Programming and CT education is primarily conducted through two approaches: the plugged approach and the unplugged approach. The plugged approach involves using digital devices such as tablets, computers, and the Internet. In contrast, the unplugged approach aims to teach programming and CT without any digital devices, instead utilizing materials like pen and paper, cards, or engaging in physical activities (Otterborn et al., 2020). Romero et al. (2018) summarized the key benefits of the unplugged approach, including embodied learning, reduced cognitive load, and concrete analogies. The unplugged approach often incorporates physical actions and tangible manipulation, aligning well with the learning styles of young children. Furthermore, compared to digital tools, incorporating unplugged materials in programming and CT education could minimize distractions that divert children’s attention and reduce cognitive load, which refers to information-processing (attentional or working-memory) demands (Block et al., 2010). Lastly, unplugged activities are built upon the construction of tangible and concrete analogies, facilitating the learning of abstract concepts related to programming and CT. Several studies have explored the effectiveness of the unplugged approach in promoting learners’ CT (del Olmo-Muñoz et al., 2020; Saxena et al., 2020; Ahn et al., 2021; Li and Yang, 2023). In this study, the way the teacher employed to teach programming and CT is the unplugged approach.

The goal of early programming and CT education is not to prepare children to become programmers or algorithmic engineers but rather to foster their CT. As argued by Resnick and Robinson (2017), children do not simply “Learn to Code” but rather “Code to Learn” and “Learn Through Coding.” Thus, our interest lies in identifying the core content of CT covered and emphasized in early childhood teachers’ instruction of programming and CT. To achieve this, we reviewed the CT content framework in ECE.

There is a lack of a consistent content framework for CT in ECE (Zhang and Nouri, 2019). After comparing different CT frameworks, Zeng et al. (2023) used Brennan and Resnick’s (2012) three-dimensional CT framework to identify CT components that were proven appropriate for young children to learn and established the CT curriculum framework for ECE. This framework articulates the core content in early programming and CT education, covers CT concepts (i.e., control flow/structures, representation, and hardware/software), CT practices (i.e., algorithmic design, pattern recognition, abstraction, debugging, decomposition, iteration, and generalizing), and CT perspectives (i.e., expressing and creating, connecting, perseverance, and choices of conduct) (Zeng et al., 2023) (see Table 1).

This section summarizes the teaching context, activity structure, pedagogical approaches, and pedagogical strategies previously used to foster children’s programming and CT skills (see Table 2).

PCK was first introduced by Shulman to emphasize the fundamental role of subject matter in (research in) teacher education and teaching in 1985. In subsequent years, PCK has been defined by different researchers in multiple ways. Despite the various definitions, researchers have identified three essential components of PCK: CK, PK, and knowledge of students’ understanding (Rojas, 2008; McCray and Chen, 2012; Zhang, 2015). Figure 1 illustrates how these three components are interrelated to the construct of PCK (McCray and Chen, 2012). This study specifically examined teachers’ CK and PK of programming and CT.

CK is the knowledge of what to teach. It encompasses knowledge of the discipline to be taught, a thorough understanding of that knowledge, and an understanding of the relationships between topics of the discipline (Krauss et al., 2008). In this study, we focused specifically on the first two aspects, i.e., whether the teachers knew the programming and CT knowledge to be taught and whether teachers had a deep understanding of them.

PK is the knowledge of how to teach. There are two types of PK: general pedagogical knowledge (GPK) and content-specific pedagogical knowledge (CPK). GPK comprises comprehension of various educational philosophies and learning theories, general knowledge of learners and basic teaching rules, and familiarity with classroom management principles and strategies (Grossman and Richert, 1988). CPK is the knowledge of instructional strategies unique to a particular subject or topic (Zhang, 2015). In this study, we examined both the GPK and CPK.

Given the scant existing literature in this field, we conducted a comprehensive review focusing on the PCK of both preservice and in-service teachers across all educational levels. Several researchers have discovered that both pre-service and in-service teachers possess limited knowledge of CT and little knowledge of how to teach programming and CT (Bower and Falkner, 2015; Chalmers, 2018; Sands et al., 2018).

Accordingly, it has been suggested by researchers that there is a pressing need to enhance teachers’ PCK through pre-service and in-service training programs to facilitate the integration of CT into their classrooms (Yadav et al., 2017; Chalmers, 2018; Haines et al., 2019). Chalmers (2018) specifically emphasized that a deeper understanding of CT concepts, practices, and perspectives is crucial for teachers to effectively incorporate CT into the primary curriculum. Çakıroğlu and Kiliç (2020) proposed a course model and evaluation tools aimed at improving teachers’ PCK for teaching CT via robotic programming.

Within the context of ECE, Strawhacker et al. (2018) found that teachers who possessed a solid foundation of CK exhibited more purposeful use of the programming tool and gave more explicit support. Similarly, Wang et al. (2020) found that the case teacher intentionally employed various strategies in his programming and CT instruction because of his clear understanding of CT skills that young children need to develop.

Previous research has indicated a need to improve teachers’ PCK through training to help them implement programming and CT education (Yadav et al., 2017; Chalmers, 2018; Haines et al., 2019). To provide targeted training to help teachers acquire the necessary PCK and effectively deliver programming and CT education, it is crucial to clearly understand the status of teachers’ PCK in programming and CT education. However, based on our thorough review of the existing literature, there is a lack of research specifically examining the status of CK and PK of programming and CT among early childhood teachers. As teachers’ CK and PK can be demonstrated in their teaching (Zhang, 2015), to examine early childhood teachers’ CK and PK, we proposed the following questions:

This study was conducted in a provincial first-class public kindergarten located in Wenzhou, China, with a specific focus on STEM education. Recognizing the increasing importance of programming and CT education, the kindergarten embarked on a new educational initiative to integrate programming and CT into its curriculum. As an initial step, instead of implementing programming and CT education across all classes, the kindergarten decided to initiate a pilot program. They selected one class from each of the age groups: K1 (3-year-olds), K2 (4-year-olds), and K3 (5-year-olds), led by one teacher in each class.

We chose the K3 class for observation because the teaching content of the unplugged curriculum in the K3 class was built upon that of K1 and K2 and covered all the CT skills of the unplugged curriculum, thus allowing us to examine RQ1 comprehensively. The K3 class consisted of 32 children aged 5–6 years, along with two teachers and a nurse. For the purposes of this study, we selected Ms. Wu, who was responsible for teaching programming and CT in this class and who enthusiastically volunteered to join our study.

Initially, the three experimental classes utilized a plugged-in programming tool named MOBLO. MOBLO is a hybrid kit that enables young children to program a virtual character on the screen by manipulating tangible programming blocks. Due to concerns regarding the potential damage of screen usage on children’s eyesight, the kindergarten developed an unplugged, screen-free programming tool and an unplugged programming curriculum, taking inspiration from MOBLO. Subsequently, the three experimental classes conducted the unplugged programming curriculum. Notably, Ms. Wu not only implemented the unplugged curriculum but also participated in designing the unplugged programming tool and the unplugged curriculum.

Ms. Wu possessed 11 years of work experience in the field of ECE. She initially graduated from a local normal university with an Associate’s degree in ECE. Following 7 years of work experience, she pursued a Bachelor’s degree in ECE through adult correspondence education. However, neither of these programs included any courses related to early programming or CT education. Ms. Wu’s exposure to programming education came exclusively from the MOBLO company. To implement programming education with the MOBLO programming tool in the kindergarten, the MOBLO company provided training to teachers. This training primarily focused on instructing teachers on the utilization of the MOBLO programming tool and how to teach programming using the lesson plans provided by the MOBLO company.

The unplugged programming tool developed by this kindergarten, like other coding sets, consists of three parts: (1) The object being programmed: The object being programmed in this coding set is a pawn named Qiqi, who is also the protagonist of the stories in the unplugged curriculum; (2) Programming tasks: The teacher or children set up programming tasks by putting the Tool Blocks (representing tools Qiqi needs to obtain to solve problems) and Scenario Blocks (representing the characters, place, and things that happen in the story) on a 10 by 10 Grid Map. (3) Programming blocks: Children program the routes Qiqi takes by placing wooden Programming Blocks (including Directional Blocks, Number Blocks, Loops Block, and Conditional Instruction Card) in the Programming Area. For example, in the context of exploring outer space, Qiqi first needs to collect tools such as the spacesuit, oxygen kit, and translator. On his journey to other planets, he must avoid meteorites. When encountering problems, he needs to use tools (for example, using a translator when meeting an alien). Eventually, he reaches other planets (see Figure 2). Appendix 3 shows how to make a similar coding set using readily available materials.

As teachers’ PCK can be demonstrated in planning, implementing and reflecting on teaching (Zhang, 2015), lesson plans, videotaped programming activities, and audiotaped interviews were collected as our data to establish triangulation (Yin, 2009), as well as memos following each observation and interview.

To analyze the CT concepts, practices and perspectives that are covered and emphasized in the early childhood teacher’s teaching of programming and CT, we used the CT curriculum framework for ECE (Zeng et al., 2023), which has a detailed and clear definition of each CT component, as the CK Framework (see Table 1). The CK Framework includes three dimensions: CT concepts, practices, and perspectives.

To examine how the early childhood teacher supported children’s programming and CT learning, we developed the PK Framework (see Table 2). The PK Framework comprises four dimensions: teaching context, activity structure, pedagogical approaches, and pedagogical strategies. We constructed the indicators under each dimension based on the aforementioned literature review. Moreover, we provided a clear definition for each indicator in the PK framework (see Table 2).

Then the first author and the second author used the two frameworks to analyze the video data, the interview data, and the lesson plans. The following explains the process of data analysis. Appendix 1 shows a few examples of data analysis.

This study was conducted with the ethical approval from the Human Research Ethics Committee (HREC), the authors’ university (Reference No. 2021-2022-0334). A letter outlining the research and consent forms was provided to the kindergarten principal and Ms. Wu, and permission was obtained from them. Since the children in this study were 5–6 years old, letters and consent forms were also provided to parents/guardians through kindergarten.

The findings were validated through data triangulation, member checking and inquiry auditing (Creswell, 2014). We collected data from multiple sources for triangulation. Member checking was conducted by re-interviewing Ms. Wu to ensure that her ideas stayed in line with her previous interview responses and the researchers’ interpretations. In addition, two senior researchers in ECE acted as the auditors to ensure the rigor of the research procedure and confirm that the findings appropriately reflected important aspects of the observations, interviews, and lesson plans.

Our evidence revealed that Ms. Wu emphasized CT concepts across her programming and CT teaching instead of CT practices and CT perspectives (see Table 3).

Ms. Wu employed group activities to teach programming and CT with highly structured, task-based activities (see Table 4). She integrated CT skills into tasks that gradually increased in difficulty and guided children to learn CT skills by completing these tasks. Ms. Wu provided ample time and support for the children’s self-exploration and acted as a facilitator and collaborator rather than an authority figure.

Ms. Wu utilized a range of pedagogical strategies to support children’s programming and CT learning. Our analysis identified eight strategies, five of which were effective: the unplugged approach, contextualization, embodied cognition, external memory support scaffolding, and a step-by-step strategy for teaching loops. Ms. Wu embraced the unplugged approach to teaching programming and CT, which involves screen-free and hands-on activities. Additionally, she employed contextualization to provide meaningful contexts for programming and CT learning, such as integrating loops learning into the narrative of aiding Qiqi in exploring planetary mysteries. Furthermore, she utilized embodied cognition. The children interacted with an external surrogate named Qiqi, manipulating it to navigate a grid map based on provided instructions. They also physically moved within the grid on the floor, following the given directives. External memory support scaffolding was another strategy Ms. Wu employed, such as using visible Programming Blocks for notating children’s algorithms to support their thinking and problem-solving. Furthermore, she implemented a step-by-step teaching strategy, drawn from early mathematics, to effectively teach loops.

However, the strategies of demonstration, pair programming, and differentiated instruction were not effectively utilized. While Ms. Wu often used demonstrations to exhibit how to identify coordinates, verify routes, and organize materials, she did not model problem-solving skills like debugging and decomposition, nor did she exemplify cooperation and a positive attitude toward mistakes. In addition, Ms. Wu employed pair programming, wherein two children with neighboring school numbers collaborated on programming tasks. However, observations indicated that Ms. Wu did not intentionally and actively observe and facilitate children’s collaboration. Consequently, pair programming proved ineffective, frequently resulting in a lack of genuine interaction and cooperation between the two children, or in some instances, one child would assume a dominant role while the other remained disengaged. Furthermore, Ms. Wu implemented differentiated instruction after recognizing that less capable children struggled to keep up and remained less engaged during programming and CT activities. However, her approach simply involved segregating children into two groups based on their abilities, slowing down the teaching pace, removing challenging tasks for the less capable group, and failing to provide targeted scaffolding for these children to grasp programming and CT concepts.

Although this study is the first to examine an early childhood teacher’s CK and PK for early programming and CT, it does come with certain limitations. Firstly, this study was based on a single teacher as a case study. While this chosen case has provided insights into the teaching of programming and CT within the context of Chinese early childhood education, caution should be exercised when generalizing the findings to broader contexts or other educators. Secondly, this study solely focused on the early childhood teachers’ CK and PK of programming and CT, without investigating the teacher’s knowledge of students, which is a crucial component of teachers’ PCK. Future studies should also explore early childhood teachers’ understanding of students’ learning in programming and CT.

This study has important implications for practice. CT encompasses more than just CT concepts; it also involves CT practices and perspectives (Brennan and Resnick, 2012). When introducing CT in ECE settings, the goal is not simply to teach young children to “Learn to Code” but rather to equip them with problem-solving skills and attitudes that can be applied beyond programming (Lye and Koh, 2014). CT practices and perspectives are exactly related to problem-solving skills and attitudes. However, it was found that the case teacher’s intentional teaching in CT practices and perspectives was limited, and her knowledge of CT practices and perspectives was weak. Thus, it is crucial to provide teachers with professional support to help them understand the goal of programming and CT education and to enhance their knowledge of CT practices and perspectives. This will help teachers move from a focus on teaching CT concepts to intentionally integrating CT practices and perspectives into their teaching practices.

Furthermore, Ms. Wu’s teaching approach was limited to whole-group activities. She lacked awareness of providing opportunities for children to learn and apply programming and CT in their daily lives. By developing a clear goal for programming and CT education and acquiring a strong CK in CT practices and perspectives, teachers can become equipped with the awareness, knowledge, and ability to integrate programming and CT into children’s daily lives. This inclusive approach ensures that programming and CT skills are accessible to all children, particularly those from disadvantaged backgrounds.

Additionally, there are various pedagogical approaches for teaching early programming and CT, ranging from a task-based approach, where learning activities are centered around tasks guided by adults (McCormick and Hall, 2021), to project-based learning, which emphasizes student-centeredness (Kokotsaki et al., 2016). However, Ms. Wu solely employed a highly structured task-based approach, which prioritized guidance but overlooked students’ autonomy and creativity. Therefore, teachers should adopt a flexible way by combining different pedagogical approaches to teach programming and CT. This enables the provision of necessary guidance while also encouraging students’ autonomy and creativity.

Lastly, based on the unplugged programming tool developed by the case kindergarten, a guide for crafting an unplugged coding set has been innovatively proposed. Programming tools play a crucial role in implementing programming and CT education. This age-appropriate, board game-like coding set extends young children the opportunity to engage in programming and CT activities within both formal and informal settings. Additionally, this unplugged coding set boasts ease of reproduction, as it can be made following straightforward steps and utilizing readily available materials.

This study makes an important contribution to the research. The constructed PK framework for programming and CT, based on an extensive literature analysis, provides a useful tool for analyzing teachers’ PK in teaching programming and CT. In addition, the study presents a model case showcasing the application of CK and PK frameworks to investigate teachers’ CK and PK in early programming and CT education.

Moreover, this study found that teachers had limited CK in CT practices and perspectives and insufficient content-specific pedagogical knowledge (CPK). Therefore, further research could explore ways to enhance teachers’ pedagogical content knowledge in programming and CT education through training programs. It is especially important to investigate how teachers can effectively apply the acquired CK and CPK to their own teaching context to facilitate the integration of programming and CT into classrooms. Previous studies have demonstrated that coaching is critical in facilitating the transfer of training content to teachers’ specific teaching situations (Neuman and Cunningham, 2009). Hence, future researchers could explore transferring the coaching model to the Chinese cultural context to enhance teachers’ intentional and effective teaching of programming and CT.