A Proposed TPACK Model of Teaching STEM with AI Components:

Evaluating a Teacher Development Course for Fostering Digital

Creativity

Siu-Cheung Kong

1,2 a

, Yin Yang

and Wing Kei Yeung

Artificial Intelligence and Digital Competency Education Centre, The Education University of Hong Kong, HKSAR, China

Department of Mathematics and Information Technology, The Education University of Hong Kong, HKSAR, China

Keywords: Artificial Intelligence, Digital Creativity, Internet of Things, Problem-Solving, STEM, TPACK.

Abstract: In this study, a novel design of STEM activities was proposed with a focus on concepts of Internet of Things

(IoT) and artificial intelligence (AI) components for developing the problem-solving abilities and digital

creativity of in-service primary teachers. This study evaluated the effectiveness of a 6-hour course for

developing primary school teachers’ competency in teaching AI-integrated STEM activities and developing

their digital creativity. One hundred and ninety-one teachers from 108 primary schools attended the

development course and completed survey and creativity evaluation. Teachers’ responses to surveys on the

TPACK model and digital creativity evaluation sheets were collected. The paired t-test results indicated

statistically significant improvement on all 17 TPACK items with a large effect size (Cohen’s d = 1.213). For

the digital creativity evaluation completed, 81.15% of the teachers demonstrated digital creativity and

expressed their ideas in designing IoT systems and many of the designs included AI components. The paper

concluded the implications of this study and future work were discussed.

1 INTRODUCTION

Artificial intelligence (AI) technology has become

more prevalent in daily life in the digital era (Kong et

al., 2021). To ensure that teaching and learning

content is meaningful and appropriate, we suggest

integrating the essential AI experiences, such as the

interaction with and training of AI models, into

STEM activities. Integrating AI models in STEM

systems makes them richer and more human-friendly

(Ouyang et al., 2023). Another objective of

promoting STEM education with AI components is to

strengthen students’ problem-solving abilities and

inspire their digital creativity. Incorporating AI

components in STEM activities provides

opportunities for students to cooperate with peers in

the problem-solving processes when designing

systems with appropriate AI algorithm and training

models (Lin et al., 2021). Furthermore, these

integrated activities provide opportunities for

https://orcid.org/0000-0002-8691-3016

https://orcid.org/0000-0002-9966-248X

https://orcid.org/0009-0009-3845-7448

students to be inspired with digital creativity because

they can transfer what they have learnt to new

scenarios.

We select the internet of things (IoT) as the source

of the key concepts in the design of our STEM

activities because IoT is assumed to be permeate

every part of daily life and become omnipresent in the

physical world in the digital era (Sedrati et al., 2022).

In this study, we try to develop teachers’

understanding of the latest development through

building physical artefacts using IoT and AI concepts

in STEM activities of our professional development

course, ultimately enabling them to promote AI and

STEM Education in their schools. Although there are

researchers working with the development of

students’ AI Literacy (Touretzky et al., 2019), there

is a lack of research on integrating AI and IoT

concepts with STEM activities. We propose a novel

approach to include AI and IoT concepts in STEM

activities of primary schools. However, as most of the

in-service teachers have limited STEM-related

Kong, S., Yang, Y. and Yeung, W.

A Proposed TPACK Model of Teaching STEM with AI Components: Evaluating a Teacher Development Course for Fostering Digital Creativity.

DOI: 10.5220/0012511900003693

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

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 2, pages 163-170

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

163

technological knowledge and lack confidence in

teaching STEM activities in this context, there is a

need for teacher development (Cavlazoglu & Stuessy,

2017; Schreiter et al., 2023). The specific research

questions addressed were as follows:

(1) What effects does the teacher development

course have on teachers’ competence in teaching

STEM with AI components?

(2) How do teachers who have participated in the

teacher development course developed their digital

creativity?

2 LITERATURE REVIEW

2.1 Internet of Things Concepts,

Problem-Solving Abilities, and

Digital Creativity

The major focus of our STEM activities is to promote

IoT concepts, problem-solving abilities, and digital

creativity. Students are guided to create IoT systems

using microprocessor and block-based programming

on AI-enabled platforms connected with the

microprocessor. Playing with and building a

functional system is helpful for students to understand

how an IoT system works. When Ashton (2009) first

introduced IoT, he proposed that computers can

observe and understand the world using sensor

technology without any help from humans. The

functionality of IoT is delivered by six elements:

object identification, sensing, communication,

computation, services, and semantics (Al-Fuqaha et

al., 2015). Object identification and semantics are

important elements of IoT. It would be too

complicated for primary students if we include all six

elements. We adopt a simplified IoT concepts of

sensing, reasoning, and reacting, which are the

essential elements of automation systems, for

designing our STEM activities in the primary school

context. Sensing refers to the use of sensors or a

microprocessor with sensors to detect and transfer

data. Reasoning refers to the use of a microprocessor

and programming codes that process the data with

computation and determine the reactions of a system.

Reacting refers to the final reaction of a system to

provide services after communication among the

devices involved. The implementation of sensing,

reasoning, and reacting in IoT systems can support

automation and interaction with humans based on

programming codes and communication between

devices.

STEM education aims to develop students’

problem-solving abilities. Sullivan and Heffernan

(2016) proposed that STEM activities work to

develop four aspects of problem-solving abilities:

casual reasoning, sequencing, conditional reasoning,

and engineering systems thinking. Kong (2023)

proposed a pedagogical design which focuses on

developing students’ problem-solving skills in four

aspects within cross-disciplinary STEM activities.

Causal reasoning refers to the identification of

casualty, which is the relationship between the causes

and effects of an incident (Van Vo & Csapó, 2023).

Students can exercise this skill by investigating the

cause of system failure and fixing bugs. Sequencing

is defined as “the ability to put items in a specific

order” (Sullivan & Heffernan, 2016, p. 8). As

students learn basic programming knowledge during

the activities, they develop sequencing skills by

arranging the programming blocks into a specific

order with the aim of automating the system.

Conditional reasoning is the process involved with

statements of the form “if A then B,” in which A is

the antecedent and B is the consequent (Nickerson,

2015). Conditional reasoning also refers to logical

reasoning, which is important in the process of

building an automated system (Sullivan & Heffernan,

2016). Engineering systems thinking is “the ability to

understand the whole system or perceive how the

components (i.e., person, part) function as part of a

system” (Frank, 2002, p. 1351). In STEM activities,

it is crucial for students to understand the function of

each component and how they interact in the whole

system.

Despite the fact that recent studies have

highlighted the significance of digital creativity and

the necessity for students to evolve from mere

consumers to innovative problem-solvers in the

digital age (Kong & Lai, 2021), there is a noticeable

gap in understanding the extent to which students are

able to transfer the skills and knowledge acquired

from STEM-based educational activities to practical,

real-world scenarios. This research would provide

valuable insights into educational strategies that can

better prepare students to contribute innovatively and

adaptively to the rapidly evolving society.

2.2 Integrating STEM with AI

Components

As technology advances rapidly, new concepts can be

added to STEM activities to motivate students in

learning actively. Promoting AI literacy is one of the

new learning objectives of our proposed STEM

activities. AI literacy can be defined as the

CSEDU 2024 - 16th International Conference on Computer Supported Education

164

“understanding of AI concepts and competencies in

using AI concepts for evaluation and using AI

concepts for understanding the real world” (Kong &

Zhang, 2021, p. 12). To help students to develop AI

literacy, we introduce the five big ideas in AI:

perception, representation and reasoning, learning,

natural interaction, and societal impact (Touretzky et

al., 2019). These five big ideas are used to develop

students’ basic understanding of AI. Among the five

big ideas, we emphasise giving primary students

experience in interacting with AI and, through hands-

on activities, develop the concept of how machines

“learn”, which is a key concept of AI at present.

Students are guided to figure out how computers learn

from data as they take part in the data training

process. We also inspire primary students to apply

what they have learnt from the AI examples by

figuring out potential uses in real-life situations. This

can help to develop their digital creativity and give

them a better understanding of the societal impact of

AI.

AI can help promote IoT concepts and provide

flexibility in building STEM systems. One of the

learning objectives of our proposed STEM activities

is to promote IoT concepts. Students usually use

sensors, microprocessors, and actuators to automate

systems in STEM activities. AI can help to promote

IoT concepts and facilitate more interactions between

humans and the system (Ghosh et al., 2018). AI and

IoT can be integrated to develop intelligent

applications that can benefit users (Katare et al.,

2018). Such an integration of AI and IoT is known as

the artificial intelligence of things (AIoT), which

makes applications “smarter” by giving them the

ability to collect and process data (Qiu et al., 2023).

The machine learning process allows the system to

learn complicated human behaviours and react

accordingly.

2.3 Pedagogy of “to Play, to Inquire, to

Assemble, to Code, to Reflect, to

Create”

Based on Kong and Lai’s (2021) pedagogy of “to

play, to think, to code, to reflect” for teaching

computational thinking through programming to

enable students understand how the programs work,

plan before coding, learn problem-solving skills in

the coding process and finally students are guided to

reflect on what they have learnt and think about the

possible use of what they have learnt in other

occasions. This pedagogy aims at developing

problem-solving skills and digital creativity of

students.

In this study, we proposed a pedagogy of “to

play,” “to inquire,” “to assemble,” “to code,” “to

reflect,” and “to create”. This pedagogy provides

students with an introduction to playing around with

a STEM system. The objective of the playing

activities is to arouse their interest in STEM, which is

important in the primary school context. This can

trigger students’ interest in inquiring how the STEM

system works. Students will then disassemble the

STEM system and understand how each

component—digital and non-digital—works and then

reassemble the components into a working system,

following some coding and data training activities if

AI models are involved. Students will be asked to

reflect on what they have learnt in building the STEM

system and to propose and discuss other possible

STEM systems using similar technologies to those in

the example system. Pedagogy of “to play” and “to

inquire” at the beginning of each STEM activity lays

a good foundation for students to proceed with the

STEM activities of “to assemble,” and “to code”

which involve building STEM artefacts, thus helping

students to develop their problem-solving abilities.

2.4 TPACK Framework in Guiding

and Evaluating Teacher

Development

The Hong Kong Education Bureau (2015) has

highlighted the importance of providing teacher

development in promoting STEM education. Indeed,

there is a genuine need for teacher development

courses to support teachers to teach STEM to develop

students’ problem-solving abilities and digital

creativity. In this study, we use the TPACK

framework to guide the design of a teacher

development course in teaching STEM activities with

AI components (Mishra & Koehler, 2006). TPACK

framework includes seven components, including

three major domains of content knowledge (CK),

pedagogical knowledge (PK) and technological

knowledge (TK), and their interactions which forms

technological content knowledge (TCK),

technological pedagogical knowledge (TPK),

pedagogical content knowledge (PCK) and

technological pedagogical content knowledge

(TPACK).

In our context, CK refers to the IoT and AI

concepts involved in STEM activities and the

problem-solving skills involved in building the

STEM system; TK refers to the general knowledge in

using technology including computer, coding

platforms, microprocessors and different electronic

components; PK refers to the general pedagogical

A Proposed TPACK Model of Teaching STEM with AI Components: Evaluating a Teacher Development Course for Fostering Digital

Creativity

165

knowledge in teaching STEM activities, such as to

guide students to use discussion and/or group

activities to accomplish the problem-solving process

and to generate ideas for digital creativity. More

importantly, teachers need to guide students to learn

from “to play” and “to inquire” so that students know

why and how to complete the “to assemble” and “to

code” process and finally learn from reflection on

problem-solving and become more creative. TCK

refers to the understanding of the technological

functions of each digital and non-digital component

used for building the STEM system; TPK refers to the

use of technology and pedagogy in general for

teaching outside our scope of STEM; PCK involves

the pedagogical design for achieving the learning

objectives of our STEM activities; TPACK refers to

the synthetisation of knowledge in our learning

context of teaching STEM with AI components.

3 METHODOLOGY

3.1 Participants and Procedure

Two hundred and one teachers from 108 primary

schools attended the 6-hour teacher development

course. The course introduces our method of teaching

STEM with AI components in primary schools. Data

included pre- and post-surveys on TPACK and digital

creativity designs. A total of 192 responses were

collected from pre- and post-surveys. A total of 191

teachers expressed their digital creativity ideas in

writing and sketches after the course.

3.2 Instruments

Surveys on TPACK: The instrument consists of 17

items. A 5-point Likert scale from 1 (strongly

disagree) to 5 (strongly agree) was used to score each

item. Cronbach’s alpha for the pre- and post-course

survey was above 0.85. Sample items are as shown

below: TCK—I understand the functions that sensor,

microprocessor and actuator perform in the IoT

systems; TPACK—I can teach STEM lessons that

appropriately combine the content of STEM,

technological innovation, and proper teaching

approaches.

Digital creativity evaluation sheets: After the

course, we also asked the primary school teachers to

suggest new STEM applications other than those they

had learnt in the course. These design suggestions

were used to have a brief understanding on the digital

creativity development of these teachers. Criteria of

the creative ideas are listed in Table 1.

Table 1: Marking criteria of creative ideas.

Criteria Mark

Ideas that were identical to the applications

discussed in the course

Description of the application could be clearer 1

New application designs that used the IoT

concepts of sensing-reasoning-reacting

3.3 The Teacher Development Course

The course was made up of three teaching units. The

first unit was about the design of a maze game, in

which a character on a monitor is controlled by a

hand-made physical joystick connected to an internal

microprocessor. In playing with the system, the

teachers gain an initial understanding of the core IoT

concepts of sensing-reasoning-reacting. The second

unit is a table tennis game that includes IoT concepts

and AI concepts. To complete the system, teachers

are required to train AI models with machine learning

by striking different poses in front of webcams.

Trained AI models are then extracted and installed in

the game system, which reacts to poses and motion

sensed by an accelerometer in the hand-made

physical racket. The third unit is a Chinese face-

changing performance game, which again involves

IoT concepts and AI concepts. Teachers can perform

a face-changing show using the system. Teachers are

required to train AI models for recognising facial

expressions. The models are then extracted for use in

the performance game application. By developing

these AI models and engaging in further discussion,

teachers develop basic AI concepts and learn how to

use them in building STEM systems with AI

components.

In the introduction to each unit in the development

course, the teachers were briefed on the technologies

involved in the unit and the content knowledge to be

involved (TK, CK, TCK). They were then

encouraged to go through the teaching process of “to

play” “to inquire,” “to assemble,” “to code,” “to

reflect,” and “to create” in each teaching unit to give

them experience of the teaching process and the

methods for developing problem-solving skills and

inducing digital creativity (PK, PCK, TPK and

TPACK).

Taking Teaching the Smart Ping Pong Course

Unit as an example, the teachers were briefly

introduced to the technology components involved

with the application before they began to play with it.

They then experienced the pedagogical processes

with the STEM systems, from “to play” through to “to

create”. Towards the end of this teaching session, the

teachers were guided to reflect on the IoT and AI

CSEDU 2024 - 16th International Conference on Computer Supported Education

166

concepts they had learnt and the pedagogical

practices for teaching the unit. The teachers were

finally asked to share ideas with their learning peers

for making new systems by applying what they had

learnt in the teaching unit, including the technology

components and the experience of training and using

AI models in the STEM system. This was an

important session to develop digital creativity in the

teacher participants and to give them knowledge on

how to inspire digital creativity in their students when

they return to their school to teach. Smart Ping Pong,

a table tennis gaming system, is used here to illustrate

the pedagogy. An AI model is trained, extracted, and

stored in the Smart Ping Pong application. Webcams

connected to the computer continuously capture

images of players and send them to the computers.

The computers use the trained AI models to reason

and categorise these images into “left” and “right”

strokes in the table tennis game and then react to the

strokes of the players in the computer monitor. Figure

1 shows an illustration of delivering the IoT concepts

of sensing-reasoning-reacting with AI components in

the Smart Ping Pong system.

Figure 1: The IoT concepts of sensing-reasoning-reacting

with AI components in the Smart Ping Pong system.

The built-in accelerometer of the microprocessor

installed in the physical racket held by the player

continuously senses the acceleration status and

transmits the status data to the computer using

Bluetooth. To interact with the system, teachers held

the physical rackets and performed “left” and “right”

strokes in front of the webcams, in reacting to the

strokes of the virtual players on computer monitors.

Teacher participants learnt about the AI concept of

confidence level, and understand that decisions made

were not always accurate.

After playing with the system, the teacher

participants were asked to move on to the “to inquire”

process. The teachers were asked to consider how the

system in Figure 2 works as a whole by sensing,

reasoning, and reacting with parts. The parts include

the physical racket containing a microprocessor with

accelerometer, battery box for the microprocessor,

webcam, computer, and computer monitor. The

teachers developed engineering systems thinking in

putting all these parts together to work as a system.

Before moving on to the “to assemble” and “to code”

elements, the teachers were introduced to the process

of teaching the machine to learn to categorise the

strokes in the table tennis game as left and right

strokes. The AI model was therefore trained to learn

the left and right strokes by collecting these images

and assigning them labels of “left” and “right”. In the

process of training these AI models, teachers learnt

about the five big ideas of AI. After exporting the

model to the Smart Ping Pong application, the

teachers coded for the responses to the two conditions

after the AI model recognises the poses of the players.

The virtual racket on the monitor moves to the right

on the monitor when the “left” pose is recognised and

to the left when the “right” pose is recognised (see

Figure 2).

Figure 2: Teachers practise conditional reasoning with

coding for movement of the virtual racket in the opposite

direction indicated by the AI recognition results.

In the “to assemble” and “to code” elements, the

teachers were asked to build the system by

assembling the physical part and coding on block-

based programming environment in driving the

system to work together. In the process of building

the system, the teachers exercised four aspects of

problem-solving skills. Figure 3 shows one of the

STEM activities in constructing the physical racket

step-by-step with a cardboard tube, battery box,

microprocessor, cardboard base and cover.

Figure 3: Teachers practise sequencing by constructing the

physical racket step-by-step with a cardboard tube, battery

box, microprocessor, cardboard base, and cardboard cover.

After a whole system is assembled with parts, it might

not function as expected. In the “to assemble”

A Proposed TPACK Model of Teaching STEM with AI Components: Evaluating a Teacher Development Course for Fostering Digital

Creativity

167

activity, the teachers exercise causal reasoning to find

out the source of errors and rectify them. Figure 4

shows some possible sources of malfunction. For

example, a malfunction could be caused by the

electricity supply failure of the microprocessor

(Figure 4a), low-quality photos causing AI pose-

recognition failure in deploying the AI model (Figure

4b), and/or losing the Bluetooth connection between

the computer and microprocessor (Figure 4c).

Figure 4: Teachers practise causal reasoning while

assembling the Smart Ping Pong system.

In the “to create” activity, the teachers were given

time to discuss and develop new ideas for solving

real-world problems by introducing other AI models

using sensors and actuators that might interest them

and putting them together with the microprocessor

and computer that they had just experienced. The

brainstorming process already helped them to apply

what they had learnt in these lessons.

In the “to reflect” session at the end of each lesson,

the teachers were asked to reflect on what they had

learnt and how to improve the pedagogy of teaching

the unit, with an emphasis on the development of

problem-solving abilities and inspiring digital

creativity. The teachers were also guided to reflect on

their experiences in building the system and to

understand that dealing with failures in the process of

building these systems is part of the learning process

in STEM activities.

4 RESULTS AND DISCUSSIONS

4.1 Teachers’ TPACK Development

A paired t-test was carried out to calculate the

significance of changes in teachers’ TPACK after

completing the teacher development course. The

results show significant improvement on all items

with a medium to large effect size. A significant

improvement was found in overall result [t(190)

=16.770, p < .001]. For Cohen’s d, a value of 0.2

indicates a small effect, 0.5 a medium effect, and 0.8

a large effect. The overall result indicated a

significant improvement with a large effect (Cohen’s

d = 1.213). A significant improvement was found in

all TPACK items, with a large effect size on CK

[t(190) = 18.539, p < .001] (Cohen’s d = 1.341), PK

[t(190) = 11.616, p < .001] (Cohen’s d = 0.841), TCK

[t(190) = 15.346, p < .001] (Cohen’s d = 1.110), TPK

[t(190) = 14.299, p < .001] (Cohen’s d = 1.035) and

TPACK [t(190) = 13.604, p < .001] (Cohen’s d =

0.984), and a medium effect size on TK [t(190) =

9.967, p < .001] (Cohen’s d = 0.721) and PCK [t(190)

= 10.398, p < .001] (Cohen’s d = 0.752). Descriptive

data is shown in Table 1. The paired t-test result

suggested that the teacher development course greatly

improved the teachers’ confidence in teaching IoT

concepts and problem-solving abilities and inspiring

digital creativity (CK) with the use of proper

technology tools (TCK). The paired t-test result also

suggested that the teacher development course has

greatly improved the teachers’ confidence in the use

of general pedagogical knowledge (PK), such as

collaborative learning in teaching STEM activities,

technological pedagogical knowledge (TPK), which

is the use of proper pedagogy in delivering

technological knowledge to students, and

technological pedagogical content knowledge

(TPACK), which is the teaching of STEM activities

in the specific context of STEM lessons. The result

also suggested that there is improvement on teacher’s

confidence in general technological knowledge (TK)

and the use of content knowledge in handling the

learning difficulties of students (PCK) after attending

the course.

Table 2: Paired t-test results of the TPACK Survey.

Pre Post

t-value

Mean SD Mean SD

CK 3.02 0.91 4.05 0.60 18.539***

TK 3.47 0.84 3.99 0.69 9.967***

PK 3.34 0.70 3.89 0.67 11.616***

PCK 3.35 0.78 3.91 0.72 10.398***

TCK 3.11 0.89 3.98 0.66 15.346***

TPK 3.07 0.91 3.94 0.71 14.299***

TPACK 3.16 0.83 3.91 0.69 13.604***

Overall 3.21 0.74 3.96 0.62 16.770***

4.2 Evaluation of the Digital Creativity

Development of the Teacher

Participants

Two members of the research team were assigned to

mark the teachers’ designs independently. The inter-

rater reliability was ICC = 0.844 (p < .001) with 95%

confidence intervals ranging from 0.797 to 0.880,

CSEDU 2024 - 16th International Conference on Computer Supported Education

168

indicating substantial agreement between the two

raters. Among the 191 teachers who completed the

surveys, 60 (31.41%) were given 2 marks, 95

(49.74%) were given 1 mark, 26 (13.61%) were given

a 0 mark for their answers, and 10 (5.24%) did not

respond to the question. To summarise, 155 (81.15%)

of the teachers showed their digital creativity in

designing IoT or AIoT systems after attending the

development course. Figure 5 shows one such design

by a teacher participant, which is for a system to help

students to learn action verbs. The teacher participant

proposed the use of a video camera to sense students’

postures. The application would pick an action verb

from its vocabulary at random and show it on the

computer monitor. The students would be required to

act out their understanding of the action verb, and a

trained pose-recognition AI model would judge

whether the action was correct. For example, if the

computer screen displayed “jump” on the monitor,

and a jumping posture was captured by the video

camera and recognised by the pose-recognition AI

model, “correct” would be shown on the computer

monitor.

Figure 5: Teacher’s design for a system on learning English

action verb vocabulary.

Overall, the design examples demonstrate that

teachers were able to design IoT systems and were

very often able to enrich their designs by integrating

AI perception models for recognising images, poses,

hands, bodies, faces, facial expressions, and sound.

5 CONCLUSION AND FUTURE

WORK

The results of this study suggest that the teacher

development course is helpful in developing teachers’

competency in teaching AI-integrated STEM

activities and developing their digital creativity. With

the result of the paired t-test analysis of the pre-course

and post-course of TPACK survey completed by 191

participants, the course was found to be significantly

helpful in improving teachers’ confidence in all

TPACK dimensions, including TK, PK, CK, TCK,

TPK, PCK, TPACK. The results of the survey and the

teacher participants’ design artefacts show the growth

of CK and TCK among the teachers, including IoT

concepts, AI concepts, and problem-solving skills.

This growth is helpful for teachers in teaching and

supporting STEM activities with AI components in

their schools, and in developing the related concepts

and problem-solving abilities in their students in this

context. The teachers were also inspired by the

pedagogy of “to play,” “to inquire,” “to assemble,”

“to code,” “to reflect,” and “to create” to deliver

STEM activities for teaching IoT and AI concepts and

were exposed to ways to develop problem-solving

abilities and foster digital creativity in learners.

Two limitations of this study are identified. First,

without data from primary students whose teachers

participated in the development courses, it remains

unclear if the observed improvements in teachers'

competency in teaching STEM activities with AI

components effectively enhance students'

understanding of related concepts, problem-solving

skills, and digital creativity. Second, the confidence

and capabilities of teachers in teaching STEM with

AI components may fluctuate after they apply this

teaching in real-world classrooms. In this respect, this

study suggests two directions for future

investigations. First, there is a need to analyse

primary students’ progression in IoT and AI concepts

and their views of their problem-solving abilities and

digital creativity development after participating in

STEM activities in their schools. A multi-level

analysis can then be conducted on the effect of the

development course on teachers and whether this

effect cascades to the achievement of students taught

by these teachers. Second, it would be interesting to

further investigate the teachers’ views of their

teaching competency after they teach their students in

STEM classrooms, and especially to compare their

responses with those collected from the post-course

survey.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the funding

support of this Coolthink@JC project from the Hong

Kong Jockey Club Charities Trust (Project No.

EdUHK C1136).

REFERENCES

Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari,

M., & Ayyash, M. (2015). Internet of things: A survey

on enabling technologies, protocols, and applications.

A Proposed TPACK Model of Teaching STEM with AI Components: Evaluating a Teacher Development Course for Fostering Digital

Creativity

169

IEEE Communications Surveys Tutorials, 17(4), 2347–

2376. https://doi.org/10.1109/COMST.2015.2444095

Ashton, K. (2009). That “internet of things” thing. RFID

Journal, 22(7), 97–114.

Cavlazoglu, B., & Stuessy, C. (2017). Changes in science

teachers’ conceptions and connections of STEM

concepts and earthquake engineering. The Journal of

Educational Research, 110(3), 239–254.

Frank, M. J. K. (2002). What is “engineering systems

thinking”? Kybernetes: The International Journal of

Systems and Cybernetics, 31(9,10), 1350–1360.

Ghosh, A., Chakraborty, D., & Law, A. (2018). Artificial

intelligence in Internet of things. CAAI Transactions on

Intelligence Technology, 3(4), 208–218.

https://doi.org/10.1049/trit.2018.1008

Hong Kong Education Bureau. (2015). Promotion of STEM

Education: Unleashing Potential in Innovation.

Katare, G., Padihar, G., & Qureshi, Z. (2018). Challenges

in the integration of artificial intelligence and internet

of things. International Journal of System and Software

Engineering, 6(2), 10-15.

Kong, S. -C. (2023). Pedagogical design of STEM activities

for developing problem-solving skills and digital

creativity of primary students in the internet of things

era: Six-step STEM pedagogy. In W. M. So & Z. H.

Wan, & T. Luo (Eds.), Cross-disciplinary STEM

learning for Asian primary students: Design, practices

and outcomes (pp. 147–163). NY: Routledge.

Kong, S.-C., & Lai, M. (2021). A proposed computational

thinking teacher development framework for K-12

guided by the TPACK model. Journal of Computers in

Education. Advance online publication.

Kong, S.-C., Lai, M., & Sun, D. (2020). Teacher

development in computational thinking: Design and

learning outcomes of programming concepts, practices

and pedagogy. Computers & Education, 151, 103872.

Kong, S.-C., & Zhang, G. (2021). A conceptual framework

for designing artificial intelligence programmes for

educated citizens. In Kong, S. C. et al. (Eds.),

Proceedings of the 25th Global Chinese Conference on

Computers in Education, GCCCE 2021. Hong Kong:

The Education University of Hong Kong.

Lin, C.-H., Yu, C.-C., Shih, P.-K. & Wu, L.-Y. (2021).

STEM-based artificial intelligence learning in general

education for non-engineering undergraduate students.

Educational Technology & Society, 24(3), 224–237.

Mishra, P., & Koehler, M. J. (2006). Technological

pedagogical content knowledge: A framework for

integrating technology in teacher knowledge. Teachers

College Record, 108(6), 1017–1054.

Nickerson, R. S. (2015). Conditional reasoning: The unruly

syntactics, semantics, thematics, and pragmatics of

“if”. Oxford University Press.

Ouyang, F., Dinh, A. T., & Xu, W. (2023). A systematic

review of AI-driven educational assessment in STEM

education. Journal for STEM Education Research, 6,

408–426. https://doi.org/10.1007/s41979-023-00112-x

Ouyan, F., Jiao, P., McLaren, B. M., & Alavi, A. H., (2023).

Artificial intelligence in STEM education: The

paradigmatic shifts in research, education, and

technology. CRC Press, Taylor & Francis Group.

Qiu, X., Yu, J., Zhuang, W., Li, G., & Sun, X., (2023).

Channel prediction-based security authentication for

artificial intelligence of things. Sensors, 23(15), 6711.

https://doi.org/10.3390/s23156711

Schreiter, S., Friedrich, A., Fuhr, H., Malone, S., Brünken,

R., Kuhn, J., & Vogel, M., (2023). Teaching for

statistical and data literacy in K-12 STEM Education:

A systematic review on teacher variables, teacher

education, and impacts on classroom practice. ZDM –

Mathematics Education.

Sedrati, A., Ouaddah, A., Mezrioui, A., & Bellaj, B.,

(2022). IOT-Gov: An IOT governance framework

using the blockchain.

Computing, 104(10), 2307–2345.

https://doi.org/10.1007/s00607-022-01086-1

Sullivan, F. R., & Heffernan, J. (2016). Robotic

construction kits as computational manipulatives for

learning in the STEM disciplines. Journal of Research

on Technology in Education, 48(2), 1–24.

https://doi.org/10.1080/15391523.2016.1146563

Touretzky, D., Gardner-McCune, C., Martin, F., &

Seehorn, D. (2019). Envisioning AI for K–12: What

should every child know about AI? In Proceedings of

the Thirty-Third AAAI Conference on Artificial

Intelligence (AAAI-19) (pp. 9795–9799). AAAI Press.

Van Vo, D., & Csapó, B. (2023). Exploring inductive

reasoning, scientific reasoning and science motivation,

and their role in predicting STEM achievement across

grade levels. International Journal of Science and

Mathematics Education, 21, 2375–2398.

https://doi.org/10.1007/s10763-022-10349-4

CSEDU 2024 - 16th International Conference on Computer Supported Education

170