You are on page 1of 17

Asia Pacific Journal of Education

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/cape20

Expert or machine? Comparing the effect of


pairing student teacher with in-service teacher
and ChatGPT on their critical thinking, learning
performance, and cognitive load in an integrated-
STEM course

Tingting Li, Yu Ji & Zehui Zhan

To cite this article: Tingting Li, Yu Ji & Zehui Zhan (2024) Expert or machine? Comparing the
effect of pairing student teacher with in-service teacher and ChatGPT on their critical thinking,
learning performance, and cognitive load in an integrated-STEM course, Asia Pacific Journal of
Education, 44:1, 45-60, DOI: 10.1080/02188791.2024.2305163

To link to this article: https://doi.org/10.1080/02188791.2024.2305163

Published online: 21 Jan 2024.

Submit your article to this journal

Article views: 242

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=cape20
ASIA PACIFIC JOURNAL OF EDUCATION
2024, VOL. 44, NO. 1, 45–60
https://doi.org/10.1080/02188791.2024.2305163

Expert or machine? Comparing the effect of pairing student


teacher with in-service teacher and ChatGPT on their critical
thinking, learning performance, and cognitive load in an
integrated-STEM course
Tingting Li , Yu Ji and Zehui Zhan
School of Information Technology in Education, South China Normal University, Guangzhou, China

ABSTRACT ARTICLE HISTORY


For student teachers’ professional development, the emergence of gen­ Received 4 August 2023
erative artificial intelligence (AI) represents both opportunity and chal­ Accepted 2 January 2024
lenge. This exploratory quasi-experimental study aims to investigate the KEYWORDS
effects of “Human-Human” and “Human-Machine” collaborative learning STEM education;
approaches on the SETM teaching training performance of student tea­ collaborative learning;
chers. Twenty-three student teachers were divided into two groups within ChatGPT; student teacher;
a single class, each adopting one learning method. The experiment lasted AI-assisted learning
for two months with weekly three-hour sessions. Data were analysed
focusing on critical thinking, learning performance, and cognitive load
between the groups. The results indicated that student teachers using
ChatGPT showed higher critical thinking systematicity, task completion
efficiency, and experienced lower cognitive load. Student teachers paired
with in-service teachers slightly outperformed those with ChatGPT on the
final teaching design proposal. These findings underscore the potential
and varying strengths of AI tools like ChatGPT and human teachers. For
further research, refined collaborative learning scaffolding are recom­
mended to explore the impact and potential of AI-assisted and in-
service teacher-involved collaboration. The study’s implications could
guide educators, policymakers, and AI developers in optimizing the AI-
enhanced collaborative learning strategies and shed light on the new
formation of human-machine collaborative intelligence in the scope of
education.

Introduction
In recent decades, there has been a shift in educational priorities, with STEM education being
integrated at a strategic level by numerous countries (Zhan et al., 2022a; Zhan & Niu, 2023).
Integrated STEM (iSTEM) education refers to the STEM themed learning that is formed through
active consultation among multiple disciplines due to the inter-related content or methodology
(Kelley & Knowles, 2016). Learners are expected to go beyond the perspective of a single discipline,
integrating knowledge, methodologies, and ways of thinking from at least two disciplines to under­
stand and address complex problems in real-world scenarios (English, 2016; Haas et al., 2021; Zhan
et al., 2022b). In this context, conducting iSTEM teaching is a complex and challenging task,
necessitating teachers to have both professional theoretical knowledge and practical skills in STEM

CONTACT Zehui Zhan zhanzehui@m.scnu.edu.cn School of Information Technology in Education, South China Normal
University, No.55, West of hongshan Avenue, Tianhe District, Guangzhou 510631, China
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
© 2024 National Institute of Education, Singapore
46 T. LI ET AL.

education (Haas et al., 2021). The availability of well-qualified teachers is deemed essential for the
realization of this vision, driving universities and educational departments to enhance the training of
student (or pre-service) teachers to meet the real teaching demands and teach iSTEM topics with
high-level STEM pedagogical expertise (Chai, 2019; Morris et al., 2021).
Given these challenges, integrating generative AI into education, such as the conversational
chatbot ChatGPT developed by OpenAI, presents a promising solution for teacher training. In
iSTEM education, ChatGPT can participate as a conversationalist, content provider, assistant, and
evaluator (Lo, 2023), offering a variety of functions including behaviour detection, knowledge
transfer, question generation, performance assessment, tutoring, and resource recommendation
(Xu & Ouyang, 2022). ChatGPT, as an accelerator for student teachers to understand STEM teaching
knowledge and solve problems (Ji et al., 2023), can facilitate interdisciplinary teaching innovation
(Dwivedi et al., 2023). Clearly, the integration of AI in education highlights the evolving collaborative
learning relationship between humans and machines (Xu & Ouyang, 2022; Zhang & Aslan, 2021).
However, caution is needed regarding the potential risks and challenges posed by generative AI,
such as the production of incorrect or false information (Cooper, 2023), circumventing plagiarism
detectors (Lo, 2023), causing adversarial impacts on cognitive autonomy, weakening individual
agency, and leading to superficial and homogenized thinking (Rani et al., 2023). Therefore, the
effective application of ChatGPT in teacher education systems requires careful consideration.
Teachers and students should possess a certain level of digital literacy and ethical standards to
face these challenges (Krügel et al., 2023) and devise appropriate strategies for integrating AI tools
into the curriculum (Rudolph et al., 2023; Tan, 2022).
Besides, collaborating with experienced teachers is also a key approach to enriching the prepara­
tion of student STEM teacher candidates. These interactions enable student teachers to more
effectively grasp the dynamics of real classroom environments (K. Huang et al., 2011), and help
bridge the often-seen theory-practice gap in teacher education (Brown, 2017). Moreover, in an era of
rapid technological advancement, the value of human interaction has been underscored (Guzman &
Lewis, 2020). Such interactions significantly foster the identity formation of student teachers, reduce
alienation among practitioners, and aid in the intergenerational transmission of the teaching
profession (Estapa & Tank, 2017; Geeraerts et al., 2018; Harlow & Cobb, 2014; Näykki et al., 2021).
While practitioner pairing learning can offer rich, context-sensitive insights, practical constraints such
as availability, power dynamics, depth of collaboration, sustainability, and scalability pose significant
challenges (Näykki et al., 2021). The effectiveness of this expert practitioner-supported learning
model, as compared to AI-driven learning, remains an unresolved question in the context of STEM
teacher training.
Recognizing the potential benefits and challenges of AI-assisted learning and collaboration with
in-service teachers, this study aims to compare the effects of “human-human” and “human-machine”
collaborative learning approaches on the performance of student STEM teacher training pro­
grammes through quasi-experimental research, elucidating their respective potentials and
shortcomings.

Literature review
For STEM teachers, they need to cultivate key competences which are critical to the success of
teaching iSTEM content more efficiently, like cognition (e.g., knowledge of STEM education), teach­
ing plan design and implementation (e.g., collaborate with teachers of other subjects in building and
assigning STEM topics) (X. Chen, 2020; Dare et al., 2018; Shernoff et al., 2017; Song & Zhou, 2021).
Besides, in STEM education, cultivating 21st-century skills, especially critical thinking, highlights the
crucial need for enhancing teachers’ proficiency in critical thinking (e.g., by framing and stimulating
about why they do what they do) (Song & Zhou, 2021; Tiatri et al., 2023). At the same time, given the
complexity of STEM education, it is essential to consider the challenges in teachers’ professional
learning to ensure an appropriate cognitive load (Timothy et al., 2023; Wu et al., 2022). This is crucial
ASIA PACIFIC JOURNAL OF EDUCATION 47

for enhancing STEM teaching resilience and teacher retention. Accordingly, for STEM teachers,
collaboration has become the most commonly reported method of STEM teacher training (B.
Huang et al., 2022; Willegems et al., 2017). Effective collaboration enhances STEM teachers’ knowl­
edge, abilities, and emotional performance in STEM education (X. Chen, 2020; Song & Zhou, 2021).
With the advancement of AI technology, the scope of collaboration partners has expanded from
humans to machines. Based on this, the following text will review existing research on collaboration
between humans and human-machine collaborative learning.

AI-assisted learning in student teacher training


In the era of Industry 4.0, the integration of AI into the educational landscape has been extended to
student teacher training (Carrillo & Flores, 2020; Lo, 2023). ChatGPT, representing conversational AI,
can simulate various human emotions and tones and understanding and responding to complex
requests in a near-natural manner (King & ChatGPT, 2023). Compared to traditional chatbots,
ChatGPT can be integrated through API into various platforms and tools, enabling real-time inter­
actions with student teachers. Consequently, there is emerging evidence suggests that AI tools,
exemplified by ChatGPT, can be effectively integrated into student teacher training as expert
mentors, intelligent study companions, and smart assistants, enhancing learning across various
environments (Lo, 2023; Stokel-Walker, 2022).
For example, ChatGPT can enhance the understanding of STEM subject knowledge, pedagogical
content knowledge, and teaching methodologies (Patel & Lam, 2023). Playing the role of a one-to-
one tutor, it acts as an accelerator of knowledge dissemination. Incorporating ChatGPT’s expertise
allows for real-time negotiations based on student teachers’ unique requirements. By processing
relevant data, and generating cross-modal teaching resources (Cooper, 2023), ChatGPT provides
pertinent cognitive feedback derived from data-driven insights, helping reduce learners’ low-order
cognitive load. Additionally, ChatGPT can assist teachers in generating cross-modal teaching
resources, such as creating instructional designs and templates consistent with STEM themes. This
provides particular value for navigating the nuances of STEM teaching practices. It equips student
teachers to efficiently filter and assimilate information, sparking discussions that foster divergent
thinking (Organisciak et al., 2023). This, in turn, frees up cognitive resources for higher-order
cognitive activities (Farrokhnia et al., 2023). It can be observed that the convenience and effective­
ness brought by ChatGPT in assisting STEM student teacher training have been recognized by
education professionals.
However, the integration of AI in student teacher training is not without its challenges (X. Chen
et al., 2020; X. Chen et al., 2022). Concerns have also been raised, including the ability of AI to cater to
the relational facets of teaching, such as empathy and emotional understanding (Taber, 2019).
Questions also persist about ethical and equity issues, as not all student teachers might have
equal access to or familiarity with such advanced technologies. Additionally, the effective integration
of AI tools into the curriculum remains a complex issue that requires careful planning and considera­
tion (J. Kim et al., 2022).
Overall, at present, there is no unified conclusion on ChatGPT’s impact on individual learning, and
some studies may overstate its effects. Moreover, the potential negative impact of pre-service
teachers’ dependency on ChatGPT for learning outcomes requires insightful conclusions. These
areas warrant further investigation in future studies.

Collaboration with in-service teachers in student STEM teacher’s learning


Student STEM teachers face real challenges in teacher training programmes, including limited
disciplinary or interdisciplinary knowledge, difficulties in adapting to pedagogical methods and
classroom roles, lack of time and expertise in curriculum integration, and the absence of knowledge
regarding relevant standards and assessments (Chai, 2019; B. Huang et al., 2022). Besides, the current
48 T. LI ET AL.

teacher education system, largely dominated by universities, has been criticized for various issues
such as a detachment from practical experience, arbitrary assembly of teacher education courses,
fragmented and incoherent content, and a lack of connection with primary and secondary schools
(Kostiainen et al., 2018; Radloff & Guzey, 2016). Traditional student teacher and in-service teacher
professional development programmes often exacerbate this problem by emphasizing the devel­
opment of teachers’ skills and knowledge in one domain, rather than preparing them for interdisci­
plinary domains (Brown, 2017; Chai, 2019; Honey et al., 2014; B. Huang et al., 2022; Lo, 2021).
Neglecting targeted training for iSTEM teaching capabilities, it leads to new teachers finding
themselves ill-equipped to handle STEM teaching once they enter the profession. Consequently,
teachers have expressed concerns about their readiness to carry out STEM education, particularly
iSTEM (Lau & Jong, 2022). To overcome these barriers, the cross-boundary integration of multiple
resources for STEM teacher education is emerging as a promising solution (Kelley et al., 2020; B. Kim
& Bastani, 2017). It has been suggested that higher education institutions must bolster their alliances
with primary and secondary schools to address present challenges. This form of collaboration can
expose student teachers to the real-world dynamics of a classroom, aiding them in understanding
the practical implications of their academic learning (Schmid & Hegelheimer, 2014). In-service
teachers, with their hands-on experience and insights, can provide student teachers with valuable
mentorship (Sá & de Almeida, 2016). This mentorship can take several forms, including providing
advice on teaching strategies, classroom management, student assessment, and navigating the
broader educational landscape (Näykki et al., 2021).
However, collaborative learning between student teachers and in-service teachers is not
without its difficulties. Among these are the power imbalance between the two groups,
cognitive differences obstructing consensus in collaborative learning, and the necessity for
an intermediary mechanism to foster cooperation (Li et al., 2023; Meschede et al., 2017;
Willegems et al., 2017). Moreover, the conflict between work and learning constraints often
undermines the potential for sustained, in-depth collaboration. In addition, existing research
is mostly conducted outside the formal university classroom, the collaborative learning of
student teachers or teachers is mostly based on peer collaboration, and there is still a lack of
understanding that teachers and student teachers are regarded as a teacher learning com­
munity. The collaboration with in-service teachers during student teacher training has not
been extensively studied and recognized for its potential benefits.
Drawing from the reviewed literature, we recognize the potential of both AI-assisted
learning, specifically through platforms like ChatGPT, and collaborative learning with in-
service teachers, in shaping student teachers’ experiences in the context of iSTEM education.
However, a gap remains in understanding how these two collaborative learning methods
compare in terms of enhancing key components of student teachers’ STEM teaching training,
such as critical thinking, learning performance, and cognitive load. This uncharted territory
necessitates empirical investigation. Consequently, this study will venture into this under­
explored domain and address the identified gap through the lens of the following research
questions. These questions aim to discern which method more effectively augments student
teachers’ proficiency in the context of collaborative STEM learning course. Accordingly, the
following three research questions were proposed:
Research Question 1 (RQ1): Which collaborative learning method can better promote the devel­
opment of critical thinking skills among student teachers within a iSTEM teaching training course?
Research Question 2 (RQ2): Which collaborative learning method is more effective in improving
the learning performance of student teachers in the iSTEM teaching training course?
Research Question 3 (RQ3): Which collaborative learning method leads to a lower cognitive load
for student teachers throughout their iSTEM teaching training?
ASIA PACIFIC JOURNAL OF EDUCATION 49

Methodology
Participants and procedure
This quasi-experimental study adopted in-service teacher-student teacher pairing learning (TSPL)
and ChatGPT-student teacher pairing learning (CSPL) for both experimental treatments. The parti­
cipants were 23 master’s students majoring in educational technology at a public normal university
in Guangdong Province, China (with a mean age of 24.4 years), and a teaching team composed of
one lead instructor and four collaborative teaching professors with extensive teaching and research
experience in STEM education from the same university. Two doctoral students also participated in
the course as teaching assistants. The study adopted a randomized grouping design. Before the
experiment, the teacher informed students of the two pairing methods and understood their
preferences. Regarding the potential issue that selection bias might impact the results, students
were then divided into two groups by a random lottery draw in the course: an TSPL group and a CSPL
group. Finally, there were one male and eleven female students in TSPL group (N = 12), and four male
and seven female students in CSPL group (N = 11). Besides, students from the TSPL group were
individually paired with twelve in-service teachers, including seven females and five males, all with
over three years of STEM teaching experience. These teachers, voluntarily participating from
Guangdong Province, China, are integral to the ongoing long-term STEM teaching research led by
the research team.
Before implementing the course curriculum, teaching assistant A spent an hour introducing
student teachers paired with ChatGPT to its usage. This encompassed how to use ChatGPT, the
tool’s limitations, and its applications in teaching. To ensure the integrity of the experiment, this
session was not made available to student teachers paired with in-service teachers. Meanwhile,
teaching assistant B guided the students paired with in-service teachers, discussing strategies for
effective collaboration, scheduled communication times, mutual expectations, and contact details.
The purpose of offering this iSTEM teaching training course is to equip student teachers with the
theories and methods of STEM education and to allow them to try their hands at designing iSTEM
lessons. Consequently, the students attended eight 3-h lessons focused on the fundamentals of
iSTEM education and the development of iSTEM teaching cases. As shown in Table 1, the learning
content covered five main themes, encompassing both theoretical knowledge and practical teach­
ing case design guidance. iSTEM teaching design and practice follow the “C-Pote” framework
proposed in previous research (Zhan et al., 2023), including conceptual group, problem chain,
objective layer, task cluster and evidence set (see Table 1 for detailed explanation). After class,
students were assigned to complete four tasks based on the course, and task records were meticu­
lously maintained in both the TSPL group and the CSPL group. These records provided invaluable
data, capturing the trajectory of students’ problem-solving journeys within their respective colla­
borative learning environments. Finally, except for the tasks, each student also needed to complete
a proposal design of iSTEM-based interdisciplinary thematic teaching and implementation. Besides,
ten-minute pre- and post-test questionnaires were administered to the students on the first and
last day of the course, respectively. After the course, students also submitted their own reflection
reports on course learning. This feedback helps to continuously optimize the course. During the
learning process, both groups of students were exposed to identical content and were required to
accomplish the same post-lesson tasks and the final teaching design proposal.

Instruments
To verify the effectiveness of both ChatGPT and in-service teacher assistance for student teachers’
collaborative learning in the iSTEM teaching training course, this study compared the effects of these
two collaborative learning methods on three variables: critical thinking, learning performance, and
cognitive load. The overall research design is illustrated in Figure 1.
50 T. LI ET AL.

Table 1. Course plan of the student teacher’s iSTEM teaching training.


Theoretical knowledge learning Practice training Weeks
1. Course introduction: overview of iSTEM-based Defining the task requirement: establishing a partnership 1st week
interdisciplinary thematic learning design and with in-service teachers or ChatGPT, and design iSTEM
implementation topics according to the “C-POTE” framework.
2. Orientation, levels, and innovative methods of Task 1: Choose a iSTEM topic and generate the big ideas (or 2nd week
iSTEM education clusters of concepts) for it
[Conceptual clusters are pivotal to achieving
interdisciplinary integration, serving as a thread that
weaves throughout the teaching process.]
3. Design and application of project-based learning Task 2: Build upon the generated big ideas for the iSTEM 3rd week
in iSTEM education teaching topic, formulate a series of driving questions
[The chain of problems, cantered around multiple
overarching concepts, effectively harnesses the guiding
role of questions in learning].
4. Evaluation of project-based learning in iSTEM Task 3: Clarify the learning objectives for the iSTEM theme 4th week
education based on the big ideas and driving questions [The object
layer serves as a tangible manifestation of core
competencies].
5. Information technology education and instruction Task 4: Design a series of learning tasks for the iSTEM theme 5th week
in the context of interdisciplinary integration [Task clusters, act as key tools to achieve students’
problem-solving skills and facilitate the transfer of
learning].
6. Review of course content of theoretical learning Task 5: Design an assessment plan for the learning of the 6th week
and Q&A (questions and answers) iSTEM theme [Evidence set is a method used to evaluate
students’ core competencies].
7. Project design of iSTEM-based interdisciplinary thematic teaching and implementation 7th week
8. iSTEM teaching project presentation 8th week

Figure 1. Research design of the study.

Critical thinking questionnaire


The critical thinking skills of the student teachers were assessed using a modified version of
C. C. Chen’s (2000) questionnaire, which is mainly based on the California Critical Thinking
Disposition Inventory (CCTDI) (Facione et al., 2000). The test includes five measurement dimensions:
truth-seeking (3 items, Cronbach’s α = .830), such as “When I realize my opinion is incorrect, I am
willing to change it”, open-mindedness (7 items, Cronbach’s α = .887), like “During discussions with
others, I strive to understand and listen to opinions different from my own”, analyticity (3 items,
Cronbach’s α = .632), like “When I encounter challenges, I try to find out the reasons and think of
solutions”, systematicity (8 items, Cronbach’s α = .836), such as “When others raise questions, I reflect
on matters related to their concerns”, and inquisitiveness (7 items, Cronbach’s α = .826), such as “I
find that I enjoy solving challenging problems”.
ASIA PACIFIC JOURNAL OF EDUCATION 51

Learning performance evaluation


The task performance evaluation was designed to examine the student teachers’ iSTEM
teaching knowledge and skills. It consisted of the four process learning tasks based on the
course framework, and an open-ended project design task also asked student teachers to
“Design a iSTEM teaching project through collaboration with your partner”. The assessment
of student teachers’ performance in both process-based and summative learning tasks is
conducted using the Consensual Assessment Technique (CAT) (Baer et al., 2004). This
approach relies on the student teachers’ submitted records of process-based tasks and
their final design proposals to evaluate their performance. Specifically, one of the teaching
assistants anonymized the assignments. Subsequently, the student’ assignments were eval­
uated by both the course’s lead professor and the two teaching assistants, and these
evaluators have no conflict of interest with the student teachers. This collaborative assess­
ment was intended to ensure robust, professional evaluations and consistent feedback.

Cognitive load questionnaire


Cognitive load was assessed using a single-item measure as an indicator of the mental effort
expended by the student teachers for learning this course, drawing on the work of Paas et al.
(2003). The student teachers were asked to rate their perceived cognitive load on a scale from 1 (very
low) to 9 (very high), like “On a scale of 1–9, how would you rate the effort you put into the learning
activities of this course? (1: Very easy, requiring minimal mental effort; 9: Very challenging, requiring
a significant amount of mental effort)”. This measure provides an indication of self-perceived
cognitive burden associated with the learning process, thus contributing a key element to our
comprehensive evaluation of the two collaboration methods on student teachers’ iSTEM teaching
training.

Content analysis of learning reflections and interviews


Additionally, to gain a deeper understanding of how student teachers learning through colla­
boration with ChatGPT or in-service teachers in the iSTEM training course, the research incorpo­
rated qualitative content analysis. Specifically, by examining their submitted course reflection
reports, 6 student teachers were selected from two groups (3 from each group) for interviews
(15–30 minutes each) to comprehensively explore the impact of different collaboration
approaches on them.

Results
Considering the small sample sizes, non-parametric tests were employed in the research to compare
the learning performance in critical thinking, STEM teaching learning performance and cognitive
load between the TSPL and CSPL groups. To answer the research questions, Wilcoxon Signed-Rank

Table 2. Effects of collaborative learning on student teachers’ critical thinking (pre-post test comparison).
CSPL a(N=11) TSPL a (N=12)

Variable Mean Rank Sum of Ranks Z p Mean Rank Sum of Ranks Z p


Truth seeking 2.00 2.00 −1.802b 0.072 4.40 22.00 −.060b 0.952
Analyticity 2.50 5.00 −1.852b 0.064 1.50 3.00 −1.225c 0.221
Systematicity 6.00 6.00 −1.990b 0.047 5.00 25.00 −.256b 0.798
Inquisitiveness 4.67 14.00 −1.012b 0.311 6.50 39.00 .000d 1.000
Open-mindedness 4.50 9.00 −1.604b 0.109 7.10 35.50 −.276b 0.783
Sub-total-critical thinking 6.00 6.00 −2.197b 0.028 4.88 39.00 .000d 1.000
a. Wilcoxon signed-rank test
b. Based on negative ranks
c. Based on positive ranks
d. Sum of negative ranks equals sum of positive ranks
52 T. LI ET AL.

Table 3. Differences of collaborative learning on student teachers’ STEM teaching training (post-test comparison).
CSPL (N=11) TSPL (N=12) CSPL Vs. TSPL

Variable Mean Rank Sum of Ranks Mean Rank Sum of Ranks Mann–Whitney U Z p
Truth seeking 12.14 133.50 11.88 142.50 64.500 −.094 0.925
Analyticity 12.64 139.00 11.42 137.00 59.000 −.443 0.658
Systematicity 15.23 167.50 9.04 108.50 30.500 −2.200 0.028
Inquisitiveness 12.32 135.50 11.71 140.50 62.500 −0.217 0.828
Open-mindedness 12.86 141.50 11.21 134.50 56.500 −0.589 0.556
Sub-total-critical thinking 13.86 152.50 10.29 123.50 45.500 −1.265 0.206
STEM proposal 9.14 100.50 14.63 175.50 34.500 −1.947 0.051
Task 1 15.00 165.00 9.25 111.00 33.000 −2.062 0.039
Task 2 16.91 186.00 7.50 90.00 12.000 −3.345 0.001
Task 3 15.55 171.00 8.75 105.00 27.000 −2.410 0.016
Task 4 15.05 165.50 9.21 110.50 32.500 −2.075 0.038
Task means 16.73 184.00 7.67 92.00 14.000 −3.202 0.001
Cognitive load 9.09 100.00 14.67 176.00 34.000 −2.046 0.041

Tests were performed to analyse differences between pre- and post-test scores (see Table 2). Besides,
Mann-Whitney U Tests were conducted to assess the differences between groups in post-test scores
(see Table 3). SPSS 25 was used to analyse the data.

Critical thinking
Regarding the critical thinking factor, the CSPL group exhibited improvements in all items. Notably,
systematicity (Z=-1.990, p = 0.047) and the overall critical thinking score (Z=-2.197, p = 0.028) achieved
statistical significance. On the other hand, the TSPL group also showed slight improvements in scores
but did not achieve statistical significance. In the post-test, aside from systematicity (Z=-2.200, p =
0.028), no notable differences in critical thinking were observed between the two groups.
In addition, in their reflective reports, student teacher from CSPL group noted that “When seeking
straightforward answers to simple questions, ChatGPT’s responses are more direct and clear-cut
compared to conventional search engines, reducing the costs of information filtering and practical
application. However, the accuracy remains to be validated”. Another feedback states, “When I asked
ChatGPT for literature recommendations, it provided titles, but no matter how I searched, I couldn’t
locate those articles. Hence, we should approach ChatGPT’s responses with a critical mindset”. These
feedbacks provided supplementary evidences on ChatGPT’s prowess in fostering critical thinking
among student teachers. Meanwhile, the TSPL group’s student teacher L also noted: “Teacher
A could not directly respond to my questions, and her answers were often beside the point.
However, her practical experience insights were highly beneficial to me”.

Learning performance
In terms of learning performance, the post-test scores for the CSPL group for iSTEM teaching
proposal design was lower than the TSPL group, but not reached statistical significance (Z=-1.947,
p = 0.051). While in task 1 (Z=-2.062, p = 0.039), task 2 (Z=-3.345, p = 0.001), task 3 (Z=-2.410, p =
0.016), and task 4 (Z=-2.075, p = 0.038), the CSPL group was found performed significantly better
than the teacher group in all four items. Similarly, in terms of the task average scores, the CSPL group
significantly outperformed the TSPL (Z=-3.202, p = 0.001).
Student teacher also mentioned, “when initially formulating an interdisciplinary concept, I hadn’t
considered the connection between AI-enabled palace lantern manufacturing and the concept of
‘sustainable development’. Upon querying ChatGPT, I learned that creating digitalized palace
lanterns using AI technology can reduce energy consumption, promoting societal sustainability”.
ASIA PACIFIC JOURNAL OF EDUCATION 53

Cognitive load
For the cognitive load factor, the post-test score of the CSPL group (Mean Rank = 9.09, Sum of Ranks
= 100.00) was lower than the TSPL group (Mean Rank = 14.67, Sum of Ranks = 176.00), indicating
a significant difference between the two groups (Z=-2.046, p = 0.041).
The student teachers from the CSPL group also mentioned, “After training or role-playing with
ChatGPT, the overall user experience is positive, and it can answer my questions at any time. During
the interactive communication, I noticed that many of its responses surpassed my initial precon­
ceived answers to the questions, which helped expand my learning horizons”. “Through error
correction and follow-up questions, ChatGPT’s application enhanced my understanding of the
core concepts of the chosen iSTEM theme, allowing me to further inquire about more abstract and
advanced content”. “When posing questions to ChatGPT and providing a thought framework, its
responses are structured and logical. Although there is still a lack in comprehensiveness and depth, it
can help fill in the gaps”. “Although the course was challenging, when completing tasks with certain
cognitive loads, there are still significant takeaways. By practically solving a specific teaching design
task, I can gain more practical experience in interdisciplinary theme learning design”.
While feedback from the TSPL group suggested “strengthening the guidance on iSTEM teaching
theories for in-service teachers. Before the course starts, these teachers should first be introduced to
the relevant theories, or student teachers should be organized to discuss them together”. Other
student teachers commented, “My own understanding can limit the effectiveness of pairing with
teachers. Without a deep understanding of the frontline situation, it’s challenging to communicate
well with paired teachers. Such communication can decrease my enthusiasm, making me feel that
I cannot complete my assignment with the paired teacher”.

Discussion
According to the results reported in section 4, both the CSPL group and the TSPL group demon­
strated the effectiveness of the collaborative learning method. And the ChatGPT and teacher
collaborations each exhibited strengths and weaknesses in the STEM teaching training course for
student teachers.

RQ1: Which collaborative learning method can better promote the development of critical think­
ing skills among student teachers within a iSTEM teaching training course?

As for RQ1, there is a significant difference in the pre-test and post-test scores among CSPL group,
demonstrating that the use of ChatGPT can enhance student teachers’ critical thinking skills. This
discovery is in line with previous studies, reinforcing the assertion that AI chatbots such as ChatGPT
can play a crucial role in fostering student teachers’ critical thinking (Rusandi et al., 2023). Due to the
significant challenges for the iSTEM teaching training course, and considering the uncertainty
brought about by AI technology, the teacher apprised student teachers of ChatGPT’s potential
inaccuracies and the possibility of misleading information (Fergus et al., 2023). As a result, when
student teachers faced cognitive conflicts provoked by ChatGPT’s responses (Ciechanowski et al.,
2019), they were spurred to critically reflect upon these answers and cross-verify their thoughts with
reputable journal articles, or to improve the prompts, reflecting certain aspects of critical thinking.
The traits of critical thinking primarily encompass questioning, reasoning, and reflection. It’s the
capability to make rational decisions about what to believe and what to do.
Similarly, TSPL group also witnessed a slight increase in critical thinking scores. The reasons might
be, the present iSTEM teaching training course, designed to equip student teachers with the
theoretical knowledge and “C-POTE” practice model of interdisciplinary learning in the context of
the Chinese new curriculum, introduces numerous fresh concepts and design paradigms. Curiosity
54 T. LI ET AL.

and desire for knowledge are important foundations for cultivating critical thinking, therefore, the
course content itself could foster the critical thinking. Questions posed by student teachers, such as,
“What are the big ideas and their hierarchical systems?” or “What are the pathways to extracting the
big ideas of the interdisciplinary themes?” – were questions that in-service teachers often struggled
to provide explicit answers to certain queries. Furthermore, because of the complexion and open­
ness of these questions, coupled with the limited training or seminars, in-service teachers had to
communicate with student teachers by relying solely on their practical iSTEM teaching experiences.
It leaves room for independent and critical thinking by student teachers. Hence, while completing
their assignments, student teachers need critically consider the answers provided by in-service
teachers.
The Mann-Whitney U test showed no statistical difference of the whole score of critical thinking
between the two pairing methods. One reason could be the small sample size in both groups, which
could affect the results. Another reason might be the high difficulty level of the course, requiring
student teachers to critically evaluate answers provided by either ChatGPT or in-service teachers
when confronted with new concepts and instruction philosophies. However, a statistical difference
was found in the post-test scores for the systematicity dimension. As ChatGPT can assist student
teachers in solving problems anytime and anywhere, they were able to maintain high concentration
on problem-solving. Meanwhile, due to their heavy workload, in-service teachers often struggled to
communicate timely and effectively to student teachers’ questions and difficulties. Therefore, CSPL
group scored higher on the “systemic” dimension compared to TSPL group.

RQ2: Which collaborative learning method is more effective in improving the learning perfor­
mance of student teachers in the iSTEM teaching training course?

In addressing RQ2, it was observed that student teachers who collaborated with ChatGPT generally
submitted those 4 task assignments of higher quality, thus achieving greater scores compared to those
paired with in-service teachers. Consequently, using ChatGPT can enhance student teachers’ task
performance, corroborating the existing findings (Biswas, 2023; Strzelecki, 2023). In this study, the
complexity of the four assigned tasks was purposefully structured to prevent copying, demanding
thoughtful analysis and reflection for completion. Additionally, interaction logs with ChatGPT were
mandated for submission by student teachers. ChatGPT’s rapid response to queries provided diverse
insights on STEM topics, aiding student teachers in their analytical and creative thinking, and identify­
ing assignment issues (Zhai, 2023), thereby producing assignments of superior efficiency and quality.
The qualitative content further enriches the interpretation of this outcome. As the student teacher
mentioned, “Throughout the course, I gained an initial understanding of how ChatGPT can empower
teachers in preparing for STEM-themed lessons”. Specifically, first, with ChatGPT serving as an
advanced knowledge management tool, it can refine and apply insights from a wealth of materials,
thereby reducing the burden of lesson planning for teachers. Second, ChatGPT acts as an “all-subject
teacher”, assisting in interdisciplinary knowledge integration during collaborative discussions and
facilitating co-planning for specialized teachers. Third, GPT contributes to fostering interdisciplinary
teaching innovation by assisting teachers in providing prompt clarifications and facilitating diver­
gent thinking. When dealing with segmented process-oriented learning tasks, student teachers
showed a generally positive response to the overall assistance provided by ChatGPT.
In the broader context of STEM teaching proposal design, the integration of these four assign­
ments was pivotal. Sharing of exemplary assignments among student teachers encouraged inter-
group learning and emulation. Conversely, student teachers who partnered with in-service teachers
benefitted from practical insights, resulting in STEM teaching proposal designs that were more
comprehensive, grounded, and applicable, with stronger consideration of students’ actual needs and
conditions. Consequently, the proposal design outcomes of student teachers collaborating with in-
service teachers proved more effective. This result further reaffirms the importance of collaborative
ASIA PACIFIC JOURNAL OF EDUCATION 55

learning with in-service teachers in the STEM teaching training course. Given the inherent complexity
of STEM teaching, student teachers need to immerse themselves in schools to experience authentic
teaching situations. They should understand how teachers organize students for inquiry-based
activities and acquire practical STEM teaching knowledge, methods, and experiences (Vangrieken
et al., 2015). Thus, during this course, six student teachers proactively reached out to in-service
teachers to observe STEM classroom teaching in primary and secondary schools. They discussed
face-to-face the cognition, design, implementation, and reflection related to STEM teaching, enhan­
cing their overall understanding of integrated STEM education. Some student teachers commented
after their firsthand experience, “I never realized that actual STEM teaching was so fascinating, yet so
challenging to execute”.
In short, for short-term, well-defined, knowledge-based and less innovative STEM collaborative
teaching design tasks, pairing with ChatGPT led to better outcomes compared to pairing with in-
service teachers. However, for complex, highly innovative, open-ended and systematic collaborative
teaching design tasks that require sustained in-depth interaction, pairing with in-service teachers
was more beneficial. Therefore, a synergistic combination of these collaborative strategies, lever­
aging strengths and compensating for weaknesses, promises to optimize the learning performance
of student teachers.

RQ3: Which collaborative learning method leads to a lower cognitive load for student teachers
throughout their iSTEM teaching training?

For RQ3, a significant difference was observed in the post-test scores of cognitive load between
student teachers paired with ChatGPT and those paired with in-service teachers. This implies that the
aid of ChatGPT in completing tasks and resolving ill-structured problems can reduce cognitive load,
consistent with the existing findings (Ji et al., 2023; Ye et al., 2023). Because ChatGPT can interpret
subtle linguistic variations and react appropriately (Zhai et al., 2023), it can mitigate the mental effort
expended by student teachers in processing and understanding iSTEM-related knowledge and
concepts. Despite not always providing completely accurate answers, however, its expeditious
responses and its capacity to promote both divergent and convergent thinking (Mukherjee &
Chang, 2023) can still lessen the cognitive effort for student teachers.
In contrast, student teachers paired with in-service teachers often struggle to get accurate
explanations for complex and abstract problems and concepts in their assignments. They frequently
need to invest additional time and energy in literature review and external assistance, thereby
maintaining a high cognitive load. Indeed, the primary goal of this iSTEM teaching training course
is to promote active learning and inquiry through complex and authentic problems or concepts
(Dwivedi et al., 2023; Patel & Lam, 2023), which might impose additional cognitive demands on
student teachers, hence the high cognitive load in both groups (Gupta & Zheng, 2020).
While ChatGPT exhibits proficiency in handling common sense problems (Ray, 2023), it falls short
in processing experiential and sociocultural knowledge, essentially functioning as an inexperienced
problem solver in these domains. Hence, in-service teachers, with their wealth of experience, can
provide such knowledge to student teachers, potentially reducing the mental effort involved in
understanding such knowledge. However, given the personal, impartible, and experiential nature of
such knowledge, it is challenging to quantify, offering fertile ground for further in-depth exploration.

Conclusion
To cultivate student teacher with excellent iSTEM teacher characteristics, the deployment of AI
tools and the collaboration with in-service teachers stand out as promising strategies in con­
temporary student teacher education. Despite calls from existing research to strengthen colla­
borative learning, there remains a dearth of in-depth empirical research in this area. Our study
56 T. LI ET AL.

attempts to bridge this gap by exploring the efficacy of different collaborative learning methods –
AI-assisted learning with ChatGPT and collaboration with in-service teachers – in enhancing
student teachers’ iSTEM teaching training course. Both collaborative learning methods show­
cased their distinct strengths. On one hand, ChatGPT’s user-friendliness, high interactivity, and
robust database support proved beneficial in helping student teachers learn complex iSTEM
teaching knowledge, hone critical thinking skills, improve learning efficiency, and reduce cogni­
tive load. On the other hand, the involvement of in-service teachers facilitated a two-way
exchange of theoretical knowledge and practical wisdom, offering a more realistic understanding
of iSTEM teaching scenarios. This interaction helped student teachers develop high-quality iSTEM
teaching proposals and exhibit improved iSTEM teaching ability.
We propose that a blend of both collaborative learning methods would cater to the diverse and
personalized iSTEM teaching needs of student teachers. For instance, while ChatGPT could be
efficiently employed for acquiring basic knowledge and skills, the invaluable practical wisdom of in-
service teachers could be leveraged for more complex and experiential learning. To facilitate these
collaborative learning methods, educators and policymakers must provide further support, such as
structuring progressive learning tasks, managing learning difficulty, and fostering long-term colla­
boration between universities and primary and secondary schools. Moreover, care must be taken to
provide clear guidance on the use of tools like ChatGPT to avoid potential ethical risks. Overall, both
collaborative approaches have their advantages in cultivating student teachers. Future studies can
further explore ways to integrate and apply them.
However, our study’s results might be impacted by limited participant numbers, restricted to
technical major graduate students, and the short experimental period. In this exploratorystudy, even
though we have attempted an empirical comparative study, some of the effects are borderline in
terms of significance, and the effect sizes are also small on some indicators. Besides, more formal and
comprehensive assessment of ChatGPT on its ability to help with iSTEM education-related topics are
needed, especially when the teaching content is related to localized iSTEM teaching training course.
Therefore, future research involving longer experimental periods, more adaptable learning scaffold
design, and a more diverse and larger participant group is required to train high-quality iSTEM
teachers for an AI-driven future.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This research was financially supported by the National Natural Science Foundation in China (62277018; 62237001), the
Ministry of Education in China Project of Humanities and Social Sciences (22YJC880106), the Major Project of Social
Science in South China Normal University (ZDPY2208), the Degree and graduate education Reform research project in
Guangdong (2023JGXM046).

Notes on contributors
Zehui Zhan is a PhD, professor, doctoral supervisor in South China Normal University, PI of the Smart Educational
Equipment Industry-University-Research Cooperation Base. Her research interest includes learning science, STEAM
education, smart education, entrepreneurial education.
Tingting Li is a PhD candidate in the School of Information Technology in Education at the South China Normal
University. Her major research interests lie in the area of STEM education, information technology education in primary
and secondary schools, artificial intelligence education.
Yu Ji is a PhD candidate in the School of Information Technology in Education at the South China Normal University. His
current research focuses on STEM education, information technology education in primary and secondary schools,
artificial intelligence education.
ASIA PACIFIC JOURNAL OF EDUCATION 57

ORCID
Tingting Li http://orcid.org/0000-0003-1095-7597
Yu Ji http://orcid.org/0009-0007-1249-3366
Zehui Zhan http://orcid.org/0000-0002-6936-1977

Data availability statement


Data available on request from the corresponding author Zehui Zhan at zhanzehui@m.scnu.edu.cn.

References
Baer, J., Kaufman, J.C., & Gentile, C.A. (2004). Extension of the consensual assessment technique to nonparallel creative
products. Creativity Research Journal, 16(1), 113–117. https://doi.org/10.1207/s15326934crj1601_11
Biswas, S. (2023). Role of chat GPT in education. Available at SSRN 4369981. https://ssrn.com/abstract=4369981
Brown, K.T. (2017). Pre-service teachers’ acquisition of content knowledge, pedagogical skills, and professional disposi­
tions through service learning. Science Education & Civic Engagement, 9(2), 13–26.
Carrillo, C., & Flores, M.A. (2020). COVID-19 and teacher education: A literature review of online teaching and learning
practices. European Journal of Teacher Education, 43(4), 466–487. https://doi.org/10.1080/02619768.2020.1821184
Chai, C.S. (2019). Teacher professional development for Science, Technology, Engineering and Mathematics (STEM)
education: A review from the perspectives of Technological Pedagogical Content (TPACK). The Asia-Pacific Education
Researcher, 28(1), 5–13. https://doi.org/10.1007/s40299-018-0400-7
Chen, C.C. (2000). An experimental study about implementing the teaching strategies of critical thinking in social class of
5th grade elementary school [Unpublished master thesis]. National Chengchi University. (Master’s thesis, National
Chengchi University). http://thesis.lib.nccu.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dallcdr&s=id=%22A2002001685%22.
&searchmode=basic
Chen, X. (2020). Pre-service teachers’ self-efficacy of interdisciplinary team teaching through the use of collaborative
concept map. International Journal of Technology in Teaching and Learning, 15(2), 76–94. https://doi.org/10.37120/
ijttl.2019.15.2.01
Chen, X., Xie, H., Zou, D., & Hwang, G.J. (2020). Application and theory gaps during the rise of artificial intelligence in
education. Computers and Education: Artificial Intelligence, 1, 100002. https://doi.org/10.1016/j.caeai.2020.100002
Chen, X., Zou, D., Xie, H., Cheng, G., & Liu, C. (2022). Two decades of artificial intelligence in education: Contributors,
collaborations, research topics, challenges, and future directions. Educational Technology & Society, 25(1), 28–47.
https://www.jstor.org/stable/48647028
Ciechanowski, L., Przegalinska, A., Magnuski, M., & Gloor, P. (2019). In the shades of the uncanny valley: An experimental
study of human–chatbot interaction. Future Generation Computer Systems, 92, 539–548. https://doi.org/10.1016/j.
future.2018.01.055
Cooper, G. (2023). Examining science education in ChatGPT: An exploratory study of generative artificial intelligence.
Journal of Science Education and Technology, 32(3), 444–452. https://doi.org/10.1007/s10956-023-10039-y
Dare, E.A., Ellis, J.A., & Roehrig, G.H. (2018). Understanding science teachers’ implementations of integrated STEM
curricular units through a phenomenological multiple case study. International Journal of STEM Education, 5(1), 4.
https://doi.org/10.1186/s40594-018-0101-z
Dwivedi, Y.K., Kshetri, N., Hughes, L., Slade, E.L., Jeyaraj, A., Kar, A.K. et al (2023). Opinion paper: “so what if ChatGPT
wrote it?” multidisciplinary perspectives on opportunities, challenges and implications of generative conversational
AI for research, practice and policy. International Journal of Information Management, 71, 102642. https://doi.org/10.
1016/j.ijinfomgt.2023.102642
English, L.D. (2016). STEM education K-12: Perspectives on integration. International Journal of STEM Education, 3(1), 1–8.
https://doi.org/10.1186/s40594-016-0036-1
Estapa, A.T., & Tank, K.M. (2017). Supporting integrated STEM in the elementary classroom: A professional development
approach centered on an engineering design challenge. International Journal of STEM Education, 4. https://doi.org/
10.1186/s40594-017-0058-3
Facione, P.A., Facione, N.C., & Giancarlo, C.A.F. (2000). The California critical thinking disposition inventory: CCTDI test
manual. California Acad. Press.
Farrokhnia, M., Banihashem, S.K., Noroozi, O., & Wals, A. (2023). A SWOT analysis of ChatGPT: Implications for educational
practice and research. Innovations in Education and Teaching International, 1–15. https://doi.org/10.1080/14703297.
2023.2195846
Fergus, S., Botha, M., & Ostovar, M. (2023). Evaluating academic answers generated using ChatGPT. Journal of Chemical
Education, 100(4), 1672–1675. https://doi.org/10.1021/acs.jchemed.3c00087
58 T. LI ET AL.

Geeraerts, K., Tynjälä, P., & Heikkinen, H.L.T. (2018). Inter-generational learning of teachers: What and how do teachers
learn from older and younger colleagues? European Journal of Teacher Education, 41(4), 479–495. https://doi.org/10.
1080/02619768.2018.1448781
Gupta, U., & Zheng, R.Z. (2020). Cognitive load in solving mathematics problems: Validating the role of motivation and
the interaction among prior knowledge, worked examples, and task difficulty. European Journal of STEM Education, 5
(1), 5. https://doi.org/10.20897/ejsteme/9252
Guzman, A.L., & Lewis, S.C. (2020). Artificial intelligence and communication: A human–machine communication
research agenda. New Media & Society, 22(1), 70–86. https://doi.org/10.1177/1461444819858691
Haas, B., Kreis, Y., & Lavicza, Z. (2021). Integrated STEAM approach in outdoor trails with elementary school pre-service
teachers. Educational Technology & Society, 24(4), 205–219. https://doi.org/10.30191/ETS.202110_24(4).0016
Harlow, A., & Cobb, D.J. (2014). Planting the seed of teacher identity: Nurturing early growth through a collaborative
learning community. Australian Journal of Teacher Education (Online), 39(7), 70–88. https://doi.org/10.14221/ajte.
2014v39n7.8
Honey, M., Pearson, G., & Schweingruber, H.A. (Eds.). (2014). STEM integration in K-12 education: Status, prospects, and an
agenda for research. National Academies Press.
Huang, B., Jong, M.S.Y., Tu, Y.F., Hwang, G.J., Chai, C.S., & Jiang, M.Y.C. (2022). Trends and exemplary practices of STEM
teacher professional development programs in K-12 contexts: A systematic review of empirical studies. Computers &
Education, 104577, 104577. https://doi.org/10.1016/j.compedu.2022.104577
Huang, K., Lubin, I.A., & Ge, X. (2011). Situated learning in an educational technology course for pre-service teachers.
Teaching and Teacher Education, 27(8), 1200–1212. https://doi.org/10.1016/j.tate.2011.06.006
Ji, Y., Zou, X., Li, T., & Zhan, Z. (2023). The effectiveness of ChatGPT on pre-service teachers’ STEM teaching literacy,
learning performance, and cognitive load in a teacher training course. In 2023 7th International Conference on
Education and Multimedia Technology (ICETM). ACM, New York, NY, USA. https://doi.org/10.1145/3637907.3637948
Kelley, T.R., & Knowles, J.G. (2016). A conceptual framework for integrated STEM education. International Journal of STEM
Education, 3(1), 1–11. https://doi.org/10.1186/s40594-016-0046-z
Kelley, T.R., Knowles, J.G., Holland, J.D., & Han, J. (2020). Increasing high school teachers’ self-efficacy for integrated STEM
instruction through a collaborative community of practice. International Journal of STEM Education, 7(1), 1–13.
https://doi.org/10.1186/s40594-020-00211-w
Kim, B., & Bastani, R. (2017). Students as game designers: Transdisciplinary approach to STEAM education. Alberta
Science Education Journal, 45(1), 45–52.
Kim, J., Lee, H., & Cho, Y.H. (2022). Learning design to support student-AI collaboration: Perspectives of leading teachers for AI
in education. Education and Information Technologies, 27(5), 6069–6104. https://doi.org/10.1007/s10639-021-10831-6
King, M.R., & ChatGPT. (2023). A conversation on artificial intelligence, chatbots, and plagiarism in higher education.
Cellular and Molecular Bioengineering, 16(1), 1–2. https://doi.org/10.1007/s12195-022-00754-8
Kostiainen, E., Ukskoski, T., Ruohotie-Lyhty, M., Kauppinen, M., Kainulainen, J., & Mäkinen, T. (2018). Meaningful learning
in teacher education. Teaching and Teacher Education, 71, 66–77. https://doi.org/10.1016/j.tate.2017.12.009
Krügel, S., Ostermaier, A., & Uhl, M. (2023). ChatGPT’s inconsistent moral advice influences users’ judgment. Scientific
Reports, 13(1), 4569. https://doi.org/10.1038/s41598-023-31341-0
Lau, W.W., & Jong, M.S. (2022). Typology of teachers’ stages of concern for STEM education. Research in Science &
Technological Education, 1–19. https://doi.org/10.1080/02635143.2022.2064447
Li, T., Zhang, X., Lai, R., Zhang, M., Zhong, C., & Zhan, Z. (2023). Internet+ technology empowering C-STEAM educational
collaboration in the greater Bay Area: A cross-regional project of “Chinese ancient hollowed windows”. In Proceedings
of the 2023 5th International Workshop on Artificial Intelligence and Education (WAIE 2023). Tokyo, Japan.
Lo, C.K. (2021). Design principles for effective teacher professional development in integrated STEM education.
Educational Technology & Society, 24(4), 136–152. https://doi.org/10.30191/ETS.202110_24(4).0011
Lo, C.K. (2023). What is the impact of ChatGPT on education? A rapid review of the literature. Education Sciences, 13(4),
410. https://doi.org/10.3390/educsci13040410
Meschede, N., Fiebranz, A., Möller, K., & Steffensky, M. (2017). Teachers’ professional vision, pedagogical content
knowledge and beliefs: On its relation and differences between pre-service and in-service teachers. Teaching &
Teacher Education, 66, 158–170. https://doi.org/10.1016/j.tate.2017.04.010
Morris, J., Song, Y., Soloway, E., & Norris, C. (2021). Teacher professional development in STEM education. Journal of
Educational Technology & Society, 24(4). https://doi.org/10.30191/ETS.202110_24(4).0006
Mukherjee, A., & Chang, H. (2023). Managing the creative frontier of generative AI: The Novelty-Usefulness Tradeoff.
California Management Review. https://ssrn.com/abstract=4469983
Näykki, P., Kontturi, H., Seppänen, V., Impiö, N., & Järvelä, S. (2021). Teachers as learners–a qualitative exploration of pre-
service and in-service teachers’ continuous learning community OpenDigi. Journal of Education for Teaching, 47(4),
495–512. https://doi.org/10.1080/02607476.2021.1904777
ASIA PACIFIC JOURNAL OF EDUCATION 59

Organisciak, P., Acar, S., Dumas, D., & Berthiaume, K. (2023). Beyond semantic distance: Automated scoring of divergent
thinking greatly improves with large language models. Thinking Skills and Creativity, 49, 101356. https://doi.org/10.
1016/j.tsc.2023.101356
Paas, F., Tuovinen, J.E., Tabbers, H., & Van Gerven, P.W. (2003). Cognitive load measurement as a means to advance
cognitive load theory. Educational Psychologist, 38(1), 63–71. https://doi.org/10.1207/S15326985EP3801_8
Patel, S.B., & Lam, K. (2023). ChatGPT: The future of discharge summaries? Lancet Digital Health, 5(3), e107–e108. https://
doi.org/10.1016/S2589-7500(23)00021-3
Radloff, J., & Guzey, S. (2016). Investigating preservice STEM teacher conceptions of STEM education. Journal of Science
Education and Technology, 25(5), 759–774. https://doi.org/10.1007/s10956-016-9633-5
Rani, P.S., Rani, K.R., Daram, S.B., & Angadi, R.V. (2023). Is it feasible to reduce academic stress in net-zero energy buildings?
Reaction from ChatGPT. Annals of Biomedical Engineering, 51(12), 1–3. https://doi.org/10.1007/s10439-023-03286-y
Ray, P. P.(2023). ChatGPT: A comprehensive review on background, applications, key challenges, bias, ethics, limitations and
future scope. Internet of Things and Cyber-Physical Systems, 3,121–154. https://doi.org/10.1016/j.iotcps.2023.04.003
Rudolph, J., Tan, S., & Tan, S. (2023). ChatGPT: Bullshit spewer or the end of traditional assessments in higher education?
Journal of Applied Learning & Teaching, 6(1). https://doi.org/10.37074/jalt.2023.6.1.9
Rusandi, M.A., Ahman Saripah, I., Khairun, D.Y., Khairun, D.Y., & Mutmainnah. (2023). No worries with ChatGPT: Building
bridges between artificial intelligence and education with critical thinking soft skills. Journal of Public Health, 45(3),
e602–e603. https://doi.org/10.1093/pubmed/fdad049
Sá, M.V., & de Almeida, M.J.B.M. (2016). Physics teachers: A holistic plan for professional education during both the
pedagogical stage and the probation year. Teachers & Teaching, 22(4), 504–518. https://doi.org/10.1080/13540602.
2015.1082733
Schmid, E.C., & Hegelheimer, V. (2014). Collaborative research projects in the technology-enhanced language classroom:
Pre-service and in-service teachers exchange knowledge about technology. ReCALL, 26(3), 315–332. https://doi.org/
10.1017/S0958344014000135
Shernoff, D.J., Sinha, S., Bressler, D.M., & Ginsburg, L. (2017). Assessing teacher education and professional development
needs for the implementation of integrated approaches to STEM education. International Journal of STEM Education,
4(1), 13. https://doi.org/10.1186/s40594-017-0068-1
Song, H., & Zhou, M. (2021). STEM teachers’ preparation, teaching beliefs, and perceived teaching competence:
A multigroup structural equation approach. Journal of Science Education and Technology, 30(2), 394–407. https://
doi.org/10.1007/s10956-020-09881-1
Stokel-Walker, C. (2022). AI bot ChatGPT writes smart essays-should academics worry? Nature. https://doi.org/10.1038/
d41586-022-04397-7
Strzelecki, A. (2023). To use or not to use ChatGPT in higher education? A study of students’ acceptance and use of
technology. Interactive Learning Environments, 1–14. https://doi.org/10.1080/10494820.2023.2209881
Taber, K.S. (2019). Experimental research into teaching innovations: Responding to methodological and ethical
challenges. Studies in Science Education, 55(1), 69–119. https://doi.org/10.1080/03057267.2019.1658058
Tan, E. (2022). ‘Heartware’for the compassionate teacher: Humanizing the academy through mindsight, attentive love,
and storytelling. Journal of Applied Learning & Teaching, 5(2), 152–159. https://doi.org/10.37074/jalt.2022.5.2.ss1
Tiatri, S., Veronica, C., Fiscarina, C., Nurkholiza, R., Wakano, V.Y., Ie, M., & Beng, J.T. (2023). Elementary school teachers’
perceptions of critical thinking in stem learning. International Journal of Application on Social Science and Humanities,
1(1), 648–658. https://doi.org/10.24912/ijassh.v1i1.25924
Timothy, V., Fischer, F., Watzka, B., Girwidz, R., & Stadler, M. (2023). Applying cognitive load theory in teacher education:
An experimental validation of the scale by leppink et al. Psychological Test Adaptation and Development, 4(1), 246.
https://doi.org/10.1027/2698-1866/a000052
Vangrieken, K., Dochy, F., Raes, E., & Kyndt, E. (2015). Teacher collaboration: A systematic review. Educational Research
Review, 15, 17–40. https://doi.org/10.1016/j.edurev.2015.04.002
Willegems, V., Consuegra, E., Struyven, K., & Engels, N. (2017). Teachers and pre-service teachers as partners in
collaborative teacher research: A systematic literature review. Teaching and Teacher Education, 64, 230–245.
https://doi.org/10.1016/j.tate.2017.02.014
Wu, C.H., Liu, C.H., & Huang, Y.M. (2022). The exploration of continuous learning intention in STEAM education through
attitude, motivation, and cognitive load. International Journal of STEM Education, 9(1), 35. https://doi.org/10.1186/
s40594-022-00346-y
Xu, W., & Ouyang, F. (2022). The application of AI technologies in STEM education: A systematic review from 2011 to
2021. International Journal of STEM Education, 9(1), 1–20. https://doi.org/10.1186/s40594-022-00377-5
Ye, Y., You, H., & Du, J. (2023). Improved trust in human-robot collaboration with ChatGPT. Institute of Electrical and
Electronics Engineers Access, 11, 55748–55754. https://doi.org/10.1109/ACCESS.2023.3282111
Zhai, X. (2023). ChatGPT for next generation science learning. XRDS: Crossroads, the ACM Magazine for Students, 29(3),
42–46. https://doi.org/10.1145/3589649
Zhai, X., Nyaaba, M., & Ma, W. (2023). Can AI Outperform Humans on cognitive-demanding tasks in science? Available at
SSRN (May 17, 2023). https://doi.org/10.2139/ssrn.4451722
60 T. LI ET AL.

Zhang, K., & Aslan, A.B. (2021). AI technologies for education: Recent research & future directions. Computers and
Education: Artificial Intelligence, 2, 100025. https://doi.org/10.1016/j.caeai.2021.100025
Zhan, Z., Ji, Y., & Lai, Y. (2023). How to implement interdisciplinary thematic learning based on the compulsory
education curriculum standards: general guideline and operational model. Modern Distance Education Research, 35
(1), 49–58. http://kns.cnki.net/kcms/detail/51.1580.G4.20230111.0950.007.html
Zhan, Z., & Niu, S. (2023). Subject integration and theme evolution of STEM education in K-12 and higher education
research. Humanities and Social Sciences Communications, 10(1), 1–13. https://doi.org/10.1057/s41599-023-02303-8
Zhan, Z., Shen, W., Xu, Z., Niu, S., & You, G. (2022a). A bibliometric analysis of the global landscape on STEM education
(2004-2021): Towards global distribution, subject integration, and research trends. Asia Pacific Journal of Innovation
and Entrepreneurship, 16(2), 171–203. https://doi.org/10.1108/APJIE-08-2022-0090
Zhan, Z., Yao, X., & Li, T. (2022b). Effects of association interventions on students’ creative thinking, aptitude, empathy,
and design scheme in a STEAM course: Considering remote and close association. International Journal of Technology
and Design Education, 33(5), 1–23. https://doi.org/10.1007/s10798-022-09801-X

You might also like