Education and Information Technologies

https://doi.org/10.1007/s10639-024-13084-1

Facilitating the development of preservice teachers’

geometric thinking through artificial intelligence (AI)
assisted augmented reality (AR) activities: The case
of platonic solids

Tuğba Uygun1 · Ali Şendur1 · Beyza Top2 · Kadriye Coşgun‑Başeğmez3

Received: 7 June 2024 / Accepted: 27 September 2024

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature
2024

Abstract
Although augmented reality has become one of the most commonly used materi-
als that support learning, especially in learning geometric concepts, it is avoided to
be used in the lessons due to its complex structure. At that point, artificial intel-
ligence working as a personal assistant in many fields can help us learn to produce
our own model with augmented reality by coping with its complexity. In this study,
it is aimed to examine both the effectiveness of Artificial Intelligence (AI) assisted
Augmented Reality (AR) activities on the development of the preservice teachers’
(PSTs) geometric reasoning and their practicality in use. To this end, an interactive
learning environment with AI assisted AR activities including Chat Generative Pre-
Trained Transformer (ChatGPT), Blender and MyWebAR, and lasting five weeks
have been designed. The participants in this case study were 15 preservice math-
ematics teachers selected by maximum variation and criterion sampling techniques.
The data were collected through AI assisted AR activities, two semi-structured
interview forms, focus group interviews, anecdotes, activity forms and test. The
analysis results have been reported under three groups about the PSTs’ test results,
experiences on AI assisted AR activities and thoughts about the activities. To con-
clude, it has been observed that the PSTs’ conceptualization and geometric thinking
levels have been improved. Moreover, their thoughts have been illustrated with the
themes of “cognitive dimension”, “affective dimension”, “technological dimension”
and “employability dimension”.

Keywords Artificial intelligence · Augmented reality · Geometric thinking ·

Platonic solids · Preservice mathematics teachers

Extended author information available on the last page of the article

Vol.:(0123456789)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

1 Introduction

In addition to make sense of the bridge between mathematics and real life, learn-
ing geometry supports various high level thinking skills such as logical thinking
abilities, spatial understanding about 3-dimensional world and conceptualizing
mathematical arguments (Suydam, 1985) related to geometric thinking. A human
being who has been in touch with geometry since the moment he was born with
his efforts to discover 3-dimensional world, tries to learn geometry by working
on 2-dimensional plane in geometry lessons and develop geometric thinking. The
basis of geometrical thinking is the skills of visualization and drawing of geomet-
ric concepts and the examination of the spatial situations of the world in which we
live in integrity by creating generalizations based on them (Köse, 2008). Geom-
etry courses taught in our schools are mainly about teaching plane geometry. In
this process, teaching 3-dimensional geometric objects on 2-dimensional plane
geometry may cause human beings to experience difficulties in learning geomet-
ric concepts and developing geometric thinking (Alcañiz et al., 2010; Ma et al.,
2009). Also, the human beings may have difficulties related to cognitive filter by
imagining and identifying orthographic and perspective projections as kinds of
representation types of 3-dimensional objects (Ma et al., 2009: Pittalis & Chris-
tou, 2010). In other words, it is very difficult to depict space on a plane. This
endeavor, which forms the basis of geometrical thinking, also has an important
place in teaching 3-dimensional geometry. However, the drawings of a 3-dimen-
sional object on a 2-dimensional plain paper are either incomplete, causing optical
illusions and different perceptions, or even if they are perfect drawings, the static
nature of the environment does not allow the views of the shapes from different
facades to be seen in a single drawing. Studies have shown that in space geom-
etry courses taught based on plane geometry, it becomes difficult for students to
see the relationships between geometric objects and that they reach wrong conclu-
sions with different perceptions. In addition to this limitation related to geometri-
cal thinking, inaccurate or insufficient teaching methods, techniques and materials
cause student unsuccess and obstructing the development of geometry thinking.
When traditional teaching methods and techniques are preferred by the teachers
without considering their students’ learning and abilities in geometry lessons, this
may result in the students’ unsuccess by encouraging their learning difficulties
(Ozdemir et al., 2020).
The students’ difficulty and unsuccess on geometrical thinking may result from
the difference between reality and perception of reality because of the close con-
nection of geometry with the real life. By considering the fact that reality can
never be changed, but how it is perceived can be improved, the lessons can be
supported by applications near to real life. In this respect, integration of tech-
nology into geometry lessons has become inevitable (Dikovic, 2009). At that
point, Augmented Reality (AR) tools can propose some facilities as an interface
between reality and how it is perceived (Sahu et al., 2021). By using AR tools
as a kind of instructional technology in geometry education, effective learning
environments in which the human beings can get experiences close to real life

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

and perceive what happens in real life as realistically as possible can be designed.
Hence, AR tools can be beneficial to overcome these difficulties in cognitive filter
with the help of the connection among the real life and virtual elements.

1.1 Integration of AR into education

AR can be described as a kind of virtual reality and human-computer interaction

technology placing 3-dimensional virtual objects into real life (Azuma et al., 2001)
so that it becomes possible to experience the reality by superimposing virtual objects
in real time (Kaufmann, 2004). AR integrates digital elements into the real world,
allowing users to see both physical and virtual elements through devices like smart-
phones, tablets, or head-mounted displays (Alcañiz et al., 2010). This technology
serves as a connection between the real and virtual worlds, offering a blended expe-
rience where users can interact with digital information overlaid onto their physical
surroundings. This concept of AR bridging the gap between reality and virtuality has
been well-documented in research, as highlighted by Milgram et al. (1995). The key
advantage of AR technology lies in its ability to seamlessly integrate virtual objects
into real-world environments, enabling both students and educators to interact with
these objects as if they were physically present. This immersive experience allows
for various forms of interaction, such as touching, moving around, and manipulating
virtual elements. This capability does not only enhance the learning experience by
making it more engaging but also facilitates a deeper understanding of complex con-
cepts. AR tools in education offers interactive and personalized learning opportuni-
ties, allowing students to explore subjects in a more tangible and accessible manner.
Moreover, it is a beneficial teaching and learning medium by reducing mental effort
when compared with paper–pencil instruction or computer-assisted instruction in
common ways (Haniff & Baber, 2003; Wang & Dunston, 2006).
Integrating AR tools in education offers a key advantage that students from dif-
ferent grade levels can perceive virtual objects as if they were part of the real envi-
ronment and interact with them collaboratively using natural gestures and actions
(Milgram et al., 1995). This direct engagement with virtual elements mirrors real-
world scenarios, enabling learners to effectively comprehend spatial problems and
relationships between objects. The utilization of AR technology within educational
settings is an emerging domain, as highlighted by Mitchell (2011). Many researchers
have focused on the effects of AR in their studies. For example, Kaufmann (2004)
suggested an immersive approach facilitating a deeper understanding of spatial con-
cepts and object relations, enhancing the learning experience. Moreover, researchers
(Cerqueira & Kirner, 2012; Coimbra et al., 2015; Del Cerro Velázquez & Morales
Méndez, 2021; Kaufmann, 2004; Lindgren & Moshell, 2011) have directed their
attention towards implementing AR technology in the mathematics classrooms.
Their focus lies in enhancing students’ academic achievements, understanding of
mathematical contents, spatial intelligence and fostering greater motivation towards
learning mathematics.
When the studies on AR in the recent years in the literature are examined, it is
seen that in most of these studies, materials and activities prepared by researchers or

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

educators have been applied to the students from various age groups rather than using
models formed by students. This case results from necessitated abilities and knowledge
in order to use AR systems. To create 3-dimensional models by AR tools, students or
educators have to learn related programming languages and experience to acquire nec-
essary knowledge and skills to produce accurate codes. In order to equip with these
knowledge, skills and experience, they have to study hard and make a lot of exercises
even to form a simple shape. Hence, in order to design a study in which the students
produce activities or materials by using AR tools, researchers must plan a long and rig-
orous process to teach the students the necessary programming language and experi-
ence it. For example, in order to make AR application with Blender, the researchers
need to teach Python programming language to the students and provide them opportu-
nities to experience it well. At that point, AI can enhance and provide the opportunities
for the use of AR systems by supporting, directing and guiding users without neces-
sitating information and for the experience of using AR systems at the beginning steps.

1.2 Supporting and assisting AR systems by AI in education

The question of whether machines could develop thinking abilities akin to humans
(Turing, 1950) sparked the inception of artificial intelligence. With this motiva-
tion, ChatGPT was developed by OpenAI as an artificial intelligence-supported
chatbot. It has critical importance and power of replacement for certain professions
that demand creativity and productivity, particularly in content creation (Adeshola
& Adepoju, 2023). This Large Language Model-powered AI chatbot is designed to
interact with users by providing some creative and productive actions. Its capabili-
ties can be exemplified such as mimicking human-like behavior, addressing various
requests, creating texts as if they were written by a human and providing guidance.
It can also be used as a mathematical assistant for querying data, acting as a search
engine and mathematical knowledge base interface. Unlike typical chatbots, Chat-
GPT offers revolutionary features in terms of creativity (Rudolph et al., 2023). On
the other hand, ChatGPT has some limitations. The examples for these limitations
are inaccurate or misleading answers (Rasul et al., 2023), unethical issues related to
plagiarism (Kasneci et al., 2023), causing the decrease of creativity (Kasneci et al.,
2023), making incorrect calculations and inability of providing figures, shapes and
drawings (Wardat et al., 2023). Moreover, there is still need to examine and note the
mathematical capabilities of ChatGPT with different aspects of mathematical tasks
by benefiting from the advances in neural information processing systems as sug-
gested in the previous research (Frieder et al., 2024).
Despite of the inability of producing 3-dimensional models in ChatGPT, AI pro-
poses great potential for enhancing flexibility, ease and effectiveness of AR systems
by helping to create required codes in related 3-dimensional software. In other words,
integrating AI technologies into educational environments allows computer-based
learning systems to function as cognitive tutors, tools or learners, as well as aids in
facilitating policymaking (Hwang et al., 2020; Xie et al., 2017). As Yufeia et al. (2020)
suggest “the typical scenarios of artificial intelligence education applications include
intelligent tutor-assisted personalized teaching and learning, intelligent assistants such

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

as educational robots, children’s partners at home, intelligent assessment, mining and

intelligent analysis of educational data, learning analysis and learning, digital portraits,
and etcetera” (p. 550), AI can be used to assist AR activities. Furthermore, Devagiri
et al. (2022) emphasize the effect of using AI on AR systems by claiming “AI has the
potential to revolutionize AR applications” (p. 2).
Artificial Intelligence in Education (AIEd) has primarily focused on the advance-
ment of computers capable of performing cognitive tasks similar to those associated
with human minds such as learning and problem-solving. Moreover, according to
Kaplan et al. (2019), AI is defined as the capacity of a system to accurately interpret
and learn from provided data, and subsequently apply that knowledge to accomplish
specific objectives. This field has been the subject of scientific exploration for more
than three decades (Luckin et al., 2016). In addition to growing interest in under-
standing and enhancing the applications of AI in education particularly in academic
institutions, there is also expansion of usage in governmental sectors. As AI technol-
ogies have become more commonly used in education, instructors have the chance
to automate routine tasks and promptly engage with students. This advancement
facilitates a more adaptive and personalized approach to teaching (Chan & Zary,
2019). Specifically, the advancements in hardware, such as high-speed graphical
processing units and the availability of diverse software libraries, have spurred the
use of AI technologies. This is particularly evident in the growing field of learn-
ing research and the adoption of data analysis techniques. Moreover, to a significant
extent, the fact that AI has already entered the field of education rapidly and actively
shows that AI will have a great impact on the developments in the field of education
in the future (Chen et al., 2020). When the studies examining the effects of ChatGPT
as the most commonly used AI technology in education in the literature have been
examined, there have been increasing research emphasizing its important roles and
beneficial facilities (Aithal & Aithal, 2023; Javaid et al., 2023).
Rather than using AI as personal assistants, the teachers can use AI as profes-
sional assistants and cognitive tutors. They can design activities, materials and
new strategies to teach geometry and organize their lessons with the help of AI. AI
can be used as a cognitive tutor by guiding them and providing assistance related
to pedagogical, mathematical and technological issues even they are not expert in
using various software. The studies in the literature emphasize that ChatGPT can be
used effectively to design and plan lessons and materials, and conceptualize contexts
(Javaid et al., 2023; Rasul et al., 2023; Tlili et al., 2023).
Wardat et al. (2023) note that ChatGPT occasionally provides incomplete or
incorrect mathematical answers and has an insufficient understanding of geometry
because of its design as a chatbot. Its inability to draw or construct geometric fig-
ures limits its ability to demonstrate the properties of these figures (Wardat et al.,
2023) and improve the students’ geometrical thinking and learning effectively. On
the other hand, AR tools can provide effective models for geometrical objects to
help the students conceptualize and improve their geometric thinking. However, its
complex nature of necessitating to know programming languages, most of the stu-
dents avoid to use them. The deficiencies of AI and the complexity of AR tools can
be resolved by using them together in an integrated way. Hence, the present study
focuses on integration of AI and AR by designing AI assisted AR activities.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

1.3 Problem statement

Due to the effect of cognitive filter, the preservice teachers can have difficulties
in realizing and conceptualizing 3-dimensional geometrical objects on studying
2-dimensional environments such as textbooks. However, it can be argued that learn-
ers are able to conceptualize these objects more effectively when they are presented
with beneficial opportunities. In this respect, it is beneficial to provide learning
experiences near to real life so integration of technology into the lessons teaching
3-dimensional objects has become inevitable. However, even if the utilities of tech-
nology, the effects of cognitive filter can remain because of the nature of these objects
and their 2-dimensional representations (Shelton & Hedley, 2004). At that point, AR
provides beneficial learning opportunities to analyze and understand geometric solids
by producing their 3-dimensonal representations as realistic as possible and by mim-
icking reality using virtual elements. Furthermore, AR can have the learners realize
and analyze the elements of geometric solids by looking them from different points
of views by rotating them and looking at their 3-dimensional representations. Moreo-
ver, learners can analyze, comprehend, manipulate or interpret the elements and the
properties of 3-dimesional objects and the relationships of their elements mentally
with the help of these activities. Therefore, there occurs a need to learn how to use
AR and design activities via AR in teaching 3-dimensional objects effectively.
In spite of many advantages of using AR tools in education, it is difficult to design
research having students designed the activities and 3-dimensional models. In order
to perform this kind of research, the students have to learn how to construct 3-dimen-
sional models by AR tools including to learn related programming language, and to
gain experiences to produce necessary codes. In order to equip with these knowl-
edge, skills and experience to use AR, the students have to study hard and make a
lot of exercises. At that point, AI can provide beneficial opportunities to the students
as cognitive tutor and personal assistant. With the help of AI, they can effectively
analyze 3-dimesional objects and their properties and could use AR without know-
ing necessary programming language. Hence, AI can propose an effective way to
decrease costs, time period and human labor by increasing efficiency in learning
process. In this respect, AI assisted AR activities can create highly effective learn-
ing environments for understanding the concept. They can learn the concept by pro-
viding real life applications of abstract concepts without being professional on any
programming language and limited experience on AR. In the light of the explana-
tions, it is aimed to fulfill a gap in the literature, providing a new usage way and
strategy of AI in teaching and learning process in geometry lessons to the preservice
teachers. Hence, motivation of this study is about to represent the development of
preservice teachers’ (PSTs) geometric thinking on the learning environment paying
attention to learn to use AR with the help of AI by providing unique opportunities
and without knowing any programming languages. In other words, the purpose of the
present study is about to design and examine the developmental process of a learn-
ing environment to illustrate how Artificial intelligence (AI) assisted Augmented
Reality (AR) activities facilitate the development of the preservice teachers’ (PSTs)
geometric thinking. In order to represent the effects of AI assisted AR activities in
the development of geometric thinking more clearly, the topic of platonic solids is

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

focused on. In other words, by researching how and to what extent the PSTs learned
the new concept and improved geometric thinking through these activities, the impact
of AI assisted AR activities used in the study could be demonstrated more effectively.
Therefore, the activities were narrowed down to the topic of the platonic solids that
the PSTs did not have any prior knowledge about. The characteristics of the activities
about platonic solids in the present study determined based on theory in the literature
and are revolved into practice in AR activities assisted by AI. Through the study, to
examine and evaluate the effects and practicality of these activities were aimed.

2 Method

The present study represents an Educational Design Research (EDR) in detail. Used
in particularly instructional design including technology, EDR can be described as a
kind of research methodology progressing through systematic steps to anticipation,
enactment and evaluation of an interactive learning environment (Richey & Klein,
2014). In this EDR, the explanations are represented based on the evaluation part of
the research.

2.1 Research design

The purpose of this EDR is to focus on how Artificial intelligence (AI) assisted
Augmented Reality (AR) activities facilitate the development of preservice teach-
ers’ geometric reasoning. In other words, the main purpose of the present study is
to examine the effects of the interactive learning environment designed based on AI
assisted AR activities with regard to geometrical thinking and learning environment.
With this motivation, the answers for the following research questions have been
examined:

1. How effective is the interactive learning environment designed based on AI assisted

AR activities to develop geometrical thinking through learning platonic solids?
2. How practical is the interactive learning environment designed based on AI
assisted AR activities to be used in developing geometric thinking in the context
of platonic solids?

The present study was designed based on explanatory case study as a kind of
qualitative research method in order to provide answers to the research questions
accurately in detail. In this study, the cases are examined and explained focusing on
how and why related to them (Yin, 1994). Benefiting from the explanatory nature of
study, how the PSTs could develop their geometrical thinking in the context of geo-
metric solids with AI assisted AR activities was analyzed and explained in detail.
Through the process, the focus point was how they used AI to produce AR repre-
sentations about the platonic solids, what they experienced and how they transferred
and explained their practices into activity papers. Their thoughts and feelings were
also examined in order to detail these issues.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

2.2 The participants

With the aim of performing in-depth analysis with rich information, non-probabilistic
sampling strategy techniques were preferred as used in most qualitative research (Vogt
et al., 2012). Among these techniques, maximum variation and criterion sampling
techniques were used in order to select the participants. With this motivation, initially,
the grade levels and sex factors were considered in order to provide variation based on
maximum variation sampling method. Then, the criteria of not knowing any program-
ming language and having experience about using any AR system was used for crite-
rion sampling method. Hence, this EDR focusing on evaluation part was conducted
to 15 preservice teachers enrolled in elementary mathematics education in a public
university in a province in southern part of Turkey. In order to provide confidentiality
for the PSTs participating in the study, pseudonyms were used. The pseudonyms of the
participants were represented in the Table 1 with their demographic characteristics.

2.3 Data collection tools

In this research, the data were collected through the interactive learning environ-
ment designed with AI assisted AR activities, two semi-structured interview forms,
focus group interviews, anecdotes, activity forms and a test including open-ended
problems related to the platonic solids prepared by the researchers. Initially, the pre-
instruction interviews were performed with the PSTs before the enactment of inter-
active learning environment. After the completion of the pre-instruction interviews,
the PSTs were conducted to the PSTs at the first week of the instruction. Then, the
tasks were performed based on the designed activity forms through 5 weeks. At the
end of the tasks of the week, the focus group interviews were performed in each
week. After completing the enactment of interactive learning environment, Platonic
Solids Geometry Test was applied as the post-test. After the completion of the test
and instructions, the post-instruction interviews were conducted to the PSTs.

2.4 Interactive learning environment with AI assisted AR activities

It was aimed to represent how the PSTs improve their geometrical reasoning about
the platonic solids in an interactive learning environment with AI assisted AR activi-
ties developed by the researchers. Initially, the researchers selected the content of the

Table 1  Characteristics of the preservice teachers in the study group

Preservice Teachers’ Codes Grade Level Female Male

PST1, ­PST2 1 PST1 PST2

PST3, ­PST4, ­PST5, ­PST6, ­PST7, 2 PST3, ­PST4, ­PST5, PST8, ­PST9, ­PST10
­PST8, ­PST9, ­PST10 ­PST6, ­PST7
PST11, ­PST12, ­PST13, ­PST14 3 PST11, ­PST12 PST13, ­PST14
PST15 4 PST15

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

interactive learning environment by discussing possible topics that the PSTs do not
have any knowledge about and skills tend to have difficulty. In accordance with the
common views of the researchers, the content of the interactive learning environment
has been prepared based on the platonic solids. Secondly, in order to identify AR and
AI tools to be used in the study, the researchers examined and criticized some AR and
AI tools based on the criteria of ease of use, accessibility, using freely, practicality,
relevance and representing relevant and beneficial products, providing teaching mate-
rials with no costs and using without necessitating any specific devices. Afterwards,
the researchers discussed their experiences and observations about these tools and
selected Blender 4.0 (Blender Foundation, 2023) to create necessary files for AR tools
and ChatGPT 3.5 (Open AI, 2023) among AI tools in the present study. At that point,
because of the expense of smart glasses and the especially criteria of providing teach-
ing material with no costs and using without necessitating any devices, the researchers
decided to have the participants used mobile phones through the enactment of interac-
tive learning environment. In order to ease the transfer of platonic solids to the real-
ity, MyWebAR (2023) as a kind of AR tool was also used in the study. In order to
visualize the solids produced by Blender 4.0, MyWebAR was used so that they could
observe the solids by having mobile phones read the qr-code and transferred the solids
into the real world. Moreover, they could study about the content independent from the
enactment of interactive learning environment whenever and wherever they wanted.
Thirdly, the activity forms were prepared in paper-based and one form for each week
was designed and they were conducted to the participants in a computer laboratory.
The tasks were directed by these activity forms. Through the enactment of interac-
tive learning environment, in the first week, the researchers had the PSTs acquired
knowledge and experience about Blender 4.0, MyWebAR and ChatGPT 3.5 because
although they have limited knowledge and experience about ChatGPT 3.5, they were
not familiar with Blender 4.0 and MyWebAR. In this week, at the beginning of the
course, the researchers talked about AR and AI and their usage in education. Then,
the researchers made presentations about Blender, MyWebAR and ChatGPT 3.5 in
order to inform the PSTs. After they had listened the presentations and discussed
the tools, they used ChatGPT 3.5 as a personal assistant and tutor to install Blender
4.0 and MyWebAR into their computers. In other words, by using and following the
directions and explanations of ChatGPT 3.5 step by step, they installed them into
the computers by themselves without any guidance of the researchers. Afterwards,
in order to provide the PSTs experience using Blender 4.0 and ChatGPT 3.5, they
made practices on the interface of Blender 4.0 in which they produced the figures
and shapes installed in Blender 4.0 without necessitating to use any Phyton codes. In
other words, the first activity form directing the tasks of the first week was designed
based on producing some figures and solids whatever they wanted by using Blender
4.0 interface without using Phyton codes, and then they produced the qr-codes with
MyWebAR to represent in the real by mobile phones. Through the process, they fol-
lowed the steps explained by ChatGPT. The instructors wanted the PSTs to examine
Blender, learn and experience its properties, commands and segments. In order to
be familiar with Blender and acquire the knowledge and experience necessitated for
AI assisted AR activities of the present study, the PSTs were motivated. In Blender,
tools such as cleaning the scene, menus, basic modes, etc. were demonstrated. The

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

PSTs were also learned how to use ChatGPT to create Phyton scripts. After receiv-
ing Phyton codes from ChatGPT, instructions were given on how to use these codes
in Blender, what to do if they encounter errors, how to save the objects that they
produced to their computers (saving in Blender interface, saving with codes), how
to color each surface of an object differently with the codes, etc. It is noted that, the
PSTs achieved this process by using ChatGPT as a Blender assistant. Hence, they
were expected to acquire some knowledge and experience about AI and AR to inte-
grate them into their lessons in the future. In the second week, the activity form
directing the tasks was designed based on geometric solids including different types
of cylinders and cones using Blender 4.0, MyWebAR and ChatGPT 3.5. By placing
the activities about geometric solids as prior knowledge being necessitated to learn
the platonic solids, it was aimed to analyze and represent the PSTs’ development of
geometric thinking on platonic solids through AI assisted AR activities more clearly
in detail. The PSTs were expected to experience their knowledge and experience in
the context of the geometric solids. Hence, the PSTs examined the topics by expe-
riencing the tools of using Blender and ChatGPT 3.5, developing their geometric
and spatial reasoning about the geometric solids as the prerequisite knowledge for
platonic solids. By placing the activities about the geometric solids, the PSTs were
expected to acquire necessary prerequisite knowledge and motivate them to learn the
platonic solids by constructing new knowledge based on the known topics. Different
from the first week, they produced the solids in Blender 4.0 by borrowing Phyton
codes from ChatGPT 3.5 without benefiting from the figures and shapes existing and
installed in Blender. In the remaining weeks, the PSTs examined the platonic solids
including tetrahedron, hexahedron, octahedron, dodecahedron and icosahedron, and
produced them in Blender 4.0 and qr-codes in MyWebAR with the tutor of ChatGPT
3.5 similar to the tasks in the second week. Therefore, the tasks were performed in
the third, fourth and fifth weeks for the platonic solids similar to the tasks and steps in
the second week. However, different from the tasks of the second week, because the
PSTs were not familiar with the platonic solids, they used ChatGPT 3.5 as a cognitive
tutor in addition to technological tutor in these weeks. The interactive learning envi-
ronment including the tasks performed by using Blender 4.0 with MyWebAR as a
kind of AR tool and ChatGPT 3.5 as a kind of AI tool. The tasks were designed based
on van Hiele geometric thinking theory in order to support the PSTs’ learning and
geometric thinking development. They examined their appearance, critical properties,
common properties shared with other objects and related theorems. The PSTs were
expected to form the different types of platonic solids and examine their attributes
and properties by producing their 3-dimensional models by AI assisted AR activities.
Hence, they could acquire opportunity to improve their skills of learning, teaching
and thinking by learning through experiencing. Training including AI assisted AR
activities was enacted by two authors of the present study and lasted 5 weeks and
3-class hours in each week (150 min). The computer laboratory in which the PSTs
studied in and their production of the platonic solid using the qr-codes formed by
MyWebAR is illustrated in Fig. 1. Through the enactment of the interactive learning
environment, the researchers performing the instructions took anecdotes. After the
instruction of the week had been completed, they wrote their thoughts, feelings and
views based on their experiences and observations on the cards in each week.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

3 The platonic solids test

The Platonic Solid Test was composed of five paper-and-pencil problems (see
Appendix). These problems have been prepared based on van Hiele Geometric
Thinking Theory. These problems have been formed related to the properties of ini-
tial four levels. This theory explains “the van Hieles developed a five-phase classifi-
cation of instruction to help educators teach students to be more sophisticated think-
ers about geometry” (McGlone, 2009, p.32.). van Hiele Geometric Thinking Theory
details the geometrical reasoning through five levels with their own characteristics
(Mansi, 2003). The focus point explains geometric reasoning based on the appear-
ance of geometric constructs at Level 1 (Visualization/ Recognition); specific prop-
erties of figures and solids at Level 2 (Description/ Analysis); relationships between
and among properties of the constructs at Level 3 (Informal deduction); conceptu-
alizing geometric statements such as axioms, definitions, and theorems at Level 4
(Deduction). By considering the properties of Level 1, the first problem on Platonic
Solid Test is about to draw cube, pentagon prism, hexagon prism, triangular cone,
tetrahedron, hexahedron, octahedron, dodecahedron, and icosahedron. In this prob-
lem, rectangular prism, cube, pentagon prism, hexagon prism and triangular pyra-
mid are also asked even if most of them are not platonic solids in order to represent
the PSTs’ conceptualization about the difference between these geometric solids and
the platonic solids including triangular, quadratic, pentagonal and hexagonal faces.
The second problem is about to define these geometric solids and the platonic solids
by identifying their critical elements based on the properties of Level 2. The third
problem consists of explanations which are not completed and prepared as fill in the
blanks form related to Level 2. The problem is about to identify the type of the pla-
tonic solid meeting the properties explained in the problem (e.g. the platonic solid
is ……. with five equilateral triangles at each vertex). The fourth problem is about
to compare the properties of the geometric solids and the platonic solids such as the
number of vertices, faces, edges and the relationship among these properties by con-
sidering Level 3. In this problem, cube, pentagon prism, hexagon prism and triangu-
lar pyramid are also asked in order to have the PSTs facilitate understanding about
the identification of the similarities and differences between these geometric solids
and the platonic solids including triangular, quadratic, pentagonal and hexagonal

Fig. 1  The computer laboratory and the production of the platonic solid

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

faces. The fifth problem related to geometric thinking Level 3 is about to produce an
equation explaining the relationship between the number of vertices, edges and faces
of the platonic solids. In other words, the problem is about to discover the equation
that the subtraction of the number of edges from the addition of the number of ver-
tices and the number of faces is equal to 2 (V + F-E = 2). This equation is named as
the Euler’s formula. The sixth problem prepared by considering Level 4 is about to
prove the accuracy of this equation.
After the researchers had prepared the Platonic Solids Geometry Test, the views
of 5 experts having the Ph.D. degree in mathematics education (3 academicians)
and mathematics (2 academicians) have been benefited from in order to provide the
sufficiency, appropriateness and accuracy of the problems by considering van Hiele
Geometric Thinking Theory based on the suggestion of Veneziano (1997) explain-
ing the need of the views of 5 to 40 experts for content validity. This test was con-
ducted to the PSTs as pre-test and post-test before and after the enactment of the
designed interactive learning environment and lasted approximately 60 min in order
to identify and compare the PSTs’ development of geometrical thinking through the
process.

3.1 The interviews

In the present study, the semi-structured interviews were conducted to each of the
PSTs two times through the interactive learning environment. Initial interviews
as pre-instruction interviews were performed before the enactment of AI assisted
AR activities. This interview form has been prepared to collect the data about the
views and familiarity of the PSTs in the study group on the use of AI and AR in
the real life and especially in education and prior knowledge about the concept of
the platonic solids. In this respect, the pre-instruction interview form was prepared
consisting six questions asking about their prior knowledge and experience about
the platonic solids, AI and AR (see Appendices). The post-instruction interviews
were conducted to the PSTs based on their experiences on AI assisted AR activi-
ties. The post-instruction interview form was prepared consisting twelve questions
asking about their acquired knowledge and experience about these activities. After
the preparation of pre-instruction and post-instruction interview questions, the views
of 6 experts having the Ph.D. degree in mathematics education (3 academicians),
education of computer and teaching technologies (3 academicians) have been ben-
efited from in order to provide the sufficiency, appropriateness and accuracy of the
content of the interview questions based on the suggestion of Veneziano (1997). The
pre-instruction interviews lasted approximately 50 min and post-instruction inter-
views lasted approximately 80 min. Through the process of enacting AI assisted
AR activities focus group interviews were performed with the PSTs and one of
the researchers of the present study. The PSTs discussed their acquired knowledge,
experiences and views about the tasks performed based on the activity forms so that
their improvement through the enactment of AI assisted AR activities was exam-
ined. In this respect, the focus group interviews were performed in each week after
the enactment.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

of the tasks of the week based their knowledge, feelings and experience acquired
by engaging in the tasks of the related week. By having focus group interview in
each week, totally, focus group interviews were performed four time through the
enactment process and each of them lasted approximately 40 min. Through all of
the interviews, in order to prevent the loss of the interview data, a voice-recording
device was used.

3.2 Data analysis

In order to analyze the qualitative data collected through the Platonic Solids Test,
anecdotes and the interviews, content analysis method (e.g., Hsieh & Shannon 2005)
was used. For the PSTs’ responses to the pre-test and post-test and the explanations
for the pre-instruction and post-instruction interviews, two different processes of
content analysis method were performed. Initially, in order to perform content analy-
sis for the test, the PSTs’ responses to the pre-test and post-test were compared for
each problem for each of the PSTs separately. Their responses for each problem for
both applications of the tests were coded as correct, partially correct, wrong, and
no answer. Moreover, the changes on their responses to the questions were repre-
sented and explained in detail. The second period of content analysis was performed
to analyze the recorded interviews, activity forms and anecdotes. Initially, the inter-
views were transcribed verbatim and converted into hardcopies. All of the raw data
collected through the study have been divided into the codes and themes emerged
through the content analysis. The content analysis process was performed following
six-step process. Initially, the qualitative data collected through the interviews tran-
scribed were organized. Secondly, the organized data were analyzed to identify the
codes. In this study, the unit of analysis was determined as words and word groups
in the PSTs’ explanations. At that point, the researchers of the present study formed
their own code lists by analyzing the data and identifying the codes independently.
Then, the researchers compared their lists in order to form the exact code list. They
accepted common codes and discussed other codes until they reached a consensus.
For example, P ­ ST2 explained “I feel like I have a book in my hand guiding me step
by step to use a very difficult tool such as Blender.”. In analyzing this explanation,
the researchers used two different codes which were “guidance” and “pathfinder”.
They discussed and accepted to use the code of “operating manual”. The same steps
were performed all of the different codes. At the end of this step, a code list includ-
ing the codes accepted by all of the researchers were formed. Thirdly, the similar
codes were grouped in order to form categories/themes. The code schema includ-
ing codes and categories are represented in Table 2. Fourthly, all of the research-
ers analyzed all of the data by using the list of codes and categories independently.
Then, the analysis of the researchers were compared whether or not there existed
individual differences in the analysis process. Fifthly, the alternative explanations
for the codes, categories and the analysis results were discussed. Lastly, the analysis
process were reported and represented below.
In order to provide evidence for the trustworthiness of the analysis process, the
transcripts of the interviews and written documents of the tests were read and coded

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

by the researchers independently. The inter-rater reliability ratio value for the pro-
cess of coding independently was calculated to be about 93%. For the remaining
7% of the coding process representing the researchers’ disagreements on the codes,
all of the researchers discussed these codes and formed new codes by reaching con-
sensus. Through the content analysis performed by all of the researchers, investiga-
tor triangulation was provided. Moreover, member checking strategy was also used.
After completion of the content analysis and the implementation of the analysis,
the findings are discussed using direct quotations from the tests and explanations
through the interviews.

4 Findings

The findings of the current study are represented into three categories grouped
based on findings related to analysis of the PSTs’ responses to The Platonic Solids
Test in order to represent the development of geometric reasoning; and the focus
group interviews and activity forms used through AI assisted AR activities to pro-
vide evidence for the development of the geometric thinking and the practicality
of these activities in the context of the platonic solids; and the pre-instruction and
post-instruction interviews to provide evidence for the practicality of AI assisted AR
activities.

4.1 Findings obtained from the platonic solids test

The PSTs’ pre-test and post-test findings obtained by conducting the Platonic Solids
Test to the PSTs are presented in Table 2. In Table 2, the PSTs’ explanations for the
questions formed based on van Hiele Geometric Thinking Theory are illustrated by
grouping into correct (C), partially correct (PC), wrong (W), and no answer (NA)
in order to highlight the changes on their geometric thinking levels in the context of
the platonic solids before and after the implementation of AI assisted AR activities.
In the first problem related to geometric thinking Level 1, the PSTs are asked to
draw the geometric solids of cube, pentagon prism, hexagon prism, triangular cone

Table 2  The number of the PSTs responding to the platonic solids test
The Pre-Test Findings The Post-Test Findings
C PC W NA C PC W NA

The 1st problem (Level 1) 15* 15

The 2nd problem (Level 2) 15* 15
The 3rd problem (Level 2) 3* 4 8 12 3
The 4th problem (Level 3) 6* 2 7 15
The 5th problem (Level 3) 15 12 3
The 6th problem (Level 4) 15 4 5 6

*: no answer or effort for the problems of the platonic solids

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

and the platonic solids of tetrahedron, hexahedron, octahedron, dodecahedron, and

icosahedron. In the pre-test, the first time taking the Platonic Solids Test, all of the
PSTs’ responses to this problem was categorized as partially correct. All of them
drew the geometric solids of cube, pentagon prism, hexagon prism and triangular
cone correctly on the other hand there was no answer and effort for drawing the pla-
tonic solids of octahedron, dodecahedron, and icosahedron. Exceptionally, they were
able to draw tetrahedron and hexahedron. For this respond, the PSTs explained that
they could realize the connection of tetrahedron with triangular pyramid and hexa-
hedron with cube but they emphasized that they had difficulty in realizing these con-
nections. In the post-test, in the second time taking the Platonic Solids Test, all of
the PSTs were able to draw the geometric solids and the platonic solids accurately.
In the second problem related to geometric thinking Level 2, the PSTs were asked
to define the geometric solids of cube, pentagon prism, hexagon prism, triangular
cone and the platonic solids of tetrahedron, hexahedron, octahedron, dodecahe-
dron, and icosahedron by identifying their critical elements. In the pre-test, all of
the PSTs’ responses to this problem were categorized as partially correct since they
were able to define the geometric solids accurately by emphasizing their critical or
basic elements on the other hand there was no answer and effort for the definition
of the platonic solids of octahedron, dodecahedron, and icosahedron. Exceptionally,
they were also able to define tetrahedron and hexahedron by realizing the connection
of tetrahedron with triangular pyramid and hexahedron with cube. In the post-test,
all of the PSTs were able to define all of the geometric solids and the platonic solids
by stating their critical or basic elements accurately.
In the third problem related to geometric thinking Level 2, the PSTs were asked
to identify the correct solids stated in the explanation. In other words, the third prob-
lem consisted of the explanations prepared as fill in the blanks form. In the pre-test,
3 of the PSTs’ responded to this problem was categorized as partially correct since
they were able to identify the explanations made by the geometric solids. For exam-
ple, they accurately completed the explanations related to the geometric solids such
as “3 squares at each vertex gives a cube”. On the other hand, 4 of the PSTs pro-
vided wrong answers for all of the explanations and the identification of “there was
no answer and effort” was performed for 8 of the PSTs. In the post-test, all of the
PSTs were able to provide correct answers for the explanations about the geometric
solids and the platonic solids.
In the fourth problem related to geometric thinking Level 3, the PSTs were asked
to compare the properties of the geometric solids and the platonic solids such as
the number of vertices, faces, edges and the relationship among these properties. In
the pre-test, 6 of the PSTs’ responses to this problem were categorized as partially
correct since they were able to identify the properties of the geometric solids. For
example, they correctly identified the vertices, faces and edges of the geometric sol-
ids by explaining their connections and drawing their opened forms. On the other
hand, 2 of the PSTs provided wrong answers for all of the solids and the identifica-
tion of “there was no answer and effort” was performed for 7 of the PSTs. In the
post-test, all of the PSTs were able to provide correct responses by explaining the
relationship among the properties of the geometric solids and the platonic solids.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

In the fifth problem related to geometric thinking Level 4, the PSTs were asked
to produce an equation explaining the relationship between the number of vertices
(V), edges (E) and faces (F) of the platonic solids. In other words, the PSTs were
expected to discover the equation named as the Euler’s formula. In the pre-test, there
was no answer and effort for this problem on all of the PSTs’ tests. In the post test,
12 of the PSTs’ responses to this problem were categorized as correct since they
were able to discover the equality of “V + F–E = 2”. On the other hand, 3 of the
PSTs provided partially correct answers for this problem.
In the sixth problem prepared by considering geometric thinking Level 4, the
PSTs were asked to prove the accuracy of the Euler’s formula (V + F–E = 2). In the
pre-test, all of the PSTs provided no answer and effort for this problem. In the post
test, 4 of the PSTs provided the correct answers by proving the Euler’s formula.
On the other hand, 5 of the PSTs provided partially correct answers by writing the
beginning part of the proof of the Euler’s formula and 6 of the PSTs provided no
answer or effort for the problem.

4.2 Findings obtained from the experiences in the interactive learning

environment with AI assisted AR activities

At this part, the findings obtained from the analysis of the data collected through the
completed activity forms, anecdotes and focus group interviews are represented. In
the first week of the interactive learning environment, the researchers learned and
discussed AI and AR, and their usage in the real life and education. Blender, MyWe-
bAR and ChatGPT were introduced to the PSTs. After the presentations had been
completed, they initially used ChatGPT as personal assistant and made some prac-
tices individually. The PSTs talked about the geometric shapes, their definitions and
properties, and they did not have any difficulty. Then, the PSTs were asked to install
Blender and MyWebAR into their computers. After they had stated that they had
not been able to install them, they asked ChatGPT how to install them and ChatGPT
directed them. The steps and explanations provided by ChatGPT about installing
Blender into the computers as illustrated in Fig. 1. By the way, in Fig. 1, it is worth
to mention that ChatGPT also provides explanations and corrections for the PSTs
on the terminology of technical words. With the help of the directions explained by
ChatGPT, they easily installed and became Fig. 2 ready to use them.
Afterwards, the PSTs made some practices in Blender and MyWebAR with Chat-
GPT by forming some geometric figures and shapes existing in Blender without
necessitating to use any Phyton codes. The commands of cleaning the scene, menus,
basic modes, etc. in Blender were practiced with the help of ChatGPT. This process
can be exemplified by the practices of ­PST5 based on the data obtained from the
activity form and anecdotes related to the observations, ­PST5 asked ChatGPT “I do
not know how to use Blender, could you direct me in order to produce a cube in the
interface of Blender without using Phyton codes?”. Then, she followed the direc-
tions of ChatGPT and formed a cube in Blender. At that point, she produced the
cube with the buttons of Blender without any coding by benefiting from the cube
existing in Blender as illustrated in Fig. 3. Afterwards, she used MyWebAR in order

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Fig. 2  The explanations of ChatGPT to install Blender into computer

Fig. 3  The explanations of ChatGPT to produce a cube by Blender

to produce the qr-code of the cube to transform it into reality. With this aim, she
asked ChatGPT “I do not know how to use MyWebAR, could you direct me in order
to form the qr-code of the cube produced in Blender?” and ChatGPT explained the
process. Then, she produced the qr-code to examine it in the real with mobile phone.
Hence, all of the PSTs made practices in order to produce some geometric figures
and solids using Blender interface without using Phyton coding and form the qr-
code with MyWebAR to represent these shapes and figures in the real by mobile
phone effectively. Based on the observations and as confirmed by their explanations
through the first focus group interview, they emphasized the facilitating role (tutor-
ing in technological context) of ChatGPT in using Blender and MyWebAR in which
they had never been used and experienced. Also, they stated that they used Blender
and MyWebAR to produce the geometric shapes and figures easily with the help of
ChatGPT. They emphasized that they did not have too much difficulty through the
activities. Also, ChatGPT did not provide any direction causing errors or useless
steps. They also explained that some of the participants struggled because it was
their first experience with AR applications, and the application was quite complex

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

as they were trying to familiarize themselves with it for the first time. For exam-
ple, they encountered problems with the tasks such as clearing the scene, pasting
and running code, and opening a new code page (new text) in AR. However, they
overcame this technical problem by following the steps explained by ChatGPT. The
instructors told them to tell the problems they faced with in AR to ChatGPT and to
ask ChatGPT how to solve these technical problems.
In the second week, the PSTs made practices related to the geometric solids
including different types of cylinders and cones using Blender, MyWebAR and
ChatGPT. Hence, the PSTs produced these geometric solids by Blender, and then
produced the qr-code by MyWebAR with the help of ChatGPT. Different from the
activities and practices performed in the previous week, they produced the solids in
Blender by using Phyton codes however they produced these codes with the help
of ChatGPT since anyone of the PSTs did not know Phyton coding. This process
can be exemplified by the practices of P ­ ST13 based on the data obtained from the
activity form and anecdotes related to the observations. Initially, ­PST13 asked Chat-
GPT “I do not know Phyton coding, could you provide me Phyton codes in order to
produce a triangular pyramid in Blender?”. Then, he followed the instructions of
and used Phyton codes produced by ChatGPT to form a triangular pyramid. At this
process, when he used Phyton codes produced by ChatGPT in Blender, he observed
that the pyramid was not produced and Blender presented error. Then, he turned
back to ChatGPT. He copied the error in Blender and pasted it on ChatGPT. He
also told to ChatGPT “Your Phyton codes did not work, give me the correct codes”.
Then, ChatGPT provided the correct codes politely and then he formed a pyramid
and its opening form in Blender. Afterwards, he asked ChatGPT “By using MyWe-
bAR, could you direct me in order to form the qr-code for the triangular pyramid
(its opening form) produced in Blender?” and they could acquire the qr-code to
examine it in the real with mobile phone. Based on the observations and as con-
firmed by their explanations through the focus group interview, they emphasized
the effective role (tutoring in technological context) of ChatGPT in using Blender
to produce Phyton codes. Also, they stated that they used Blender and MyWebAR
to produce the geometric shapes and figures easily with the help of ChatGPT. They
had some difficulty in producing correct Phyton codes since ChatGPT produced
the codes that did not work and caused the errors in Blender. In the focus group
interview, the PSTs explained their experiences and thought about the tasks. They
also talked about the facilities and difficulties observed through the process by pro-
viding evidences including their studies on ChatGPT, Blender and MyWebAR. For
example, ­PST5 expressed:
I asked ChatGPT to produce the codes for Blender to form a triangular pyra-
mid. It explained the processes step by step from beginning such as opening
Blender and switching to the scripting workspace to the end in which the codes
were presented. Then, I copied the codes from ChatGPT and pasted them in
Blender. However, something went wrong. The codes did not work and Blender
represented the error. To conclude, I could not produce the triangular pyramid
by Blender.
The explanations of ­PST5 were illustrated in Figs. 4 and 5.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Fig. 4  The dialogue between ChatGPT and ­PST5 about the codes for triangular pyramid

Fig. 5  The codes produced by ChatGPT and caused error in Blender

PST5 continued to explain her experience and tasks as follows:

I copied the error appeared in Blender and pasted it in ChatGPT. I didn’t think
it would work. But to my surprise it worked. When ChatGPT saw the error
code, it immediately realized that it made a mistake and gave the incorrect
codes. It politely admitted that it had made a mistake, apologized and given
the correct codes. When I copied these new codes and pasted them in Blender.
It worked and triangular pyramid was formed.
The explanations of P­ ST5 were illustrated in Figs. 6 and 7.
As discussed in the focus group interviews and observed in the Figs. 4, 5, 7 and 7,
the most significant technical problem observed in AI assisted AR activities was that
ChatGPT could not always produce the correct Phyton codes working in Belnder on
the first try. When the errors observed in Blender were directly transferred to Chat-
GPT, ChatGPT could analyze the errors and produced the correct codes. The most
observed technical problem related to the errors resulted from producing incorrect

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

codes experienced in AR were solved by the ability to analyze and correct the errors
of ChatGPT.
In the following weeks, the PSTs examined the platonic solids including tetrahe-
dron, hexahedron, octahedron, dodecahedron and icosahedron, and produced them
in Blender and their qr-codes in MyWebAR with the tutor of ChatGPT similar to
the tasks in the previous week. However, different from the tasks performed in the
second week, the PSTs talked to ChatGPT about the platonic solids since they were
not familiar with the platonic solids. For example, ­PST1 initiated a discussion with
ChatGPT by asking “What is dodecahedron? Define it.” By considering explana-
tions made by ChatGPT, she stated that she was not able to imagine it. Then, she
asked ChatGPT again “What are its critical elements? Explain its properties such
as edges, vertices and faces etc.” and the answers of ChatGPT were illustrated in
Fig. 8. In this process, they used ChatGPT as a cognitive tutor as it was observed
and illustrated in Fig. 8.
Afterwards, she asked ChatGPT “I do not know Phyton coding, could you provide
me Phyton codes in order to produce a dodecahedron in Blender?” similar to the
tasks in the second week. She followed the instructions and used the Phyton codes
in order to produce it by ChatGPT. At his process, the codes provided by Chat-
GPT did not work in Blender. Blender illustrated errors and dodecahedron was not
formed. Hence, she turned back to ChatGPT and explained the errors. Differently,
with the help of their experience in and knowledge about Blender and ChatGPT, she
proceeded by directing ChatGPT to ask if it was possible to throw out the inaccurate
parts of the codes and make edits on the particular parts of these codes. Then, Chat-
GPT provided the accurate codes by making the revisions on the previous codes and
formed a dodecahedron in Blender by transferring these codes from ChatGPT into
Blender. Through this process, the PSTs were encouraged to examine the codes and
analyze the differences between the correct and incorrect codes.
Different from the activities in the previous weeks, the PSTs paid attention to
coloring the faces with various colors. With this aim, they asked ChatGPT “Could

Fig. 6  ChatGPT determined the errors and produced the correct codes

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Fig. 7  Applying the correct codes of ChatGPT in Blender

Fig. 8  Some of the explanations of ChatGPT about dodecahedron

you produce the Phyton codes for Blender in order to make each surface of an octa-
hedron a different color?”. By transferring the codes of ChatGPT into Blender, it
was observed that ChatGPT produced the codes including errors and the faces of
the solids did not become colorful. At this process, the codes provided by ChatGPT
did not work in Blender. Blender illustrated the errors and octahedron with colorful

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

faces was not formed. Hence, they turned back to ChatGPT and explained the errors.
They told ChatGPT to throw out the inaccurate parts of the codes and make revi-
sions on the particular parts identified by her. Then, ChatGPT provided the accurate
codes by making revisions on the previous codes accurately and having each face of
the octahedron colored differently in Blender by transferring these codes from Chat-
GPT into Blender. Through this process, the PSTs were asked to analyze the accu-
rate and inaccurate codes and identify the parts causing errors. Afterwards, they also
asked ChatGPT “By using MyWebAR, could you direct me in order to produce these
qr-codes on the octahedron for Blender?”. Through the observations and the focus
group interview, they emphasized the facilitating roles of ChatGPT as technologi-
cal tutor in using Blender and MyWebAR, and as cognitive tutor in conceptualiz-
ing the platonic solids. Through the process, they had some difficulties in producing
the codes to form the platonic solids and color the edges, and in transferring them
into Blender. However, even if ChatGPT caused errors, it could revise the errors and
accurately produce the codes to form the platonic solids in the correct way. In the
following weeks, through engaging in the tasks related to the production of the other
remaining platonic solids, the PSTs performed similar practices and experienced
similar actions and difficulties as it was observed through the activities in the pre-
vious weeks, and they explained in the focus group interviews. In one of the focus
group interviews, all of the PSTs except ­PST13 stated similar explanations in the
following focus group interviews. In coloring the faces of octahedron with different
colors, he asked ChatGPT to provide the codes for coloring in Blender. ChatGPT
produced the inaccurate codes causing error. Instead of coping the error on Blender
and pasting it on ChatGPT, he asked ChatGPT “…the codes do not work. They are
wrong. Provide the accurate codes.” ChatGPT kindly apologized and produced
the new codes. However, the new codes did not work and they caused the errors in
Blender again. He tried this way a few more times and ChatGPT produced the incor-
rect codes in each time. Then, he copied the errors appeared in Blender and pasted
them on ChatGPT as performed by the other PSTs. At this point, ChatGPT provided
the correct codes producing octahedron with colorful faces in Blender as illustrated
in Fig. 9. In these weeks, as in the previous weeks, the most frequently observed
technical problem was that ChatGPT was not able to produce the correct Phyton
codes in the first trial. However, when the errors were transferred to ChatGPT, its
ability to analyze the errors and produce the correct Phyton codes demonstrated that
this technical problem could be easily resolved. Moreover, the explanations of Chat-
GPT for the errors in the codes showed that it provided the participants with instruc-
tive information about Blender. In this way, the participants acquired knowledge
about Blender and Phyton codes by analyzing the errors; that is, they learned how to
check and analyze, long Phyton scripts from the error analysis of ChatGPT.
We remark that, the ChatGPT dialogues with one of the PSTs can be found in Appen-
dices. After creating the platonic solids in Blender, they discovered how to display them
using MyWebAR by following the directions in ChatGPT as shown in Fig. 10.
In the focus group interviews, the PSTs discussed using ChatGPT in order to ana-
lyze the codes to be used in Blender and provided by ChatGPT. In AI assisted AR
activities, they asked ChatGPT to explain the codes on each line and the meanings
of the terms used in the codes. ChatGPT answered their questions clearly in detail so

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Fig. 9  The procedures of forming octahedron by ChatGPT and Blender

Fig. 10  Creating octahedron by MyWebAR with the directions of ChatGPT

that they used it as a technological tutor. For example, P­ ST2 explained how he used
ChatGPT to analyze and learn something about Phyton codes used in producing the
platonic solids. After producing the codes for coloring the faces of octahedron, I
asked ChatGPT how to use these codes. Then, he wanted ChatGPT to explain the
codes in the first line, the second line etc. He also asked the meanings of the terms

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

such as vertices used in Phyton codes. ChatGPT expressed the answers of their ques-
tions clearly in detail as illustrated in Fig. 11. In addition to providing Phyton codes
to be used in Blender in producing the platonic solids, ChatGPT helped the PSTs
analyze the meanings of these codes. In this process, the PSTs used ChatGPT as a
cognitive tutor in the technological content in a different way.

4.3 Findings obtained from the semi‑structured interviews

In order to provide evidence for the practicality and the effects of the interactive
learning environment with AI assisted AR activities, the PSTs’ views collected
through the pre-instruction and post-instruction interviews were analyzed. Through
the interviews, all of the PSTs emphasized that the enactment of interactive learning
environment with AI assisted AR activities were satisfactory and beneficial. When
their explanations in the interviews are analyzed in detail by identifying the codes
and categories, these analysis has been gathered under 4 categories as “cognitive
dimension”, “affective dimension”, “technological dimension” and “employability
dimension”.
The PSTs’ views about AI assisted AR activities based on their experiences have
been evaluated in the categories and codes in Table 3. It has been observed that
all of the PSTs think that AI assisted AR activities create positive effects on these
dimensions.
When the evaluations are presented separately based on these dimensions, the
PSTs emphasize that AI assisted AR activities result in positive effects on the cat-
egory of cognitive dimension. The sample explanations related to the PSTs’ views
on this dimension are as follows:

PST13… Through the activities, I realized that defining the solid sufficiently and
accurately is important and critical since the clarity of the prompts in ChatGPT
results in accurate solutions. When I asked ChatGPT to produce the code for
dodecahedron, there were errors for the codes. On the other hand, when I asked

Fig. 11  The explanations of ChatGPT in analyzing Phyton codes

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

to produce the code by defining it explaining its edges, vertices, faces, and its
properties, ChatGPT produced the accurate codes not causing errors. Hence, I
think that ChatGPT directs me to know and explain the solid sufficiently in detail
by using correct words so that I can conceptualize the solid well by explaining it
regularly….
PST7…I did not know anything about the platonic solids before the course.
With the help of ChatGPT, I learned their properties and critical elements.
Even if I discussed them with ChatGPT, I was not able to imagine them. Hence,
I produced their models by Blender and MyWebAR using the directions of
ChatGPT. These models are so beneficial to learn them and analyze the solids
and their properties in detail and effectively. I learn by taking role and being
active in every step.
PST6…ChatGPT sometimes produced information including errors or missing
parts. Hence, I have to criticize the explanations of ChatGPT. I need to analyze
whether the explanations are accurate, completed and serving or not. I think this
situation is useful since ChatGPT encourages me to search for the content and
learn it effectively.

When the PSTs’ explanations are analyzed in affective dimension, it is observed

that all of the PSTs state that the interactive learning environment with AI assisted
AR activities result in having fun, increasing efficacy belief, decreasing anxiety,
increasing motivation, and comfort to study with an advisor. Some of the PSTs’
views are as below:

PST15… When I see products created with AR tools, I think they are very useful
for use in education. However, when I examined my friends who used the appli-
cation, I was intimidated. I think it is very difficult and troublesome to learn how
to use it… When I heard about the course, I was very scared at first. But I’m glad
that I attended. I can use AR tools that I was afraid of using in the past. I will use
it easily when I become a teacher. It was great fun to learn about three-dimen-
sional objects such as platonic solids in this way.

Table 3  Categories and codes in AI assisted AR activities

Categories Codes

Cognitive dimension Conceptual knowledge, tutoring about geometrical content, critical thinking,
teaching from errors, learning individually, accessing current information,
learning by doing
Affective dimension Having fun, increasing efficacy belief, decreasing anxiety, increasing motiva-
tion, comfort to study with an advisor
Technological dimension Not necessitating technological knowledge and experience, easily formed
products, solving technical problems, revising technical errors, attaining
technical knowledge and experience individually, tutoring about technical
content, technological literacy
Employability dimension No cost, less effort, less practice, availability, easily usage, independence
from time and instruction, operating manual, coding booklet

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

PST12… The thought of using technology in real classes scares me. AR is a

complex tool to use. Whenever I encounter any problems during the process, I
suddenly become anxious. However, when I work with ChatGPT, I feel relaxed
because I have the confidence that someone knowledgeable will guide me in the
right way and solve my problems.
PST2…When I face with a problem, error or difficulty in using technological
tools, I feel like I am lost. I leave engaging in these tools since I feel tension.
While using Blender, I feel tension when I face with a problem even if using Phy-
ton codes produced by ChatGPT. However, ChatGPT helps me solve this problem
and remove the errors so that my tension becomes lost.

In the category of technological dimension, all of the PSTs has explained that AI
assisted AR activities provide some facilities such as not necessitating technologi-
cal knowledge and experience, easily formed products, solving technical problems,
revising technical errors, attaining technical knowledge and experience individually,
tutoring about technical content, technological literacy. Sample views on this topic
are as below:

PST1…It is great to produce models for the platonic solids in Blender without
knowing Phyton coding easily. Learning Phyton coding takes longer than six
months by practicing hard…In experiencing Blender by attaining the codes from
ChatGPT, even if I do not know Phyton coding, I can analyze and evaluate the
produced codes. Now, I know the general principles of coding.
PST8…Through the process, ChatGPT produced errors in the codes. I was not
able to identify the codes. At that point, in addition to producing codes, ChatGPT
worked as a software developer. It modified the errors and produced the accurate
codes. In the real, it is hard to find a software developer but I produced the solids
by discussing ChatGPT about programming and solving programming problems
individually at basic level.
PST13… At the beginning of the course, I did not know anything about Phyton
codes and AR tools of Blender and MyWebAR. However, I can produce Phyton
codes to use in Blender right now, of course, by using ChatGPT. Without Chat-
GPT, I cannot write Phyton codes. On the other hand, now, I have been famil-
iar with codes and I can understand and analyze the codes by acquiring general
knowledge about Phyton but not very well. Also, I can use Blender and produce
the objects with its interface.

In the category of employability dimension, all of the PSTs have explained that
AI assisted AR activities provide some facilities such as no cost, less effort, less
practice, availability, easily usage, independence from time and instruction, operat-
ing manual, and coding booklet. Sample views on this topic are as below:

PST2… There is no need to take any courses or do a lot of practice about AR

tools or programming languages to produce 3-dimensional shapes as taught in
the course. I feel like I have a book in my hand guiding me step by step to use a
very difficult tool such as Blender. It is such a book that it contains all the infor-

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

mation and codes I may need. I can form 3-dimensional shapes by producing
them in Blender easily in a short time and use them in my classroom in the future
effectively. This is wonderful!
PST15…As it is known, in order to use Blender, I need to know Phyton program-
ming language. Learning this language takes a lot of time and costs a lot. With
the help of AI assisted AR activities, it is possible to use Blender in any time I
want without taking any course about Phyton Programming language, of course,
at basic level. Moreover, in this way, using Blender becomes so easy and useful.

5 Conclusions

In the present research, practicality and effective aspects of the interactive learning
environment with AI assisted AR activities designed by the researchers based on
the context of the platonic solids have been examined based on the development of
geometric thinking. Along with this, The PSTs’ responses to The Platonic Solids
Test, activity forms used through the interactive learning environment and the focus
group interviews related to the activities, anecdotes taken by the researchers through
observing the enactment of the activities and the pre-instruction and post-instruction
interviews to provide evidence for their development of geometric thinking and the
practicality of AI assisted AR activities in the context of the platonic solids have
been investigated in detail. The contributions of the designed interactive learning
environment with AI assisted AR activities on the PSTs’ geometric thinking in the
context of platonic solids was asserted in the study.
In order to represent the development of the PSTs’ geometric thinking levels in
the context of the platonic solids, the PSTs’ responses to the Platonic Solids Test
conducted before and after the enactment of the interactive learning environment
with AI assisted AR activities were analyzed. In the test, in order to clearly ana-
lyze the PSTs’ geometric thinking skills in 3-dimensional solids as necessitated
prior knowledge for the platonic solids, these questions have been formed in addi-
tion to the questions about the platonic solids. The PSTs’ responses to the pre-test
questions, it was observed that the PSTs’ geometric thinking levels were higher than
Level 2 (The questions about the geometric solids prepared related to the properties
of Level 1 and Level 2 were answered correctly by the PSTs). On the other hand, it
was observed that the PSTs could not answer the questions about the platonic solids.
They could only make some efforts for tetrahedron and cube with the help of their
connection with the geometric solids. Based on the analysis of their answers for the
platonic solid problems, it can be stated that they were at Level 1 and they did not
have any knowledge about them and even know what they look like. After the enact-
ment of the interactive learning environment had been completed, the PSTs were
conducted to the Platonic Solids Test as the post-test. The PSTs’ responses to the
post-test questions were examined, it was observed that all of the PSTs could acquire
the properties of Level 3 geometric thinking and 4 of them reached Level 4 in the
context of the platonic solids. To conclude, AI assisted AR activities facilitated the
development of the PSTs’ geometric thinking levels.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

In the present study, the PSTs used ChatGPT to discuss and learn to acquire
conceptual knowledge about the platonic solids, and technological knowledge and
experiences about using Blender and producing the necessary Phyton codes. The
PSTs used ChatGPT as technological and cognitive tutor through engaging in AI
assisted AR activities. In other words, ChatGPT was used as a tutor to create neces-
sary scripts for the platonic solids in Blender and then visualize them in MyWe-
bAR. The process representing how the PSTs’ geometric thinking levels developed
was represented benefiting from the tasks performed based on the activity papers,
anecdotes and focus group interviews. Moreover, the PSTs explained their expe-
riences, thoughts and feelings through the focus group interviews and interviews.
Their experiences related to producing codes and revising the errors on them were
represented by direct quotations as represented above. Moreover, their thoughts and
feelings were analyzed by the content analysis and the themes were emerged; cog-
nitive dimension, affective dimension, technological dimension and employability
dimension. Hence, the practicality and effective aspects of the interactive learning
environment with AI assisted AR activities were represented in detail by providing
evidences benefiting from the data.

6 Discussions

Based on the comparison of the PSTs’ responses to the Platonic Solid Test enacted
as the pre-test and post-test, it was determined that their geometric thinking levels
were increased from Level 1 to Level 3 and Level 4 in the context of the platonic
solids. These findings could illustrate that AI assisted AR activities were beneficial
and effective to help the PSTs imagine the platonic solids, examine their properties
by making comparisons among the solids and discover their properties. At that point
it can be concluded that the PSTs could develop their geometric thinking levels with
the help of AR activities. The development observed on the PSTs’ geometric think-
ing level resulted from engaging in AR activities. Through engaging in AR activi-
ties, the PSTs could easily produce the models for the platonic solids to examine
them, analyze and compare their properties, and discover the rules or formulas about
them. They could develop their geometric thinking levels with the tasks by analyz-
ing the models of the platonic solids from different perspectives at their own pace,
and conceptualizing the transitions between the platonic solids by comparing them
(İbili & Şahin, 2016). To conclude, it was observed that the PSTs’ geometric think-
ing levels and conceptualization were improved in the context of the platonic solids
with the help of the enactment of the interactive learning environment including AR
tools in the present study. Moreover, it was observed that they could understand the
platonic solids rather than visually recognizing them but also transferring them into
2-dimensional plane, and reach a level where they could discuss the properties and
rules about the content. This finding is parallel to the results of the previous research
focusing on the effects of AR tools on geometric thinking (İbili et al., 2020; Kera-
walla et al., 2006). Through the research, it was also observed that the PSTs’ per-
formances on the activities were supported by AR related to the platonic solids as

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

it was suggested in the previous research (Kaufmann, 2004; Kaufmann & Dünser,
2007).
In the present study, the PSTs used ChatGPT in conceptualizing the platonic sol-
ids as a cognitive tutor and to form Phyton codes in producing the platonic solids
by Blender and MyWebAR as a tutor in technological contexts. Through the pro-
cess, ChatGPT as a chatbot providing the opportunities of asking questions related
to various contexts, discussing any topic, requesting a solution and consulting for
any issue (Kasneci et al., 2023; Rospigliosi, 2023; Strzelecki, 2023). The property
that ChatGPT is a kind of text-based artificial intelligence tool (OpenAI, 2023) pro-
vides contribution to the PSTs’ conceptualization and geometric thinking about the
platonic solids by using ChatGPT as a cognitive tutor. It explains the platonic solids
by converting the geometric representations into verbal expressions. The PSTs asked
about the platonic solids and then ChatGPT explained these solids by verbal expres-
sions (see Fig. 8). However, it was observed that the expressions of ChatGPT were
not necessary for the PSTs in order to conceptualize the platonic solids. Moreover,
they necessitated to study with 3-dimensional models for the platonic solids, and
Blender and MyWebAR were used to produce these models. At that point, ChatGPT
provided another beneficial way by expressing how to use Blender and MyWebAR,
and forming necessitated Phyton codes. In the study, it was observed that even if the
beneficial geometrical and technological expressions were produced by ChatGPT,
inaccurate Phyton codes and expressions were formed by ChatGPT. Through the
interactive learning environment with AI assisted AR activities, the PSTs developed
their technological literacy. They learned to use Blender and become familiar with
Phyton codes by using ChatGPT as a technological tutor. Hence, the PSTs neces-
sitated to criticize and evaluate what they read on ChatGPT and they realized this
case. Moreover, they emphasized the development of critical thinking skills by eval-
uating and analyzing the expressions of ChatGPT and attaining a learning way to
analyze the errors in the expressions. The errors of ChatGPT were observed mostly
on Phyton codes. This finding is parallel to the previous research emphasizing that
ChatGPT could produce wrong answers for the problems (Chiu, 2023) and incorrect
information about a topic (Zhang et al., 2023). Furthermore, in the present study,
some of the PSTs could easily acquire the codes on their first try on ChatGPT. On
the other hand, the others could attain the codes by more than two tries and using
different questions on ChatGPT. For example, when the PSTs asked ChatGPT to
give the correct codes since the previous codes had not worked, ChatGPT produced
incorrect codes again. On the other hand, ChatGPT provided correct codes when
the errors in Blender was transferred into ChatGPT. Briefly, when ChatGPT had the
opportunity to analyze the error, it could produce the correct codes. This case can be
stated by the fact that ChatGPT can produce accurate, clear and detailed expressions
and answers only if asked detailed questions (Lo, 2023). Moreover, it is necessary to
realize that ChatGPT can produce changing expressions about a particular problem
based on the specificity, accurateness and preferred words (Rospigliosi, 2023).
The PSTs in the present research emphasized the positive effects of the designed
interactive learning environment with AI assisted AR activities on their improve-
ment of their geometric thinking, competency, their thoughts and experiences.
In order to support these findings, there occurs many research representing the

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

important supportive effects of using interactive learning environments with AR

tools (İbili et al., 2020; Kaufmann, 2004; Kaufmann & Dünser, 2007; Kerawalla
et al., 2006) and AI tools (Aithal & Aithal, 2023; Javaid et al., 2023; Kasneci et al.,
2023; Rasul et al., 2023; Wardat et al., 2023; Tlili et al., 2023; Urhan et al., 2024).
Moreover, it is well known that one of the drawbacks of usage of technology is that
the users need to learn how to use the specific software/tools not only in mathemat-
ics but also in other educational disciplines. For this reason, many teachers do not
prefer to bring the technology into their classrooms. However, integration of the
software/tools (e.g. Blender 4.0 and MyWebAR in the current case) with AI (e.g.
ChatGPT 3.5 in the present study) can help them use technology effectively without
much technical knowledge of the software/tools. With this motivation, in the present
study, the effects and practicality of using these two kinds of tools in an integrated
way such as Blender and MyWebAR as a kind of AR tool assisted with ChatGPT as
a kind of AI tool were examined based on the PSTs’ views stated through the inter-
views and focus group interviews. Their views were analyzed and represented by
grouping into 4 categories as “cognitive dimension”, “affective dimension”, “tech-
nological dimension” and “employability dimension”. Through the enactment of AI
assisted AR activities, the PSTs could produce effectively and beneficially 3-dimen-
sional model for the platonic solids by Blender and MyWebAR with the help of
ChatGPT. This finding can be encouraged by the findings of the previous research
emphasizing the role of ChatGPT in designing lessons and materials (Javaid et al.,
2023; Kasneci et al., 2023; Rasul et al., 2023; Tlili et al., 2023).
As much as the effects of AI assisted AR activities on the cognitive and tech-
nological dimensions, there have been positive effects on the affective dimension.
In the interviews and focus group interviews, all of the PSTs made expressions
related to the positive contributions of AI assisted AR activities to their emotional
improvement by emphasizing the issues of “having fun, increasing efficacy belief,
decreasing anxiety, increasing motivation, comfort to study with an advisor”.
There have been research parallel to this finding in the literature highlighting posi-
tive emotional effects of AR (Claros-Perdomo et al., 2022; Fuchsova & Korenova,
2019; Gomez‐Rios, Paredes‐Velasco, Hernández‐Beleño, & Fuentes‐Pinargote,
2023) and AI (Aktay et al., 2023; Kamite et al., 2019; Kolchenko, 2018; Oranga,
2023). Moreover, the PSTs’ expressions analyzed using these codes have showed
that Blender, MyWebAR and ChatGPT as kinds of technological tools have helped
them improve their emotional development especially in the field of using tech-
nology as it has been emphasized in the previous research (Schreyer-Bennethum
& Albright, 2011). The most emphasized codes by the PSTs are having fun and
increasing motivation in the present study. When the previous research are exam-
ined, it is observed that nearly all of the studies have focused on the opportunity
of the AR and AI tools of being interesting and motivating (Kamite et al., 2019;
Kolchenko, 2018; Yao et al., 2024).
In the present study, in addition to the positive effects of AI assisted AR
activities on the “cognitive dimension”, “affective dimension”, and “technological
dimension”, the PSTs emphasized the important roles of the activities related to
the opportunities of no cost, less effort, less practice, availability, easily usage,
independence from time and instruction, operating manual, coding booklet under

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

the theme of “employability dimension”. In the findings, it was observed that the
PSTs tended to avoid using technological tools especially AR tools because of their
complex nature such as necessitating to Phyton language through pre-instruction
interviews. In other words, despite living in an era where technology impacts every
aspect of our lives, the teachers involved in the research reported insufficient levels
of technological literacy and highlighted internal factors affecting technology use
in learning environments. This inadequacy among teachers has negatively impacted
the effective use of technology in educational settings. Numerous studies in the
literature support this finding (Salehi & Salehi, 2012). In the present study, with the
help of ChatGPT, the PSTs could use Blender without knowing Phyton language
and become familiar with Blender and Phyton language by no cost, less effort, less
practice in the present study. Moreover, by following the errors in Phyton scripts
together with ChatGPT, they learned following long scripts (the process and the
mission of each row through long scripts), making necessary corrections throughout
the algorithm and updating Phyton codes. It is noted that, this experience was not
in expert level but in almost competent level. Hence, this kind of AI assisted AR
activities can be benefited from in the courses such as teaching technologies in
teacher education programs. Moreover, AI tools can be used beneficially to support
the course of programming such as the course of Phyton language. At the end of
the interactive learning environment with AI assisted AR activities, the PSTs stated
that they have been motivated to use AR tools and willing to use and examine
new tools. These changes observed through this research can be supported by the
previous research in the literature (Kolchenko, 2018). Moreover, as it has been
observed in the present study, AI assisted AR activities can effectively reduce costs
and enhance efficiency during learning the mathematical contents. By providing the
right information at the right time and in the right format without distracting the
user, these activities can lower cognitive load and support human labor (Abraham
& Annunziata 2017), helping the PSTs can acquire necessary knowledge and
experience about AI and AR tools.
In the present study, the PSTs used ChatGPT as cognitive tutor and technologi-
cal tutor. They discussed the platonic solids and when they asked the attributes and
critical properties of the platonic solids, ChatGPT provided correct explanations.
Moreover, the PSTs mostly used ChatGPT as technological tutor through the activi-
ties. They used it as a guide since it explained the steps about installing Blender,
understanding and experiencing its commands and menus accurately. In producing
3-dimensional models for the platonic solids by Blender, ChatGPT provided Phyton
codes. However, some of Phyton codes did not work. At that point, this was not
a critical problem since when Phyton codes were transferred to Blender, Blender
tested the codes and sometimes produced the errors. After transferring the errors
in Blender to ChatGPT, the errors could be analyzed by ChatGPT and the accurate
codes could be produced. Hence, it can be stated that the explanations provided by
ChatGPT are not always accurate. This issue is supported by the previous study.
Wardat et al. (2023) emphasize that ChatGPT can produce incomplete or incor-
rect answers and explanations. Therefore, it is important to check the accuracy of
the explanations provided by ChatGPT and while the PSTs acquire knowledge and
experience about using ChatGPT, they should increase their awareness to check its

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

accuracy. We also note that the students also verified and cross-checked the written
knowledge of ChatGPT (e.g. the characteristic features of a specific solid) with the
visual data in Blender.

6.1 Limitations and recommendations

In this study, we examined the effect of AI assisted AR activities on the PSTs’

geometrical thinking in the context of the platonic solids. However, we could not
examine many other theorems and rules about the platonic solids in detail. There-
fore, future studies can be designed with AI assisted AR activities focusing on the
related theorems and rules and their proofs so that their development of geometric
thinking levels can be improved more effectively about the platonic solids. In order
to analyze and represent the long-term impact of the AI assisted AR activities on
the PSTs’ geometric thinking levels and retention of geometric concepts, a follow-
up study can be performed. It is noted that, one of the limitations of this study is
the absence of a control group. In the future research, this issue can be revisited
with a study that includes a control group. This would allow for a more detailed
examination of the effects of AI assisted AR activities used in this research. In
addition, the development of geometric thinking skills of the participants from dif-
ferent grade levels from primary school to university grade level can be examined
in different mathematical contexts in the future studies. Moreover, the interactive
learning environment was not performed by a large group of PSTs. Thus, these
activities can be conducted to a large group of PSTs or inservice teachers to clearly
represent their effects. The effectiveness and practicality of AI assisted AR activi-
ties have been examined using qualitative data. Benefiting from quantitative data
acquired through analysis of a rubric focusing on this issue can provide more ben-
eficial and detailed information. Hence, in the further research, a rubric analyzing
the practicality and effectiveness of AI assisted AR activities can be developed and
used so that the usage of these activities in the classrooms and examination of them
in the future studies can be encouraged. Moreover, in the present study, the train-
ing process was conducted as an individual working process for the participants. In
the future studies, this process could be carried out with student-to-student interac-
tion that includes discussions among the participants. This way, the contribution of
interaction can also be examined in detail. Additionally, the impact of the activities
on the participants’ motivation was limited to the data obtained from the interviews
with the participants. In the future studies, the motivation variable can be inves-
tigated in more detail and supported with quantitative data using the scales. Fur-
thermore, it was observed that AI assisted AR activities as practical and effective
activities could improve the PSTs’ cognitive knowledge, technological knowledge
and emotions. Hence, it is suggested to expand the usage area of these activities
such as in schools, inservice training programs and teacher education programs.
Moreover, in the study, it was observed that ChatGPT enhanced to learn, practice
and understand Phyton language so ChatGPT can be used as a beneficial tool in the
courses of algorithms and programming

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Appendix 1

Platonic Solid Test

1. Aşağıda belirtilen çok yüzlüleri çiziniz. [Draw the following objects]

Küp (Cube) Beşgen prizma (Pentagonal Prism)

Üçgen Piramit (Triangular Pyramid) Kare piramit (Square Pyramid)
Düzgün Altı yüzlü (Hexahedron) Düzgün Sekizyüzlü (Octahedron)
Düzgün Dört yüzlü (Tetrahedron) Düzgün Yirmi yüzlü (Icosehedron)
Düzgün On iki yüzlü (Dodecahedron)

2. Aşağıdaki çok yüzlüleri tanımlayınız. [Define the following objects]

Küp (Cube)
Beşgen Prizma (Pentagonal Prism)
Üçgen Piramit (Triangular Pyramid)
Kare Piramit (Square Pyramid)
Düzgün Altı Yüzlü (Hexahedron)
Düzgün Sekiz Yüzlü (Octahedron)
Düzgün Dört Yüzlü (Tetrahedron)
Düzgün On iki Yüzlü (Dodecahedron)
Düzgün Yirmi Yüzlü (Icosehedron)

3. A. Her bir köşesinde üç eşkenar üçgen bulunan çok yüzlü ……..

B. Her bir köşesinde dört eşkenar üçgen bulunan çok yüzlü ……..
C. Her bir köşesinde beş eşkenar üçgen bulunan çok yüzlü ……..
D. Her bir köşesinde üç kare bulunan çok yüzlü ……..
E. Her bir köşesinde üç beşgen bulunan çok yüzlü ……..
F. Her bir köşesinde iki dikdörtgen ve bir kare olan çok yüzlü……
G. Her bir köşesinde iki dikdörtgen ve bir üçgen bulunan çok yüzlü……
[A. Polyhedral with three equilateral triangles at each corner ……..
B. Polyhedral with four equilateral triangles at each corner. ……..
C. Polyhedral with five equilateral triangles at each corner. ……..
D. Polyhedral with three squares at each corner ……..
E. Polyhedral with three pentagons at each corner ……..
F. A polyhedron with two rectangles and a square at each corner……
G. Polyhedral with two rectangles and a triangle at each corner……]

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

4. Aşağıdaki tablodaki yazılan çok yüzlülerin belirtilen özelliklerini yazınız. [Write the specified proper-
ties of the polyhedra in the table below.]

Cismin Adı (The Object) Cismin Yüz Sayısı Cismin Köşe Sayısı Cismin Ayrıt Sayısı
(Y) (Number of (K) (Number of (A) (Number of
faces F) vertices V) edges E)
Küp (Cube)
Beşgen Prizma (Pentagonal Prism)
Üçgen Piramit (Triangular Pyramid)
Kare Piramit (Square Pyramid)
Düzgün Altı Yüzlü (Hexahedron)
Düzgün Sekiz Yüzlü (Octahedron)
Düzgün Dört Yüzlü (Tetrahedron)
Düzgün On iki Yüzlü (Dodecahe-
dron)
Düzgün Yirmi Yüzlü (Icosehedron)
5. Yukarıdaki tablo doldurulduğunda her bir cisim için belirtilen Y, K ve A sayıları arasında bir ilişki var
mıdır? (Eğer varsa bu ilişkiyi açıklayınız.) Eğer mümkünse Y, K ve A arasındaki ilişkiyi ispatlayınız.
[When the table above is filled in, is there a relationship between the numbers F, V and E specified for
each object? (Explain this relationship, if any.) If possible, prove this relationship between F, V, E.]

Appendix 2

Pre-instruction Interview Form

1. What are polyhedra? Explain.
2. Which types of artificial intelligence do you know?
3. Have you ever used artificial intelligence? For what purpose did you use it?
4. What is augmented reality? What does it do?
5. What do you know about augmented reality? Can you produce geometric
objects through augmented reality?
6. What is your knowledge about programming language?

Appendix 3

Post-instruction Interview Form

1. What topics did you cover by AI assisted AR activities?
2. Which technological tools did you use in AI assisted AR activities and for
what purposes?
3. What were your tasks in AI assisted AR activities?
4. What was the purpose of the tasks and activities?
5. How did AI assisted AR activities contribute to you?
6. What do you think about the process of learning polyhedra and the tools you
used?

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

7. How would you describe your work in the tasks in AI assisted AR activi-
ties? What were your roles and actions?
8. What distracted you through the tasks?
9. What difficulties or problems did you encounter in AI assisted AR activities?
10. Could you explain in detail how and for what purpose you use artificial
intelligence in AI assisted AR activities?
11. Now, can you produce polyhedra through augmented reality without using
artificial intelligence?
12. What are your positive and negative opinions about the use of artificial
intelligence in the process of generating polyhedra through augmented reality?

Appendix 4

Sample for the explanations in ChatGPT performed by ­PST5

https://​chat.​openai.​com/​share/​783df​a37-​99a8-​45ad-​b6eb-​b9c6e​a2ac6​2d.

Funding There is no funding for the research.

Data availability All data collected and analyzed through the present study are available from the cor-
responding author.

Declarations
Competing interests There is not any potential competing interest by the authors.

References
Abraham, M., & Annunziata, M. (2017). Augmented reality is already improving worker perfor-
mance. Harvard Business Review, 13(1–5):1.
Adeshola, I., & Adepoju, A. P. (2023). The opportunities and challenges of ChatGPT in education.
Interactive Learning Environments. Advance online publication. https://​doi.​org/​10.​1080/​10494​
820.​2023.​22538​58
Aithal, S., & Aithal, S. (2023). Effects of AI-based Chatgpt on higher education libraries. International
Journal of Management Technology and Social Sciences, 8(2), 95–108. https://​doi.​org/​10.​47992/​
IJMTS.​2581.​6012.​0272
Aktay, S., Gök, S., & Uzunoğlu, D. (2023). ChatGPT in education. Türk Akademik Yayınlar Dergisi (TAY
Journal), 7(2), 378–406. https://​doi.​org/​10.​29329/​tayjo​urnal.​2023.​543.​03
Alcañiz, M., Contero, M., Pérez-López, D. C., & Ortega, M. (2010). Augmented reality technology for
education. In S. Soomro (Ed.), New achievements in technology education and development (pp.
247–256). InTech.
Appendices 1Platonic solid test 1.Aşağıda Belirtilen çok yüzlüleri çiziniz. [Draw the following objects].
Azuma, R., Baillot, Y., Behringer, R., Feiner, S., Julier, S., & MacIntyre, B. (2001). Recent advances
in augmented reality. IEEE Computer Graphics and Applications, 21(6), 34–47. https://​doi.​org/​10.​
1109/​38.​963459
Blender Foundation (2023). Blender (Version 4.0) [Computer software]. https://​www.​blend​er.​org/

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Cerqueira, C., & Kirner, C. (2012). Developing educational applications with a non-programming
augmented reality authoring tool. Proceedings of World Conference on Educational Multimedia,
Hypermedia, and Telecommunications (pp. 2816–2825).
Chan, K. S., & Zary, N. (2019). Applications and challenges of implementing artificial intelligence in
medical education: Integrative review. JMIR Medical Education, 5(1), e13930. https://​doi.​org/​10.​
2196/​13930
Chen, X., Xie, H., Zou, D., & Hwang, G. J. (2020). Application and theory gaps during therise of artificial
intelligence in education. Computers & Education: Artificial Intelligence, 1, 100002. https://​doi.​org/​10.​
1016/j.​caeai.​2020.​100002
Chiu, T. K. F. (2023). The impact of Generative AI (GenAI) on practices, policies and research direction
in education: A case of ChatGPT and Midjourney. Interactive Learning Environments, 1–17. https://​
doi.​org/​10.​1080/​10494​820.​2023.​22538​61
Claros-Perdomo, D. C., Millan-Rojas, E. E., & Gallego-Torres, A. P. (2022). Use of augmented reality,
gamifcation and m-learning. Metaverse, 3(2), 8.
Coimbra, T., Cardoso, T., & Mateus, A. (2015). Augmented reality: An enhancer for higher education
students in Math’s learning? Procedia Computer Science, 67, 332–339.
del Cerro Velázquez, F., & Morales Méndez, G. (2021). Application in augmented reality for learning
mathematical functions: A study for the development of spatial intelligence in secondary education
students. Mathematics, 9(4), 369. https://​doi.​org/​10.​3390/​math9​040369
Devagiri, J. S., Paheding, S., Niyaz, Q., Yang, X., & Smith, S. (2022). Augmented reality and artifi-
cial intelligence in industry: Trends, tools, and future challenges. Expert System Applications, 207,
118002. https://​doi.​org/​10.​1016/j.​eswa.​2022.​118002
Dikovic, L. (2009). Applications GeoGebra into teaching some topics of mathematics at the college level.
Computer Science and Information Systems, 6(2), 191–203. https://​doi.​org/​10.​2298/​CSIS0​90219​1D
Frieder, S., Pinchetti, L., Griffiths, R. R., Salvatori, T., Lukasiewicz, T., Petersen, P., & Berner, J. (2024).
Mathematical capabilities of chatgpt (p. 36). Advances in neural information processing systems.
Fuchsova, M., & Korenova, L. (2019). Visualisation in basic science and engineering education of future
primary school teachers in human biology education using augmented reality. European Journal of
Contemporary Education, 2019, 8(1), 92–102. https://​doi.​org/​10.​13187/​ejced.​2019.1.​92
Gómez-Rios, M. D., Paredes-Velasco, M., Hernández-Beleño, R. D., & Fuentes-Pinargote, J. A. (2023).
Analysis of emotions in the use of augmented reality technologies in Education: A systematic
review. Comput Appl Eng Educ, 31, 216–234.
Haniff, D. J., & Baber, C. (2003). User evaluation of augmented reality systems. In Proceedings of the
Seventh International Conference on Information Visualization, 505–511. https://​doi.​org/​10.​1109/​
IV.​2003.​12180​32
Hsieh, H. F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative
Health Research, 15(9):1277–1288. https://​doi.​org/​10.​1177/​10497​32305​276687
Hwang, G. J., Xie, H., Wah, B. W., & Gaševi´c, D. (2020). Vision, challenges, roles and research issues
of Artificial Intelligence in Education. Comput Educ Artif Intell, 1, 100001.
İbili, E., & Şahin, S. (2016). The effect of augmented reality assisted geometry instruction on students’
achivement and attitudes. Teaching Mathematics and Computer Science, 13(2), 177–193. https://​
doi.​org/​10.​5485/​tmcs.​2015.​0392
İbili, E., Çat, M., Resnyansky, D., Şahin, S., & Billinghurst, M. (2020). An assessment of geometry
teaching supported with augmented reality teaching materials to enhance students’ 3D geometry
thinking skills. International Journal of Mathematical Education in Science and Technology, 51(2),
224–246. https://​doi.​org/​10.​1080/​00207​39X.​2019.​15833​82
Javaid, M., Haleem, A., Singh, R. P., Khan, S., & Khan, I. H. (2023). Unlocking the opportunities through
ChatGPT tool towards ameliorating the education system. BenchCouncil Transactions on Bench-
marks Standards and Evaluations, 3(2), 100115. https://​doi.​org/​10.​1016/j.​tbench.​2023.​100115
Kamite, T., Ito, T., Matsumoto, A., Munakata, T., & Inoue, T. (2019). A chatbot system for mental health-
care based on SAT counseling method. Mobile Information Systems, 2019.
Kaplan, A., Haenlein, M., & Siri (2019). Siri in my hand, who’s the fairest in the land? On the interpreta-
tions, illustrations, and implications of artificial intelligence. Business Horizons, 62(1), 15–25.
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., et al. (2023). ChatGPT for good? On opportunities
and challenges of large language models for education. Learning and Individual Differences, 103,
102274. https://​doi.​org/​10.​1016/j.​lindif.​2023.​102274
Kaufmann, H. (2004). Geometry education with augmented reality (Unpublished doctoral dissertation).
Vienna: Vienna University of Technology.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Kaufmann, H., & Dünser, A. (2007). Summary of usability evaluations of an educational augmented
reality application. In Proceedings of the Second International Conference on Virtual Reality (pp.
660–669). Beijing, China. https://​doi.​org/​10.​1007/​978-3-​540-​73335-5_​71
Kerawalla, L., Luckin, R., Seljeflot, S., et al. (2006). Making it real: Exploring the potential of augmented
reality for teaching primary school science. VirtualReal, 10(3–4), 163–174.
Kolchenko, V. (2018). Can modern AI replace teachers? Not so fast! Artificial intelligence and adaptive
learning: Personalized education in the AI age. HAPS Educator, 22(3), 249–252. https://​doi.​org/​10.​
21692/​haps.​2018.​032
Köse, N. (2008). İlköğretim 5. sınıf öğrencilerinin dinamik geometri yazılımı Cabri geometriyle simetriyi
anlamlandırmalarının belirlenmesi: Bir eylem araştırması [Determining fifth grade primary school
students? understanding of symmetry using dynamic geometry software Cabri geometry: An action
research]. Unpublished doctoral dissertation, Anadolu University, Turkey.
Lindgren, R., & Moshell, J. M. (2011). Supporting children’s learning with body-based metaphors in a
mixed reality environment. Proceedings of the 10th international conference on interaction design
and children. ACM, pp 177–180.
Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt
engineering. The Journal of Academic Librarianship, 49(4), 102720. https://​doi.​org/​10.​1016/j.​
acalib.​2023.​102720
Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed. An argument for
AI in education. Pearson Education.
Ma, H. L., Wu, D., Chen, J., & Hsieh, K. (2009). Mitchelmore’s development stages of the right rectan-
gular prisms of elementary school students in Taiwan. Proceedings of the 33rd Conference of the
International Group for the Psychology of Mathematics Education.
Mansi, K. E. (2003). Reasoning and geometric proof in mathematics education: A review of the litera-
ture. Degree of Master of Science (USA: North Carolina State University).
McGlone, C. W. (2009). A case study of preservice teachers’ experiences in a reform geometry course.
Doctoral Thesis, University of North Carolina, United States of America.
Milgram, P., Drascic, D., Grodski, J. J., Restogi, A., Zhai, S., & Zhou, C. (1995). Merging Real and Vir-
tual Worlds. Proceedings of IMAGINA ‘95 (Monte Carlo), 218–230.
Mitchell, R. (2011). Alien contact!: Exploring teacher implementation of an augmented reality curricular
unit. Journal of Computers in Mathematics and Science Teaching, 30(3):271–302.
MyWebAR (2023). MyWebAR [Augmented reality software]. https://​www.​myweb​ar.​com/
OpenAI (2023). ChatGPT (GPT-3.5). https://​chatg​pt.​com
Oranga, J. (2023). Benefits of artificial intelligence (chatgpt) in education and learning: Is ChatGPT help-
ful? International Review of Practical Innovation Technology and Green Energy (IRPITAGE), 3(3),
46–50. https://​doi.​org/​10.​54443/​irpit​age.​v3i3.​1250
Ozdemir, F., Aslaner, R., & Acikgul, K. (2020). The effect of the computer-aided mathematics teaching
on the students’ attitudes towards mathematics: A meta-analysis study. Inonu University Journal of
the Graduate School of Education, 7(13), 18–40. https://​doi.​org/​10.​29129/​inujg​se
Pittalis, M., & Christou, C. (2010). Types of reasoning in 3D geometry thinking and their relation
with spatial ability. Educational Studies in Mathematics, 75(2), 191–212. https://​doi.​org/​10.​1007/​
s10649-​010-​9251-8
Rasul, T., Nair, S., Kalendra, D., Robin, M., de Oliveira Santini, F., Ladeira, W. J., Sun, M., Day, I.,
Rather, R. A., & Heathcote, L. (2023). The role of ChatGPT in higher education: Benefits, chal-
lenges, and future research directions. Journal of Applied Learning and Teaching, 6(1), 41–56.
https://​doi.​org/​10.​37074/​jalt.​2023.6.​1.​29
Richey, R. C., & Klein, J. D. (2014). Design and development research. Handbook of research on educa-
tional communications and technology (pp. 141–150). Springer.
Rospigliosi, P., & ‘asher’ (2023). Artificial intelligence in teaching and learning: What questions should
we ask of ChatGPT? Interactive Learning Environments, 31(1), 1–3. https://​doi.​org/​10.​1080/​10494​
820.​2023.​21801​91
Rudolph, J., Tan, S., & Tan, S. (2023). ChatGPT: Bullshit Spewer or the end of traditional assessments
in higher education? Journal of Applied Learning and Teaching, 6(1), 342–363. https://​doi.​org/​10.​
37074/​jalt.​2023.6.​1.9
Sahu, C. K., Young, C., & Rai, R. (2021). Artificial intelligence (AI) in augmented reality (AR)-assisted
manufacturing applications: A review. International Journal of Production Research, 59(16), 4903–
4959. https://​doi.​org/​10.​1080/​00207​543.​2020.​18596​36

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Salehi, H., & Salehi, Z. (2012). Integration of ICT in language teaching: Challenges and barriers. In
Proceedings of the 3rd International conference on e-Education, e-Business, e-Management and
e-Learning, (Vol. 27, pp. 215–219).
Schreyer-Bennethum, L., & Albright, L. (2011). Evaluating the incorporation of technology and applica-
tion projects in the higher education mathematics classroom. International Journal of Mathematical
Education in Science and Technology, 42(1):53–63.
Shelton, B. E., & Hedley, N. R. (2004). Exploring a cognitive basis for learning spatial relationships with
augmented reality. Technology Instruction Cognition and Learning, 1(4), 323.
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. https://​doi.​org/​10.​1080/​10494​820.​2023.​
22098​81. Advance online publication.
Suydam, M. N. (1985). The shape of instruction in geometry: Some highlights from research. Mathemat-
ics Teacher, 78, 481–486.
Tlili, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023).
What if the devil is my guardian angel: ChatGPT as a case study of using Chatbots in education.
Smart Learning Environments, 10(1), 15. https://​doi.​org/​10.​1186/​s40561-​023-​00237-x
Turing, A. M. (1950). Computing machinery and intelligence. Mind, 59(236), 433–460. https://​doi.​org/​
10.​1093/​mind/​LIX.​236.​433
Urhan, S., Gençaslan, O., & Dost, Ş. (2024). An argumentation experience regarding concepts of calculus
with ChatGPT. Interactive Learning Environments. https://​doi.​org/​10.​1080/​10494​820.​2024.​23080​
93. Advance online publication.
Veneziano, L. (1997). A method for quantifying content validity of health-related questionnaires. Ameri-
can Journal of Health Behavior, 21(1), 67–70
Vogt, W. P., Gardner, D. C., & Haefele, L. M. (2012). When to use what research design. Guilford Press.
Wang, X., & Dunston, P. S. (2006). Compatibility issues in augmented reality systems for AEC: An
experimental prototype study. Automation in Construction, 15(3), 314–326.
Wardat, Y., Tashtoush, M. A., AlAli, R., & Jarrah, A. M. (2023). ChatGPT: A revolutionary tool for
teaching and learning mathematics. Eurasia Journal of Mathematics Science and Technology Edu-
cation, 19(7), em2286. https://​doi.​org/​10.​29333/​ejmste/​13272
Xie, H., Zou, D., Wang, F. L., Wong, T. L., Rao, Y., & Wang, S. H. (2017). Discover learning path for
group users: A profile-based approach. Neurocomputing, 254, 59–70.
Yao, W., Wang, L., & Liu, D. (2024). Augmented reality-based language and math learning applications
for preschool children education. Universal Access in the Information Society. https://​doi.​org/​10.​
1007/​s10209-​024-​01101-6
Yin, R. K. (1994). Discovering the future of the case study. Method in evaluation research. Evaluation
Practice, 15(3):283–290
Yufeia, L., Salehb, S., Jiahuic, H., & Abdullah, S. M. S. (2020). Review of the application of artificial
intelligence in education. International Journal of Innovation Creativity and Change, 12(8), 548–
562. https://​doi.​org/​10.​53333/​IJICC​2013/​12850
Zhang, R., Zou, D., & Cheng, G. (2023). Chatbot-based learning of logical fallacies in EFL writing:
Perceived effectiveness in improving target knowledge and learner motivation. Interactive Learning
Environments, 1–18. https://​doi.​org/​10.​1080/​10494​820.​2023.​22203​74

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under
a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of such publishing agreement and
applicable law.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 Education and Information Technologies

Authors and Affiliations

Tuğba Uygun1 · Ali Şendur1 · Beyza Top2 · Kadriye Coşgun‑Başeğmez3

* Tuğba Uygun
tugbauygun42@gmail.com
Ali Şendur
ali.sendur@alanya.edu.tr
1
Department of Mathematics and Science Education, Faculty of Education, Alanya Alaaddin
Keykubat University, Alanya/Antalya, Türkiye
2
Gradute Student, Department of Mathematics and Science Education, Postgraduate Education
Institute, Alanya Alaaddin Keykubat University, Alanya/Antalya, Türkiye
3
Mathematics Teacher, Ministry of National Education, Alanya/Antalya, Türkiye

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center
GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers
and authorised users (“Users”), for small-scale personal, non-commercial use provided that all
copyright, trade and service marks and other proprietary notices are maintained. By accessing,
sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of
use (“Terms”). For these purposes, Springer Nature considers academic use (by researchers and
students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and
conditions, a relevant site licence or a personal subscription. These Terms will prevail over any
conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription (to
the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of
the Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may
also use these personal data internally within ResearchGate and Springer Nature and as agreed share
it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not otherwise
disclose your personal data outside the ResearchGate or the Springer Nature group of companies
unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial
use, it is important to note that Users may not:

1. use such content for the purpose of providing other users with access on a regular or large scale
basis or as a means to circumvent access control;
2. use such content where to do so would be considered a criminal or statutory offence in any
jurisdiction, or gives rise to civil liability, or is otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association
unless explicitly agreed to by Springer Nature in writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a
systematic database of Springer Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a
product or service that creates revenue, royalties, rent or income from our content or its inclusion as
part of a paid for service or for other commercial gain. Springer Nature journal content cannot be
used for inter-library loans and librarians may not upload Springer Nature journal content on a large
scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not
obligated to publish any information or content on this website and may remove it or features or
functionality at our sole discretion, at any time with or without notice. Springer Nature may revoke
this licence to you at any time and remove access to any copies of the Springer Nature journal content
which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or
guarantees to Users, either express or implied with respect to the Springer nature journal content and
all parties disclaim and waive any implied warranties or warranties imposed by law, including
merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published
by Springer Nature that may be licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a
regular basis or in any other manner not expressly permitted by these Terms, please contact Springer
Nature at

onlineservice@springernature.com