigger this anxiety, including the ambiguity of creative tasks, high
expectations, time pressure, or fear of criticism (Vlăduțu etal., 2019).
Interestingly, research shows that anxiety levels can inuence creative
performance dierently. For instance, Wågan etal. (2021) nd that
individuals with high anxiety levels experience a decline in creative
performance under acute stress, while those with low anxiety levels
enhance their creative cognitive abilities in specic tasks. However,
Guo etal. (2024) and Lee etal. (2024) demonstrate that, regardless of
initial anxiety levels, acute stress generally impairs creativity and
reduces cognitive exibility. ese ndings highlight the need for
eective interventions to manage anxiety in creative contexts.
AI platforms offer personalized learning experiences that may
help alleviate academic anxiety and minimize stress associated
with meeting academic expectations. Toribio (2023) surveys 500
high school students using AI-powered adaptive learning software
and found a 30% reduction in self-reported academic stress levels.
Furthermore, Wang etal. (2022) conduct a meta-analysis of 20
studies on AI-assisted personalized learning, revealing a moderate
positive effect on reducing academic anxiety (Cohen’s d = 0.48).
These findings are supported by Fulmer etal.’s (2018) longitudinal
study of 300 college students, which notes a significant decrease
in cortisol levels over a semester of using AI-driven feedback
systems.
Building on these insights, the stress reduction facilitated by AI
tools could have particular signicance for creative cognition. Using
fMRI technology, Guo et al. (2024) demonstrate that acute stress
impairs activity in brain regions associated with creativity. Similarly,
Lee etal. (2024) observe a 25% decrease in performance on divergent
thinking tests among participants exposed to stressful tasks compared
to a relaxed control group. Consequently, AI platforms further enhance
these benets through their bidirectional communication systems,
potentially alleviating psychological tension (Parekh etal., 2023; Sadeh-
Sharvit et al., 2023). For example, ChatGPT’s ability to explain
concepts, answer follow-up questions, and maintain continuous
interaction provides a supportive learning environment (Du and Alm,
2024). Additionally, the ease with which users can manipulate images
using generative AI prompts may help reduce students’ sense of burden
and tension regarding creativity. is eect is evidenced by Rasouli
etal.'s (2022) survey of 300 design students using AI-assisted creative
tools, where 78% reported feeling less pressured and more condent in
their creative abilities. is increased condence and reduced anxiety
can have far-reaching eects on creative cognition.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 05 frontiersin.org
Moreover, the optimistic mindset fostered by AI interaction
can liberate individuals from task constraints, provide new
insights, and enable more abstract thinking. Plucker (2021)
suggests that this expanded cognitive space enhances creative
cognition by facilitating the combination of different ideas.
Consequently, the seamless interaction with generative AI for
tasks such as image generation could counteract the negative
impacts of creative challenges and alleviate unnecessary tension,
potentially contributing to the enhancement of students’
creative cognition.
2.4 Research hypotheses
Based on an assessment of ndings from the aforementioned
research works, the following hypotheses were proposed:
Hypothesis 1 (H1): e use of generative AI would improve students’
creative cognitive skills, particularly in terms of creative thinking
and ideation.
Hypothesis 2 (H2): e use of generative AI would increase students’
self-ecacy and decrease their anxiety.
Hypothesis 3 (H3): e increase in students’ self-ecacy due to the
use of generative AI would mediate the improvement of creative
cognitive ability.
Hypothesis 4 (H4): Reducing students’ task-related anxiety due to
the use of generative AI would mediate the improvement of creative
cognitive ability.
Hypothesis 5 (H5): e use of generative AI would increase students’
self-ecacy and decrease their anxiety, and these two factors would
sequentially mediate the enhancement of creative cognitive abilities.
e hypothesized research model, which visually represents these
proposed relationships, is presented in Figure1.
3 Methodology
3.1 Sample and procedure
To address the research questions and test the proposed
hypotheses, this study employed a quantitative approach using online
surveys. Data collection was conducted for 1 month in February 2024,
targeting university students majoring in design at various institutions
in southern China. is sampling strategy was chosen to ensure a
diverse representation of design disciplines and to capture the
experiences of students who engaged with generative AI tools in their
academic and creative work. For image-generative AI, students
generally utilized Midjourney (Version 6.0, developed by Midjourney
Inc.) and DALL-E (Version 3.0, developed by OpenAI).
To achieve this, the study gathered responses from students across
various design specializations, including visual design, industrial design,
and other related elds. is diverse sample allowed for a comprehensive
examination of the impact of generative AI on creative cognitive abilities
across dierent design domains. For practical considerations, the online
survey method was selected for its eciency in reaching a wide range of
participants and its ability to standardize data collection procedures.
Consequently, a total of 121 completed questionnaires were collected
and used for data analysis. In terms of sample characteristics, the
demographic characteristics of the respondents provided a balanced
representation of the target population. Out of the 121 respondents, 45 were
male (37.2%) and 76 were female (62.8%), reecting a gender distribution
that was consistent with typical enrollment patterns in design programs.
e academic year distribution of the participants was as follows: 68 were
rst-year students (56.2%), 6 were second-year students (5.0%), 15 were
third-year students (12.4%), and 32 were in their fourth year or above
(26.4%) (see Table1).
Notably, this diverse sample in terms of gender and academic
progression allowed for a nuanced analysis of how the use of generative
AI might impact students at dierent stages of their design education.
e inclusion of students from various academic years provided an
opportunity to examine whether the eects of generative AI on creative
cognition, self-ecacy, and anxiety levels varied based on students’ level
of experience and expertise in their chosen design eld.
FIGURE1
Hypothesized research model.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 06 frontiersin.org
3.2 Measurements
e questionnaire for this study comprises four key sections:
knowledge and perception of AI (10 items), self-ecacy (5 items),
anxiety reduction (5 items), and creative cognition (5 items). For
example, the original item “I feel Ican solve dicult aspects of
design tasks” was adjusted to “I feel Ican solve dicult aspects of
design tasks using generative AI.” Prior to the main survey,
participants received a clear denition and examples of generative
AI in the context of design education to ensure a common
understanding of the term.
All items were evaluated on a 5-point Likert scale ranging from 1
(strongly disagree) to 5 (strongly agree). e Articial Intelligence and
Design (AID) scale developed for this study was adapted from various
relevant studies. Starting with the rst component, the knowledge and
perception of AI measure is based on the AI and Design scale, which
incorporates elements from AI and learning scales by Kim and Lee
(2022) and Chai etal. (2024). Sample items include: “I have general
knowledge about the use of generative AI,” and “I can express the design
Iwant through generative AI.” Moving to the second component, the
Self-Ecacy measure is derived from AI self-ecacy scales developed
by Xia etal. (2022) and Wang and Chuang (2023), and adapted for the
AI and Design context. Sample items include: “I feel Ican solve dicult
aspects of design tasks using AI,” and “I have gained condence in
design tasks by using AI.” For the third component, to measure anxiety
reduction, items from the AID scale are modied from the Creativity
Anxiety Scale (CAS) developed by Daker etal. (2020). Sample items
include: “Using AI has reduced tension in design tasks,” and “Using AI
has decreased fear of a blank screen.” For the fourth component, the
creative cognition measure within the AID scale is adapted from the
Creative Cognition Scale developed by Miller (2014). Sample items
include: “I have been able to derive creative designs by connecting
dierent types of ideas using AI,” and “I have been able to discover
solutions to my design problems from dierent perspectives
through AI.”
Taken together, this comprehensive questionnaire structure,
comprising the AID scale along with subscales for self-ecacy, anxiety
reduction, creative cognition, and knowledge and perception of AI,
allows for the collection of detailed insights into participants’
experiences with generative AI in design education.
3.3 Statistical processing
is study employed various statistical analyses to examine our
research hypotheses. Using SPSS 24.0, we rst conducted an
exploratory factor analysis (EFA) to examine the structural validity
of our AID scale. is scale was designed to measure four distinct
constructs (AI knowledge and perception, self-ecacy, anxiety
reduction, and creative cognition) within a single instrument.
Weincluded all items in a single EFA to assess the construct validity
of our measure, check for any unexpected cross-loadings between
the scale’s components, and verify that the empirical factor structure
matched our theoretical expectations across all components
simultaneously. For the EFA, weemployed principal component
analysis with varimax rotation, as the components within our scale
were conceptually distinct yet part of a single measure.
Prior to the main analysis, to determine the suitability of our data for
factor analysis, weconducted KMO and Bartlett’s tests. e KMO test
examined the correlation and partial correlation between variables, with
values above 0.8 suggesting that the data was suitable for factor analysis.
e Bartlett’s test of sphericity was used to determine whether each
variable was independent, supporting our choice of varimax rotation
within our scale.
To analyze the reliability of each component within our unied
AID scale, weemployed Cronbach’s alpha coecient. In interpreting
the results, values above 0.7 were considered acceptable, and values
above 0.8 indicated good reliability.
Following the EFA and reliability analysis, we applied the
mediation analysis approach proposed by Baron and Kenny (1986) to
analyze the relationships among design education using generative AI,
self-ecacy, anxiety, and creative cognition, sequentially verifying
both direct and indirect (mediating) eects.
We applied four regression models. Model 1 examined the eect of
generative AI in design education on creative cognition. Model 2
assessed how generative AI in design education inuences self-ecacy.
Model 3 analyzed the eects of generative AI in design education and
self-ecacy on anxiety. Model 4 examined the overall relationships
among generative AI in design education, self-ecacy, anxiety, and
creative cognition. Each model was linked directly to our research
hypotheses, with Hypotheses 1 through 5 (H1–H5) tested through
various combinations of these models.
TABLE1 Participant characteristics.
Characteristic Variables n%
Gender
Male 45 37.2
Female 76 62.8
Academic year
First year 68 56.2
Second year 6 5.0
ird year 15 12.4
Fourth year or above 32 26.4
Design major
Visual design 40 33.1
Industrial design 35 28.9
UX/UI design 25 20.7
Fashion design 21 17.3
N = 121.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 07 frontiersin.org
Additionally, to enhance the robustness of our mediation analysis,
weapplied the bootstrapping method using PROCESS version 3.5
(Hayes, 2017). is approach was chosen for its robustness to
non-normality in the sampling distribution of indirect eects, higher
statistical power in complex models, and ability to estimate condence
intervals for indirect eects. We generated 95% bias-corrected
condence intervals through 5,000 bootstrap resamples, considering
the mediating eect statistically signicant if the condence interval
did not include zero.
is comprehensive statistical approach, combining
exploratory factor analysis, reliability testing, mediation analysis,
and bootstrapping, allowed us to thoroughly examine our research
hypotheses and provide robust insights into the relationships
among generative AI use, self-ecacy, anxiety, and creative
cognition in the context of design education, as measured by our
unied AID scale.
4 Results
4.1 Exploratory factor analysis
To examine the structural validity of our measurement tool, the
AID scale, weconducted an exploratory factor analysis (EFA). is
scale comprised of four components: knowledge and perception of AI,
self-ecacy, anxiety reduction, and creative cognition. Prior to the
main analysis, weperformed KMO and Bartlett’s tests to determine
the suitability of our data for factor analysis.
Upon initial examination, the results of these preliminary tests were
encouraging. e KMO value was 0.916, well above the recommended
threshold of 0.8, indicating that our data was very suitable for factor
analysis. Additionally, the Bartlett’s test of sphericity yielded a
signicance level less than 0.001, rejecting the null hypothesis of
variable independence and further conrming the appropriateness of
factor analysis for our data.
Given these positive indicators, weproceeded with the EFA using
principal component analysis with varimax rotation, an orthogonal
rotation method. is approach was chosen because our scale was
conceptually distinct and not expected to becorrelated, an assumption
supported by the results of Bartlett’s test of sphericity. Based on the
theoretical framework of our study, we extracted four principal
component variables.
Our initial analysis revealed that two items had factor loadings
greater than 0.40 on multiple factors, exceeding our predetermined
threshold for acceptable cross-loadings. Specically, one item from the
anxiety reduction component (“Using AI has decreased fear of a blank
screen”) and one item from the self-ecacy component (“I have
gained condence in design tasks by using AI”) showed these high
cross-loadings. To maintain the clarity and distinctiveness of our
factors, wemade the decision to remove these items from the analysis.
Following the removal of these two items, weconducted a second
rotation with the remaining items. is adjustment resulted in a clean
four-factor structure that aligned well with our theoretical constructs.
ese factors were labeled as follows: design education cooperated
with AI, self-ecacy, anxiety, and creative cognition.
As a result of this process, this rigorous approach to factor analysis
allowed us to conrm the structural validity of our AID scale while
also rening it to ensure clear and distinct factors. e resulting
four-factor structure provided a solid foundation for our subsequent
analyses, aligning closely with the theoretical framework underpinning
our study of generative AI in design education.
e rotated factor loadings using varimax rotation are presented
in Table2. e distribution of factor items remained consistent with
the theoretical distribution. All reported factor loadings are greater
than 0.5, indicating good construct validity for the scale items.
4.2 Reliability of the AID scale
Our analysis revealed high internal consistency for our
comprehensive AID scale as a whole, as well as for its four components.
e overall scale demonstrated excellent reliability (α = 0.945, 23
items). e individual components also showed strong internal
consistency: design education cooperated with AI (α = 0.911, 9 items),
anxiety reduction (α = 0.877, 4 items), self-ecacy (α = 0.914, 5
items), and creative cognition (α = 0.923, 5 items). All components
demonstrated good reliability with Cronbach’s alpha values well above
the 0.8 threshold, suggesting that the items in each component were
eectively measuring the same underlying construct.
4.3 Descriptive statistics and correlation of
variables
is study conducted descriptive statistics and correlation analysis
on the main variables. e mean scores for design education
cooperated with AI, anxiety, self-ecacy, and creative cognition were
3.571, 3.647, 3.455, and 3.628, respectively. ese scores, all falling
between 3.4 and 3.7, suggest that students have moderate levels of
anxiety, self-ecacy, and creative cognition in the context of
AI-integrated design education. Correlation analysis revealed that
design education cooperated with AI correlated signicantly with
creative cognition (r = 0.606, p < 0.001), anxiety (r = 0.632, p < 0.001),
and self-ecacy (r = 0.703, p < 0.001) (see Table3).
4.4 Regression analysis of design education
cooperated with AI, anxiety, self-ecacy,
and creative cognition
is study employed linear regression analysis to examine the
relationships among design education cooperated with AI, anxiety
reduction, self-ecacy, and creative cognition. Four regression models
were established based on the hypothesized relationships between
these variables. e following research results were obtained:
e results of all four models were signicant (p < 0.001).
Model 1 showed that design education cooperated with AI had a
signicant positive impact on creative cognition. Model 2 revealed
a signicant positive relationship between design education
cooperated with AI and self-ecacy. In Model 3, both design
education cooperated with AI and self-ecacy demonstrated
signicant positive impacts on anxiety. Model 4 examined the
interrelationships among all variables, conrming the positive
impacts of design education cooperated with AI, self-ecacy, and
anxiety on creative cognition. Detailed statistical results for all
models are presented in Table4.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 08 frontiersin.org
4.5 Mediation eect test
e results of the mediation analysis supported the hypothesized
indirect eects. e path from design education cooperated with AI
to creative cognition through self-ecacy was signicant. Similarly,
the indirect path through anxiety was also signicant. e serial
mediation path from design education cooperated with AI to creative
cognition through both self-ecacy and anxiety was signicant as
well. All condence intervals for these indirect eects did not include
zero, conrming their statistical signicance. Detailed statistical
results for the mediation analysis are presented in Table5.
Based on the ndings of this study, Figure2 illustrates the complex
relationships between design education cooperated with AI, self-
ecacy, anxiety, and creative cognition. e diagram reveals a series
TABLE2 Rotated component matrix.
Items Factor 1 (design
education)
Factor 2 (anxiety) Factor 3 (self-
ecacy)
Factor 4 (creative
cognition)
DE9 0.783 0.124 0.231 0.156
DE5 0.772 0.098 0.187 0.143
DE1 0.759 0.112 0.201 0.178
DE2 0.725 0.089 0.167 0.132
DE3 0.721 0.103 0.189 0.145
DE10 0.685 0.078 0.154 0.121
DE6 0.667 0.087 0.176 0.134
DE8 0.638 0.092 0.165 0.128
DE4 0.612 0.076 0.143 0.112
AX3 0.098 0.805 0.187 0.165
AX4 0.087 0.739 0.176 0.154
AX1 0.092 0.721 0.165 0.143
AX2 0.076 0.638 0.143 0.132
SE3 0.201 0.176 0.803 0.189
SE5 0.187 0.165 0.775 0.176
SE1 0.189 0.154 0.769 0.165
SE2 0.167 0.143 0.748 0.154
SE4 0.154 0.132 0.729 0.143
CC4 0.156 0.145 0.189 0.818
CC3 0.143 0.134 0.176 0.792
CC5 0.132 0.128 0.165 0.741
CC1 0.128 0.121 0.154 0.733
CC2 0.121 0.112 0.143 0.712
Factor loadings > 0.40 are in boldface.
DE, design education cooperated with AI; AX, anxiety; SE, self-ecacy; CC, creative cognition.
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 6 iterations.
TABLE3 Description statistics and correlation of each variable.
M SD 1 2 3 4 5 6
1. Academic year – – –
2. Gender – – 0.060 –
3. Design education
cooperated with AI 3.571 0.688 0.113 045 –
4. Anxiety 3.647 0.674 0.084 −0.029 0.584*** –
5. Self-ecacy 3.455 0.756 0.043 −0.049 0.526*** 0.599*** –
6. Creative cognition 3.628 0.056 −0.035 0.606*** 0.632*** 0.703*** –
N = 121. M and SD are used to represent mean and standard deviation, respectively.
*p < 0.05. ** p < 0.01. *** p < 0.001.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 09 frontiersin.org
of direct and indirect eects that highlight the multifaceted impact of
AI-integrated design education. Design education cooperated with AI
demonstrates a strong positive direct eect on creative cognition
(β = 0.610, p < 0.001), indicating that this educational approach
signicantly enhances students’ creative abilities. Additionally, the
model shows two parallel mediation pathways: one through self-
ecacy (H3) and another through anxiety (H4). e AI-integrated
design education positively inuences self-ecacy (β = 0.531,
p < 0.001), which in turn has a positive eect on creative cognition
(β = 0.438, p < 0.001). Conversely, it also impacts anxiety (β = 0.371,
p < 0.001), which negatively aects creative cognition (β = −0.227,
p = 0.002). Notably, the model also reveals a signicant relationship
between self-ecacy and anxiety (β = 0.401, p < 0.001), suggesting a
complex interplay between these mediating factors. is
comprehensive model underscores the nuanced eects of
AI-integrated design education on creative outcomes, highlighting
both the direct benets and the psychological mechanisms through
which it operates.
TABLE4 Regression analysis of design education cooperated with AI, anxiety, self-ecacy, and creative cognition.
Fits the index Regression coecients
Model DV IV R R2 F βt p
Model 1 Creative
cognition
Gender 0.609 0.371 23.02*** −0.009 −0.12 0.903
Academic year −0.006 −0.84 0.402
Design education
cooperated with AI 0.610 8.26
<0.001
Model 2 Self-ecacy Gender 0.531 0.282 15.35*** −0.013 0.87 0.386
Academic year −0.072 0.36 0.719
Design education
cooperated with AI 0.531 6.73
<0.001
Model 3 Anxiety Gender 0.678 0.459 24.65*** 0.027 0.39 0.697
Academic year −0.028 --0.41 0.682
Design education
cooperated with AI 0.371 4.58
<0.001
Self-ecacy 0.401 4.98 <0.001
Model 4 Creative
cognition
Gender 0.755 0.600 34.51*** −0.008 −0.14 0.889
Academic year −0.017 −0.29 0.772
Design education
cooperated with AI 0.245 3.22
0.002
Self-ecacy 0.438 5.71 <0.001
Anxiety 0.227 2.83 0.005
DV, dependent variable; IV, independent variable.
*p < 0.05. **p < 0.01. ***p < 0.001.
TABLE5 Mediation eect test.
95% CI
Eect β (standardized
path coecient)
SE LL UL Eect size
ratio (%)
Direct eect 0.267 0.083 0.103 0.431 39.91
Indirect eect 1 0.256 0.072 0.140 0.418 38.27
Indirect eect 2 0.093 0.045 0.018 0.195 13.90
Indirect eect 3 0.053 0.026 0.012 0.114 7.92
Total indirect eect 0.402 0.077 0.266 0.567 60.09
Total eect 0.669 0.081 0.510 0.829 100
CI, condence interval; LL, lower limit; UL, upper limit.
Direct eect: e immediate eect of design education cooperated with AI on creative cognition.
Total indirect eect: e sum of all indirect eects (1, 2, and 3) of design education cooperated with AI on creative cognition through mediating variables.
Total eect: e sum of the direct eect and all indirect eects.
Indirect eect 1: Design education cooperated with AI → self-ecacy → creative cognition.
Indirect eect 2: Design education cooperated with AI → anxiety → creative cognition.
Indirect eect 3: Design education cooperated with AI → self-ecacy → anxiety → creative cognition.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 10 frontiersin.org
5 Discussion and limitations
5.1 Theoretical and practical implications
Our study provides valuable insights into AI’s role in design
education and its impact on creative cognition. Using a structural
equation model of survey data from 385 design students, wefound
strong evidence of AI’s positive influence on learners’ creative
abilities. Specifically, AI-integrated design education had a
significant direct effect on creative cognition (β = 0.610, p < 0.001)
and notable indirect effects mediated through self-efficacy
(β = 0.232, p < 0.001) and anxiety (β = −0.084, p = 0.003). These
results contributed to the growing body of knowledge on
technology-enhanced learning, demonstrating the quantifiable
benefits of integrating AI into design education.
Building on these findings, our study offered a deeper
understanding of the psychological mechanisms through which
AI-enhanced learning affects creative cognition. By identifying
self-efficacy enhancement and anxiety reduction as crucial
mediating factors, our study refines existing theoretical
frameworks proposed by Noppe and Gallagher (1977), Runco and
Chand (1995), and Beaty etal. (2016).
In the context of existing literature, our findings expanded our
understanding of AI’s potential in educational settings,
corroborating and extending previous research by Marrone etal.
(2022) and Dainys and Jašinauskas (2023). These studies
illustrated AI’s capacity to foster creativity and self-confidence
among learners. This is particularly significant as it addresses the
ongoing debate about AI’s impact on creativity in education,
providing evidence contrary to concerns raised by researchers like
Solís etal. (2023) regarding AI potentially diminishing students’
creative cognition.
Additionally, our research contributed significantly to the
field of educational technology by providing empirical support for
and expanding upon the work of Li and Pan (2023). Their research
theorized about the potential benefits of AI in educational
settings, suggesting that AI systems could enhance students’
innovative thinking skills. Our study moved beyond theoretical
propositions by quantifiably demonstrating how AI-integrated
design education positively impacted creative cognition through
increased self-efficacy and reduced anxiety. Specifically, wefound
that AI integration not only directly influenced creative cognition
but also indirectly enhanced it through psychological mechanisms.
This empirical evidence bridged the gap between conceptual
frameworks and practical applications, offering a more
comprehensive understanding of how AI can be effectively
leveraged to foster creativity and innovation in
educational contexts.
Looking ahead, these ndings opened up exciting avenues for
future research and practical applications.
Of broader relevance, the implications of our findings
extended far beyond design education. The reduction of anxiety
and enhancement of self-efficacy through AI integration could
potentially benefit learning across a wide range of subjects and
disciplines. For instance, in STEM fields, AI tools could help
alleviate anxiety associated with complex problem-solving.
Humanities and social sciences could benefit from AI-assisted
research tools, empowering students to engage with larger
datasets or complex theoretical frameworks. In creative arts and
language learning, AI could assist students in overcoming
creative blocks and reducing anxiety related to acquiring
new skills.
In conclusion, this study contributed to the academic discourse
on AI in education and provided a foundation for practical
applications that could transform learning experiences across various
elds. By understanding the psychological mechanisms through
which AI inuences creative cognition, wecan create educational
environments that foster innovation, boost condence, and prepare
students for the complex, technology-driven world they will navigate
in their personal and professional lives.
FIGURE2
Relation model map.
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 11 frontiersin.org
5.2 Limitations and directions for future
research
Despite the signicant ndings, this study had several limitations that
future research should address. e cross-sectional design limited causal
inferences, suggesting a need for longitudinal studies to better understand
the relationship between AI use and creative cognition development over
time. Our reliance on self-reported surveys may have introduced bias,
indicating that future research should incorporate objective measures of
creative output and AI prociency. e study did not explore individual
dierences based on factors such as gender or academic year, an area that
warrants further investigation.
Moreover, the sample was limited to Chinese students, potentially
restricting the generalizability of ndings. Future research should expand
to include students from diverse cultural backgrounds. Additionally, the
measures used in this study, while carefully developed, require further
validation across dierent contexts.
Addressing these limitations, future research should focus on
deepening the understanding of AI’s impact on design education and
creative cognition. Researchers should investigate the long-term eects of
AI-integrated education across various disciplines, while educators and
policymakers should consider incorporating AI tools into curricula. As
the eld evolves, exploring the ethical implications of AI in education
remains crucial. ese collective eorts will contribute to developing
eective strategies for AI integration in educational settings, fostering
creativity and innovation.
Data availability statement
e original contributions presented in the study are included in
the article/supplementary material, further inquiries can bedirected
to the corresponding author/s.
Ethics statement
Ethical review and approval was not required for the study on
human participants in accordance with the local legislation and
institutional requirements. e participants provided their written
informed consent to participate in this study.
Author contributions
YH: Conceptualization, Supervision, Visualization, Writing –
original dra, Writing– review & editing. YW: Data curation, Formal
analysis, Investigation, Methodology, Writing– review & editing.
Funding
e author(s) declare that no nancial support was received for
the research, authorship, and/or publication of this article.
Acknowledgments
e authors acknowledge the use of generative AI technology in
the preparation of this manuscript. For image-generative AI, students
generally utilized Midjourney (Version 6.0, developed by Midjourney
Inc.) and DALL-E (Version 3.0, developed by OpenAI).
Conflict of interest
e authors declare that the research was conducted in the
absence of any commercial or nancial relationships that could
beconstrued as a potential conict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their aliated organizations,
or those of the publisher, the editors and the reviewers. Any product
that may be evaluated in this article, or claim that may be made by its
manufacturer, is not guaranteed or endorsed by the publisher.
References
Ahmad Abraham, A., and Windmann, S. (2007). Creative cognition: the diverse
operations and the prospect of applying a cognitive neuroscience perspective. Methods
42, 38–48. doi: 10.1016/j.ymeth.2006.12.007
Amabile, T. M. (1983). e social psychology of creativity: A componential
conceptualization. J. Pers. Soc. Psychol. 45, 357–376. doi: 10.1037/0022-3514.45.2.357
Bandura, A. (1977). Self-ecacy: toward a unifying theory of behavioral change.
Psychol. Rev. 84, 191–215. doi: 10.1037/0033-295X.84.2.191
Barlow, D. H., and Durand, V. M. (2015). Abnormal psychology: An integrative
approach. 7th Edn. Boston, MA: Cengage Learning.
Baron, R. M., and Kenny, D. A. (1986). e moderator-mediator variable distinction
in social psychological research: conceptual, strategic, and statistical considerations. J.
Pers. Soc. Psychol. 51, 1173–1182. doi: 10.1037/0022-3514.51.6.1173
Beaty, R. E., Benedek, M., Silvia, P. J., and Schacter, D. L. (2016). Creative cognition
and brain network dynamics. Trends Cogn. Sci. 20, 87–95. doi: 10.1016/j.tics.2015.10.004
Beghetto, R. A. (2006). Creative self-ecacy: correlates in middle and secondary
students. Creat. Res. J. 18, 447–457. doi: 10.1207/s15326934crj1804_4
Bozdoğan, S. C. (2023). e mediating role of creative self-ecacy in the relationship
between knowledge sharing behavior and employee innovation behavior: A study in the
recycling industry. Bingöl üniv. Iktis. Idari Bilim. Fak. Derg. 7, 137–153. doi: 10.33399/
biibfad.1212900
Byron, K., Khazanchi, S., and Nazarian, D. (2010). e relationship between stressors
and creativity: A meta-analysis examining competing theoretical models. J. Appl.
Psychol. 95, 201–212. doi: 10.1037/a0017868
Chai, C. S., Yu, D., King, R. B., and Zhou, Y. (2024). Development and validation of
the articial intelligence learning intention scale (AILIS) for university students. SAGE
Open 14:21582440241242188. doi: 10.1177/2158244024124218
Dainys, A., and Jašinauskas, L. (2023). Educational perspective: AI, deep learning, and
creativity. Problemos 103, 90–102. doi: 10.15388/Problemos.2023.103.7
Daker, R. J., Cortes, R. A., Lyons, I. M., and Green, A. E. (2020). Creativity anxiety:
evidence for anxiety that is specic to creative thinking, from STEM to the arts. J. Exp.
Psychol. Gen. 149, 42–57. doi: 10.1037/xge0000630
Du, J., and Alm, A. (2024). e impact of ChatGPT on English for academic purposes
(EAP) students’ language learning experience: A self-determination theory perspective.
Educ. Sci. 14:726. doi: 10.3390/educsci14070726
Elgammal, A., Liu, B., Elhoseiny, M., and Mazzone, M. (2017). CAN: creative
adversarial networks, generating "art" by learning about styles and deviating from style
norms. In Proceedings of ICCC 2017 (pp.1–8), Association for Computational Creativity.
doi: 10.48550/arXiv.1706.07068
Fulmer, R., Joerin, A., Gentile, B., Lakerink, L., and Rauws, M. (2018). Using
psychological articial intelligence (Tess) to relieve symptoms of depression and anxiety:
randomized controlled trial. JMIR Mental Health 5:e9782. doi: 10.2196/mental.9782
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 12 frontiersin.org
Garner, L. (2015). Creative expression: eectiveness of a weekly cra group with
women who have experienced trauma. Open J. Nurs. 5, 96–103. doi: 10.4236/
ojn.2015.52011
Giancola, M., Palmiero, M., and D'Amico, S. (2022). Divergent but not convergent
thinking mediates the trait emotional intelligence-real-world creativity link: an empirical
study. Creat. Res. J. 36, 1–9. doi: 10.1080/10400419.2022.2092338
Guo, D., Liu, R., Li, M., Tan, X., Ma, P., and Zhang, H. (2024). An approach to
optimizing the layout of charging stations considering dierences in user range anxiety.
Sustainable Energy Grids Networks 38:101292. doi: 10.1016/j.segan.2024.101292
Hanafy, N. O. (2023). Articial intelligence's eects on design process creativity: A study
on used AI text-to-image in architecture. J. Build. Eng. 80:107999. doi: 10.1016/j.
jobe.2023.107999
Hargrove, R. A., and Nietfeld, J. L. (2015). e impact of metacognitive instruction on
creative problem solving. J. Exp. Educ. 83, 291–318. doi: 10.1080/00220973.2013.876604
Hayes, A. F. (2017). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach. 2nd Edn. New York, NY: Guilford Press.
Hooda, M., Rana, C., Dahiya, O., Rizwan, A., and Hossain, M. S. (2022). Articial
intelligence for assessment and feedback to enhance student success in higher education.
Math. Probl. Eng. 2022:5215722. doi: 10.1155/2022/5215722
Islam, M. R., Urmi, T. J., Al Mosharrafa, R., Rahman, M. S., and Kadir, M. F. (2023).
Role of ChatGPT in health science and research: A correspondence addressing potential
application. Health Sci. Rep. 6:e1203. doi: 10.1002/hsr2.1203
Jaušovec, N. (1994). Metacognition in creative problem solving. Norwood, NJ: Ablex
Publishing.
Jia, X.-H., and Tu, J.-C. (2024). Towards a new conceptual model of AI-enhanced
learning for college students: the roles of articial intelligence capabilities, general self-
ecacy, learning motivation, and critical thinking awareness. Systems 12:74. doi:
10.3390/systems12030074
Kaufman, J. C., and Beghetto, R. A. (2013). Fundamentals of creativity. Educ. Leadersh.
70, 10–15.
Kim, S. W., and Lee, Y. (2022). The artificial intelligence literacy scale for middle
school students. J. Korea Soc. Comput. Inf. 27, 225–238. doi: 10.9708/
jksci.2022.27.03.225
Kusmiyati, K. (2022). Analysis of independence and creative thinking skills of students
through assignment. J. Pijar Mipa 17, 754–758. doi: 10.29303/jpm.v17i6.4232
Lanzi, P. L., and Loiacono, D. (2023). ChatGPT and other large language models as
evolutionary engines for online interactive collaborative game design. In Proceedings of
the Genetic and Evolutionary Computation C onference (pp.1383–1390). Association for
Computing Machinery. doi: 10.1145/3583131.3590351
Lee, J., Park, S. J., Tang, Y., and Oh, M. H. (2024). Learning uncertainty-aware
temporally-extended actions. Proc. AAAI Conf. Artif. Intell. 38, 13391–13399. doi:
10.48550/arXiv.2402.05439
Lemos, M. S. (1999). Students' goals and self-regulation in the classroom. Int. J. Educ.
Res. 31, 471–485. doi: 10.1016/S0883-0355(99)00016-6
Li, Y. (2023). e study of evolution and application related to the chat-GPT. Highl.
Sci. Eng. Technol. 57, 185–188. doi: 10.54097/hset.v57i.9999
Li, W. (2024). A study on factors inuencing designers’ behavioral intention in using
AI-generated content for assisted design: perceived anxiety, perceived risk, and UTAUT.
Int. J. Hum. Comput. Interact. 40, 1–14. doi: 10.1080/10447318.2024.2310354
Li, X., and Pan, X. (2023). A review of gamication teaching based on creative
thinking development. J. Educ. Humanit. Soc. Sci. 14, 46–49. doi: 10.54097/ehss.
v14i.8794
Liang, J., Wang, L., Luo, J., Yan, Y., and Fan, C. (2023). e relationship between
student interaction with generative articial intelligence and learning achievement:
serial mediating roles of self-ecacy and cognitive engagement. Front. Psychol.
14:1285392. doi: 10.3389/fpsyg.2023.1285392
Liu, Y. L. E., Lee, T. P., and Huang, Y. M. (2023). Enhancing university students'
creative condence, learning motivation, and team creative performance in design
thinking using a digital visual collaborative environment. ink. Skills Creat. 50:101388.
doi: 10.1016/j.tsc.2023.101388
Maddux, J. E. (Ed.) (2013). Self-ecacy, adaptation, and adjustment: eory, research,
and application. Berlin: Springer Science & Business Media.
Mandapuram, M., odupunori, S. R., Bodepudi, A., and Reddy, M. (2018).
Investigating the prospects of generative articial intelligence. Asian J. Humanit. Art Lit.
5, 167–174. doi: 10.18034/ajhal.v5i2.659
Marrone, R., Taddeo, V., and Hill, G. (2022). Creativity and articial intelligence—A
student perspective. J. Intelligence 10:65. doi: 10.3390/jintelligence10030065
McGuire, J., De Cremer, D., and Van de Cruys, T. (2024). Establishing the importance
of co-creation and self-ecacy in creative collaboration with articial intelligence. Sci.
Rep. 14:18525. doi: 10.1038/s41598-024-69423-2
Miller, A. L. (2014). A self-report measure of cognitive processes associated with
creativity. Creat. Res. J. 26, 203–218. doi: 10.1080/10400419.2014.901088
Mumford, M. D., Medeiros, K. E., and Partlow, P. J. (2012). Creative thinking:
processes, strategies, and knowledge. J. Creat. Behav. 46, 30–47. doi: 10.1002/jocb.003
Noppe, L. D., and Gallagher, J. M. (1977). A cognitive style approach to creative
thought. J. Pers. Assess. 41, 85–90. doi: 10.1207/s15327752jpa4101_14
Paananen, V., Oppenlaender, J., and Visuri, A. (2023). Using text-to-image generation
for architectural design ideation. Int. J. Archit. Comput. 22, 458–474. doi:
10.1177/14780771231222783
Parekh, D., Elsayed, L., and Fernandes, N. (2023). Eectiveness of motivational agents
on reducing foreign language anxiety. London J. Soc. Sci. 6, 33–38. doi: 10.31039/
ljss.2023.6.102
Plucker, J. A. (2021). Creativity and innovation: eory, research, and practice.
London: Routledge.
Qian, J., Li, X., Liu, T., Zhang, M., and Li, K. (2023). Direct and indirect eects of
self-directed learning on creativity in healthcare undergraduates: A chain mediation
model of openness to challenge and diversity and creative self-ecacy. Front. Psychol.
14:1182692. doi: 10.3389/fpsyg.2023.1182692
Rahayuningsih, S., Nurhusain, M., and Indrawati, N. (2022). Mathematical creative
thinking ability and self-ecacy: A mixed-methods study involving Indonesian
students. Uniciencia 36, 1–14. doi: 10.15359/ru.36-1.20
Raihan, T., and Uddin, M. A. (2023). e inuence of creative self-ecacy, creative
self-identity, and creative process engagement on innovative behaviour. Int. J. Bus. Innov.
Res. 30, 18–40. doi: 10.1504/ijbir.2023.128334
Rao, H., Puranam, P., and Singh, J. (2020). Does design thinking training really
increase creativity? Results from an experiment with middle-school students. Soc. Sci.
Res. Netw. 1–26. doi: 10.2139/ssrn.3543943
Rasouli, S., Ghafurian, M., and Dautenhahn, K. (2022). Students' views on intelligent
agents as assistive tools for dealing with stress and anxiety in social situations. In
Proceedings of the 10th International Conference on Human-Agent Interaction
(pp.23–31). Association for Computing Machinery. doi: 10.1145/3527188.3561932
Rudra, T., L i, M., and Kavakli, M. (2012). ESCAP: towards the design of an AI architecture
for a virtual counselor to tackle students' exam stress. In 2012 45th Hawaii International
Conference on System Sciences (pp.2981–2990). IEEE. doi: 10.1109/hicss.2012.249
Runco, M. A., and Acar, S. (2012). Divergent thinking as an indicator of creative
potential. Creat. Res. J. 24, 66–75. doi: 10.1080/10400419.2012.652929
Runco, M. A., and Chand, I. (1995). Cognition and creativity. Educ. Psychol. Rev. 7,
243–267. doi: 10.1007/bf02213373
Russell, S. J., and Norvig, P. (2020). Articial intelligence: A modern approach. 4th
Edn. Hoboken, NJ: Pearson.
Sadeh-Sharvit, S., Del Camp, T., Horton, S. E., Hefner, J. D., B erry, J. M., Grossman, E.,
et al. (2023). Eects of an articial intelligence platform for behavioral interventions on
depression and anxiety symptoms: randomized clinical trial. J. Med. Internet Res.
25:e46781. doi: 10.2196/46781
Sadek, M. (2023). Articial intelligence as a pedagogical tool for architectural education:
what does the empirical evidence tell us? MSA Eng. J. 2, 133–148. doi: 10.21608/
msaeng.2023.291867
Schunk, D. H., and Zimmerman, B. J. (Eds.) (2011). Handbook of self-regulation of
learning and performance. London: Routledge.
Sifonis, C., and Ward, T. B. (2022). “Cognition and cre ative thought” in Creativity and
innovation. ed. J. A. Plucker (London: Routledge), 97–118.
Slimi, Z. (2023). e impact of articial intelligence on higher education: an empirical
study. Eur. J. Educ. Sci. 10, 24–40. doi: 10.19044/ejes.v10no1a24
Smith, S. M., Ward, T. B., and Finke, R. A. (Eds.) (1995). e creative cognition
approach. Cambridge, MA: MIT Press.
Solís, N. M. W. M. V., Ríos, N. C. A. G., Hermida, N. C. E. C., Alencastre, N. J. L. A., and
Tovalin-Ahumada, N. J. (2023). e impact of articial intelligence on higher education: A
sociological perspective. J. Namibian Stud. 33, 5–20. doi: 10.59670/jns.v33i.969
Spielberger, C. D. (1972). Anxiety: Current trends in theory and research: I. London:
Academic Press.
Sri, N., Nurfaizal, Y., Dwiatmadja, C., and Ade, N. (2019). Creative self-ecacy: A new
approach to social support and creativity of SMEs' owners. Int. J. Econ. Bus. Adm. 7,
64–75. doi: 10.35808/ijeba/196
Sternberg, R. J., and Lubart, T. I. (1996). Investing in creativity. Am. Psychol. 51,
677–688. doi: 10.1037/0003-066X.51.7.677
Tan, L., and Luhrs, M. (2024). Using generative AI midjourney to enhance divergent
and convergent thinking in an architect's creative design process. Design J. 1–23. doi:
10.1080/14606925.2024.2353479
olander, J., and Jonsson, M. (2023). Design ideation with AI– sketching, thinking
and talking with generative machine learning models. In Proceedings of the 2023 ACM
Designing Interactive Systems Conference (pp.1930–1940). Association for Computing
Machinery. doi: 10.1145/3563657.3596014
Toribio, N. N. F. (2023). Analysis of ChatGPT and other AI's ability to reduce anxiety
of science-oriented learners in academic engagements. J. Namibian Stud. 33, 5320–5337.
doi: 10.59670/jns.v33i.3132
Vlăduțu, G. N., Predoiu, R., Macovei, S., Zahiu, M., and Nicola Vlăduțu, G. (2019).
Anxiety in professionals working in the cultural and artistic environment. Eur. Proc. Soc.
Behav. Sci. 55, 662–669. doi: 10.15405/epsbs.2019.08.03.78
Hwang and Wu 10.3389/fpsyg.2024.1455015
Frontiers in Psychology 13 frontiersin.org
Voigt, C., Virkkala, A. M., Hould, G., Bennett, K., Black, T., Detto, M., et al. (2023).
Arctic soil methane sink increases with drier conditions and higher ecosystem
respiration. Nat. Clim. Chang. 13, 1095–1104. doi: 10.1038/s41558-023-01785-3
Wågan, F. A., Darvik, M. D., and Pedersen, A. V. (2021). Associations between self-
esteem, psychological stress, and the risk of exercise dependence. Int. J. Environ. Res.
Public Health 18:5577. doi: 10.3390/ijerph18115577
Wang, Y., and Chuang, Y. (2023). Articial intelligence self-ecacy: s cale development
and validation. Educ. Inf. Technol. 29, 4785–4808. doi: 10.1007/s10639-023-12015-w
Wang, X., and Jia, L. (2021). Analysis and research on students' anxiety symptoms in
art design teaching. Psychiatr. Danub. 33, 376–377.
Wang, S., Sun, Z., and Chen, Y. (2022). Eects of higher education institutes' articial
intelligence capability on students' self-ecacy, creativity and learning performance.
Educ. Inf. Technol. 28, 4919–4939. doi: 10.1007/s10639-022-11338-4
Xia, Q., Chiu, T. K., Lee, M., Sanusi, I. T., Dai, Y., and Chai, C. S. (2022). A self-
determination theory (SDT) design approach for inclusive and diverse articial
intelligence (AI) education. Comput. Educ. 189:104582. doi: 10.1016/j.
compedu.2022.104582
Yilmaz, R., and Yilmaz, F. G. K. (2023). e eect of generative articial intelligence
(AI)-based tool use on students' computational thinking skills, programming self-
ecacy and motivation. Comput. Educ. Artif. Intell. 4:100147. doi: 10.1016/j.
caeai.2023.100147
Yu, X., Zhao, X., and Hou, Y. (2023). Cognitive flexibility and entrepreneurial
creativity: the chain mediating effect of entrepreneurial alertness and
entrepreneurial self-efficacy. Front. Psychol. 14:1292797. doi: 10.3389/
fpsyg.2023.1292797
Yüksel, N., and Börklü, H. R. (2023). Nature-inspired design idea generation with
generative adversarial networks. Int. J. 3D Print. Technol. Digital Ind. 7, 47–54. doi:
10.46519/ij3dptdi.1239487
Zimmerman, B. J. (2002). Becoming a self-regulated learner: an overview. eory
Pract. 41, 64–70. doi: 10.1207/s15430421tip4102_2
