Integrating Computer Science and Informatics Education in Primary Schools: Insights from a Slovenian Professional Development Initiative

Abstract

In this study, we present a professional development programme for teachers launched to introduce Computer Science and Informatics (CSI) in primary education in Slovenia. The study aims to examine which CSI core concepts teachers most frequently choose to integrate into their lessons when given the freedom to select the topics within the framework, and to explore how students engage with and respond to these activities, as reported in teachers’ reflections. This study is based on reflective feedback from forty-seven teachers from seven primary schools who implemented interdisciplinary lessons that integrate CSI content into existing primary school curricula. Qualitative data from 152 reflections were used to support our research findings. The results show that teachers most frequently introduced the concepts from the content area of algorithms and programming. In contrast, content areas such as computing systems, networks and the internet, data and analysis, and impacts of computing received less attention. Teachers reported that students were motivated and engaged, although some challenges emerged, including difficulties in solving tasks or following instructions. As this pilot study reports on the first year of a two-year initiative, the findings provide preliminary insights into how a structured professional development programme for teachers can support interdisciplinary approaches in CSI education.

1. Introduction

Sustainability increasingly intersects with computer science and informatics (CSI), where the concept has evolved from focusing on software durability in the 1970s and later on sustained organisational contexts to now encompass broader societal and ethical implications [1]. CSI education supports social sustainability [2] by bridging the digital divide—increasing access to computing skills and infrastructure, particularly for students from rural areas [3,4]; opening pathways to high-demand, well-paid careers that promote social mobility and economic equity [5]; developing students’ problem-solving skills [6,7], transferable across disciplines and contexts, enabling them to engage with complex challenges [8]; and fostering awareness of the environmental impacts of digital tools, such as energy consumption and electronic waste, encouraging responsible and sustainable technology use [9]. These aspects align with Sustainable Development Goal 4 (SDG 4: Quality Education), emphasising equitable access to quality and inclusive education for all [10].

Recognising the increasingly central role of computing, researchers and policymakers argue that CSI education should be integrated across the entire K–12 curriculum [11,12,13]. What was once considered an elective subject is now a foundational element of modern education [11]. This shift encompasses not only the inclusion of programming-related skills but also the development of a broad set of competencies, such as computational thinking (CT) [14,15]; understanding how data is collected, processed, and modelled [16]; practising digital safety and privacy [17]; and recognising the societal and ethical implications of computing [18].

The Slovenian government has acknowledged this need by funding several initiatives to introduce CSI education at different levels of education. This paper is part of the UTRINKI project (Digital transformation of education for a sustainable future—students, sustainability, computer science and informatics as a challenge), which seeks to advance this movement in primary education.

However, prior research suggests that many primary school teachers may not feel adequately prepared to teach CSI concepts or fully understand the rationale for their inclusion [19,20]. Professional development (PD) initiatives have been introduced to address these challenges, yet they are often short-term and lack the sustained, systemic support necessary for lasting impact [20].

A recent review of 47 studies on the effectiveness of PD in K–12 CSI education highlights a lack of research on student learning outcomes and emotional responses to CSI integration [21]. Additionally, how teachers translate the K–12 Computer Science Framework into classroom practice—particularly when they have the freedom to choose content—remains underexplored. To address these gaps, we present findings from the first year of a two-year PD programme, examining which core content areas from the K–12 Computer Science Framework [22] are most frequently addressed in teachers’ interdisciplinary lessons, and how students engage with CSI-related activities based on teachers’ reflections. The following research questions guide the study:

(1)

Which domains from the K–12 Computer Science Framework are most and least frequently addressed by primary school teachers in the UTRINKI project during interdisciplinary lessons implemented as part of a PD Programme?

This question examines the CSI concepts and practices introduced in the lessons, the subject areas in which the CSI concepts were integrated, and the modality of the CSI activities.

(2)

How do primary school students in the UTRINKI project engage with and respond to CSI-related activities, based on teachers’ post-lesson reflections?

This question examines students’ affective responses to the lessons and the difficulties and barriers they encountered during the activities.

2. Literature Review

2.1. CSI Education in Primary School Curricula

Computational concepts in education are promoted under various closely related terms, including programming, CT, Computer Science (CS) education, Informatics education, and Computing education [11,13]. Voogt et al. [23] distinguish between programming, CT, and CS education, highlighting that programming is just one context for engaging with both CS and CT. CT underpins how students engage with problems and design solutions [13] and is characterised by a set of its components—abstraction, decomposition, pattern recognition, algorithmic thinking, and debugging [8,24,25]. Although CT originated in CS, its problem-solving strategies extend beyond the discipline and are applicable more broadly [11].

Furthermore, Sampson et al. [26] note that informatics, CS, and computing are often conflated, although they differ in focus. Informatics emphasises the study of information systems, CS delves deeper into theoretical foundations and algorithms, and computing encompasses a broader range of practical applications [26]. This paper uses the term Computer Science and Informatics (CSI) education to reflect the discipline’s broad scope and align with terminology commonly used in the Slovenian educational context [27].

Several frameworks have been developed to define the core elements of CSI education, such as the Informatics Reference Framework for School [28] and the K-12 Computer Science Framework [22]. The latter specifies core concepts that represent major content areas in the field of CS; namely, algorithms and programming, computing systems, data and analysis, networks and the internet, and impacts of computing. The framework also defines computational practices as behaviours and thinking that computationally literate students use to engage in today’s data-rich and interconnected world. CT is at the heart of those practices and is delineated by many of them [22].

Moreover, recent scholarship has highlighted critical debates in the field, particularly regarding equity in CSI education and integrating emerging technologies such as AI into primary school curricula. Researchers argue that equitable CSI education must address systemic barriers (e.g., gender gaps, socio-economic disparities, racial and ethnic inequities, and rural access to resources) to ensure all learners benefit from computing opportunities [29,30,31]. In parallel, there is growing discussion about how AI tools can be incorporated into education, balancing technical understanding with ethical, social, and critical perspectives [32,33,34]. For instance, teaching students to craft effective prompts [35,36] and introducing more advanced applications of AI systems, such as humanoid robots [37,38,39], are examples of efforts to broaden students’ engagement with AI. However, several scholars caution against the uncritical integration of AI technologies into classrooms, arguing that such practices may exacerbate existing educational inequalities, compromise student privacy, perpetuate algorithmic bias, and reduce critical thinking and creativity of students [40,41].

There are two main approaches to integrating CSI in compulsory education: (1) as part of a dedicated computing-related subject or (2) as a cross-curricular theme shared across subjects. For instance, Kwon et al. [42] developed dedicated CSI units for elementary students, including block-based coding and project-based learning, delivered separately from other subjects. In contrast, Ozturk et al. [43] developed a PD where teachers integrated CSI concepts into existing subjects, such as English language, arts, math, science, and social studies, through problem-based learning within a standards-based curriculum.

2.2. Professional Development in CSI Education

To effectively integrate CSI in curricula in primary education, teachers need to build a strong foundational understanding of computational concepts and practices [44,45,46]. A persistent obstacle in this regard is the lack of disciplinary knowledge and experience in CSI among many primary school teachers [19,20,47,48]. Researchers and policymakers have increasingly emphasised the importance of structured PD initiatives to strengthen teachers’ conceptual and pedagogical skills in CSI [20,49,50]. A wide range of PD programmes has been developed to address these needs, some focusing on CT as a foundation [44,51,52,53] and others more directly targeting CSI education [54,55,56].

Two recent studies have examined PD in CSI education [20,57]. According to Liu et al. [20], PD initiatives ranged widely in scale, from small groups of fewer than 10 teachers to large-scale programmes involving hundreds or over a thousand participants. These were delivered through various formats, including regional initiatives and MOOCs. The duration also varied, from one-day sessions to multi-year programmes, although most PDs identified in their review were short-term workshops. A wide selection of tools and approaches was used, including online platforms like Scratch, unplugged activities, and robotics kits, mainly targeting the development of CT and, less often, concepts from CSI education. Ni et al. [57] reviewed 48 articles, with most PD programmes being held in the United States. Their review identified four common strategies for building PD communities: developing CS teacher leaders, creating resource repositories, offering MOOCs for self-paced learning, and facilitating online forums. Most often used evaluation methods of PDs include surveys, followed by interviews and teachers’ reflections. Both reviews highlight the need for PD communities to build sustainable, prolonged, and engaging learning experiences [20,57].

The following studies illustrate how PD initiatives support the integration of CSI education in primary education. Ozturk et al. [43] conducted a six-month PD programme with eleven teachers with limited CSI knowledge. Following the training, teachers implemented CSI-related content into their lessons and reported increased student engagement and enhanced collaborative learning. They emphasised that problem-based, interdisciplinary lessons were more effective than routine instruction. Mansour et al. [56] examined a three-year PD programme involving 57 teachers and 871 students. The study reported significant increases in teacher confidence and CSI content knowledge following the programme, particularly among teachers who initially reported low confidence. Students also demonstrated substantial learning gains, particularly when taught by teachers with higher confidence. El-Hamamsy et al. [55] conducted a large-scale PD initiative involving 350 primary teachers in Switzerland. Activities included unplugged approaches, robotics, and visual programming. Teachers evaluated the PD positively, and their understanding and representation of CSI improved. Voluntary adoption rates reached 97% during the programme. They remained high at 80% the following year, delivering over 2800 CS lessons across two years, most of which used unplugged methods. Similarly, Colwell et al. [54] implemented an online PD programme for integrating CSI into K–5 literacy instruction. Through five iterative cycles involving classroom implementation, written reflections, and community of practice meetings, ten teachers reported increased appreciation of CSI fundamentals. They favoured unplugged approaches, which reduced reliance on digital devices. Teachers’ attitudes toward CSI integration became more positive over time.

Our PD programme, described in Section 3.2, aligns with several features of international initiatives while offering new insights into teacher development. Like Ozturk et al. [43], we focus on integrating CSI activities within existing subjects through interdisciplinary lessons, rather than treating CSI as a stand-alone subject. Our programme involves a comparable number of teachers as the three-year PD by Mansour et al. [56], allowing for close collaboration and peer support. Like El-Hamamsy et al. [55], we combine unplugged activities with plugged ones. Similarly to Colwell et al. [54], we organise the programme into three iterative classroom cycles, which support teachers in improving their teaching.

While the reviewed studies offer valuable insights into teacher development and instructional strategies, they do not examine which content areas are teachers’ priorities in practice or provide in-depth analyses of how students engage with CSI-related lessons implemented through PD programmes. This study addresses these gaps by exploring instructional focus and student experiences based on teacher reflections.

3. Methods

3.1. Context and Participants

Many European Union Member countries have recently revised their statutory curricula to introduce basic CSI concepts in primary education to develop students’ computational skills [13]. Slovenia has been actively preparing for its inclusion yet has not formally integrated CSI into its compulsory national primary curriculum. As part of these efforts, the Ministry of Education established the Department for Digitalisation in 2021, and in the years that followed, launched several pilot projects to support the digital transformation of education. Among these is the UTRINKI project, which aims to introduce CSI education in pre- and primary school.

The project is coordinated by researchers from the Faculty of Natural Sciences and Mathematics at the University of Maribor. The consortium includes seven primary schools and one kindergarten; however, this study focuses exclusively on the results from the primary schools. The sample is regional and convenience-based. Six of the schools are in the Podravje region and one in the Koroška region, placing all participating institutions in the north-eastern part of Slovenia. Of these, one school is in a city, three are in suburban areas, and three are in rural locations, reflecting a mix of urbanisation levels across the participating schools. Forty-seven primary school teachers from these seven schools, teaching Grades 1 to 5, are involved in the project.

3.2. Professional Development Framework and Timeline

A central component of the project is the PD programme for teachers, which is designed to build capacity for integrating CSI in primary education. Our PD programme aims to strengthen teachers’ CSI content knowledge, enhance pedagogical content knowledge, and build their confidence in delivering CSI instruction to students.

The programme follows the Innovative Teacher Training and Support Model [58]. The model consists of five key stages: (i) teacher participation in training, (ii) classroom implementation, (iii) reflection, (iv) teacher acting as a trainer, and (v) teachers’ assessment of competencies.

The timeline of the PD programme (Figure 1) began in November and December 2024 with initial coordination meetings between members of the university research team and the school project coordinators and principals. During these meetings, we established the foundation for collaboration, clarified the project goals, and introduced the structure of the PD programme.

In January 2025, we held a two-day workshop with the participating teachers in the project (Stage i). During the workshop, we introduced the Slovenian CSI curriculum framework, developed by the RINOS [27] research group based on the K–12 Computer Science Framework [22]. Teachers engaged with interdisciplinary teaching strategies and participated in practical activities based on CS Unplugged [59] and the Slovenian platform Vidra [60], incorporating translated CS Unplugged resources alongside locally developed materials tailored to the Slovenian educational context.

Between February and May 2025, teachers applied the knowledge gained during the initial training by implementing CSI-focused interdisciplinary lessons across three iterative classroom cycles (Stage ii). After each lesson, they completed a structured reflection form, prepared by the research team, and shared their insights on the project’s Moodle platform (Stage iii). In addition to submitting reflections, teachers were encouraged to engage in the online forum, where they could share experiences, discuss challenges, and access lesson examples contributed by researchers and peers. Although the project did not prescribe specific topics, we asked teachers to implement at least three lessons aligned with the Slovenian K–12 Computer Science Framework version [27].

As part of Stage iv, some teachers also took on the role of trainers by opening their classrooms to colleagues, who observed the lessons, exchanged experiences, and provided peer support. This open classroom component was optional and implemented in several schools during the spring term. At the end of each classroom cycle, teachers participated in self-assessments of their competencies in collaboration with their school coordinators (Stage v). Following the first cycle, in March 2025, school coordinators met with the university research team to discuss challenges and the need for additional support. These discussions informed the ongoing refinement of the PD process.

Throughout the year, schools received project funding to acquire educational software, programmable robots, and external services, such as additional teacher workshops. These resources supported unplugged and plugged approaches to teaching CSI content, allowing teachers to implement diverse tools in their classroom settings.

In June 2025, we held a closing workshop that brought together participating teachers to reflect on the project’s first year and prepare for the second. We focused the seminar on content areas that had received less attention, particularly those beyond algorithms, and placed special emphasis on plugged activities in CSI education. We also introduced additional resources, including Scratch, Code.org, Barefoot Computing, and BBC Bitesize.

3.3. Instrument

We developed a structured reflection tool to examine how primary school teachers implemented CSI activities in the PD programme and how students experienced these activities. The instrument combined quantitative and qualitative items to support teacher reflection and was administered online via the 1 ka questionnaire tool. Teachers completed it immediately after each implemented lesson, and all responses were collected anonymously to ensure confidentiality. The questionnaire included four content-focused questions aligned with the study’s research aims (

Table 1. Questionnaire items.

No.	Item	Response Scale
1	Please indicate the extent to which this lesson focused on the following core content areas of CSI: computer systems, data and analysis, algorithms and programming, networks and the internet, impacts of computing	1 = does not apply at all … 5 = fully applies
2	Please specify the topic and content of the lesson that was conducted.	Open response
3	Please rate how the students experienced the lesson. (This evaluation should be conducted together with the students.) Included dimensions: motivation, engagement, enjoyment, curiosity, enthusiasm, surprise, exclusion of certain students, dominance of individual students, stressfulness	1 = very low … 5 = very high
4	Please evaluate the effectiveness of the implemented lesson in CSI.	Open response
5	Year level	1, 2, 3, 4, 5

).

First, teachers rated the extent to which the implemented activity addressed the five core CSI content areas (computing systems, data and analysis, algorithms and programming, networks and the internet, and impacts of computing) on a five-point Likert scale. Second, they described the lesson content in an open-text response. Third, teachers and students evaluated nine affective dimensions of the lesson experience, each using a single-item five-point Likert scale. The nine affective dimensions were: (i) motivation (willingness to participate), (ii) engagement (active involvement during the activity), (iii) enjoyment (perceived fun or satisfaction), (iv) curiosity (interest in exploring further), (v) enthusiasm (energy and excitement shown), (vi) surprise (novelty or unexpectedness of the activity), (vii) exclusion of certain students (extent to which some students were left out), (viii) dominance of individual students (the extent to which a few students controlled the activity), and (ix) stressfulness (perceived pressure or difficulty). Each dimension was evaluated using a single-item rating on a 5-point Likert scale. Single-item measures were used for each dimension to allow teachers, in collaboration with students, to provide an overall assessment of the lesson experience. Fourth, teachers reflected on the overall success of the activity. Additionally, teachers reported the grade level at which the lesson was conducted.

All Likert-type ratings were presented on a five-point scale with clear labels (1 = very low … 5 = very high or 1 = does not apply at all … 5 = fully applies). The instrument was completed immediately after each CSI lesson, ensuring fresh reflections.

Content validity was established through expert review to ensure alignment with research questions and constructs. Internal consistency coefficients (e.g., Cronbach’s alpha) were not calculated; each affective dimension was measured with a single item. Prior research has shown that single-item measures can provide adequate reliability and validity for concrete, unidimensional constructs, especially in applied contexts where brevity is critical [61,62,63]. Moreover, triangulation with qualitative reflections strengthens construct validity. Standardised instructions supported objectivity, and sensitivity was ensured by the five-point response scale.

3.4. Data Collection and Analysis

A total of 152 teacher reflections were submitted and included in the analysis. The first reflection was submitted on 20 January 2025, and the final on 26 June 2025. The majority were submitted between early February and late May, corresponding with the main period of CSI activity implementation in classrooms.

Once survey responses were collected, we analysed the data. The Likert-scale data were analysed in SPSS (version 29.0.2.0) using descriptive and inferential statistics (Spearman correlations, and the Mann–Whitney U test).

Open-ended responses were analysed using Atlas.ti software (version 9.24.0), following an inductive approach. We applied descriptive coding and in vivo coding techniques. Descriptive coding summarises a passage of qualitative data with a word or short phrase—most often a noun—capturing its basic topic [64]. In vivo coding uses participants’ own words or phrases as codes, reflecting their actual language [64]. In practice, we sometimes coded directly using participants’ terms, while other times we created codes based on the underlying meaning of the text.

Firstly, we conducted the initial coding and created a code list. Coders familiarised themselves with the codebook through pilot coding 10 reflections, which were discussed before formal coding began. The unit of analysis was defined as an individual reflection, with multiple codes assigned when a reflection contained distinct ideas. This allowed a single reflection to be associated with more than one CSI concept and practice. Next, a different researcher reviewed all coding decisions collaboratively with the initial coder until a complete consensus was reached. To assess reliability, another researcher independently coded 20% of the responses for each question using the established code list, resulting in Cohen’s kappa values of 0.84 for Question 2 and 0.77 for Question 4, indicating substantial agreement.

An illustrative example can be drawn from Question 2 (Core content areas of CSI) with Teacher no. 13.:

“The students were divided into pairs, and each pair received a card with arrows that guided them to specific letters on the grid (code: algorithm). One student in each pair read the arrows, while the other moved across the grid (code: control). Upon reaching a letter, the student recorded it on the card according to instructions. The letters were intentionally mixed, and students had to rearrange them to form words. A competitive element was introduced by awarding victory to the pair that found the greatest number of letters or words within one lesson.”

While the study included 47 teachers, we collected 152 lesson reflections. For simplicity, we refer to these responses collectively as “teachers” throughout the text.

For Question 2, we generated 50 codes, which we grouped into 19 categories based on shared meanings and further organised into four overarching themes reflecting key patterns in participants’ responses. For Question 4, we applied the same coding process, producing 16 codes grouped into two categories, which we analysed under a single overarching theme related to students’ experience of CSI activities.

Accordingly, we present the results of Question 2 thematically, while the results of Question 4 are presented through categorised reflections. Question 1 provides quantitative data on the extent to which teachers implemented CSI content, supporting the qualitative findings of Question 2. Similarly, Question 3 offers quantitative insights into students’ affective experiences, complementing the qualitative reflections captured in Question 4.

4. Results

4.1. Patterns in the Integration of CSI in Primary Education

This section presents the results from the analysis of structured teacher reflections collected after lesson implementation. The findings are organised into four overarching themes that emerged through qualitative coding: (1) Core CSI content areas (CSI concepts) most frequently addressed; (2) Emergence of cognitive and thinking skills (CSI practices) during lessons; (3) Integration of CSI within school subjects and curriculum areas; and (4) Interfaces and modalities through which the lessons were delivered.

4.1.1. Core CSI Content Areas

The analysis of teacher ratings revealed that algorithms and programming was most frequently included in lessons, receiving the highest average rating (M = 4.42, SD = 1.03). Lower scores were observed for data and analysis (M = 3.11, SD = 1.30), computing systems (M = 2.62, SD = 1.25), impacts of computing (M = 2.47, SD = 1.24), and networks and the internet (M = 2.20, SD = 1.15) (see

Table 2. Teacher ratings of core CSI content areas.

Core Content Area	N	Mean	Std. Deviation	Median	IQR
Computing systems	130	2.62	1.25	2	2
Data and analysis	136	3.11	1.30	3	2
Algorithms and programming	152	4.42	1.03	5	1
Networks and the internet	132	2.20	1.15	2	2
Impacts of computing	130	2.47	1.24	2	2

).

The correlations between teachers’ ratings of the five core content areas showed significant associations (see

Table 3. Spearman’s correlations between teacher ratings of core CSI content areas.

	Computing Systems	Data and Analysis	Algorithms and Programming	Networks and the Internet	Impacts of Computing
Computing systems	1	0.371 **	0.002	0.640 **	0.618 **
Data and analysis		1	−0.222 *	0.383 **	0.412 **
Algorithms and programming			1	−0.042	−0.030
Networks and the internet				1	0.722 **
Impacts of computing					1

Note: Spearman’s rho coefficients are shown. N varies by pairwise deletion (126–151). ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

). Computing systems was correlated with both networks and the internet (ρ = 0.640, p < 0.01) and impacts of computing (ρ = 0.618, p < 0.01), suggesting that teachers who reported addressing computing systems also tended to integrate related topics about digital networks and societal impacts. Data and analysis were positively related to networks and the internet (ρ = 0.383, p < 0.01) and impacts of computing (ρ = 0.412, p < 0.01), indicating a potential trade-off in emphasis between data-related skills and networks, and impacts of computing. A robust correlation emerged between networks and the internet and impacts of computing (ρ = 0.722, p < 0.01), reflecting their conceptual and pedagogical proximity. In contrast, algorithms and programming was significantly correlated only with data and analysis (ρ = − 0.222, p < 0.05).

Qualitative coding emphasises algorithms and programming (N = 118). Computing systems (N = 29), impacts of computing (N = 28), and networks and the internet (N = 28) were referenced less frequently but at comparable levels, consistent with their lower average ratings in the quantitative data. In contrast, while teachers rated data and analysis as the second most addressed content area, it appeared least frequently (N = 24) in the qualitative reflections (see

Table 4. Frequency of content area references in qualitative data.

Category	Codes	Frequency
Algorithms and programming	algorithm, modularity, control, programming	118
Computing systems	devices, binary system, troubleshooting, pixels	29
Impacts of computing	collaboration, safety, everyday life	28
Networks and the internet	cybersecurity	28
Data and analysis	data collection, data storage, data transformation, optimisation, data visualisation, patterns, inference	24

).

After the initial round of coding, data and analysis appeared at relatively low frequencies (N = 18). However, given its significant negative correlation with algorithms and programming (ρ = –0.222, p < 0.05) and the relatively high ratings reported on the Likert-scale by teachers, we conducted a sensitivity analysis by re-examining algorithms and programming excerpts. This review revealed four additional instances where students made inferences based on the information and data provided within algorithmic tasks, and two more excerpts involving sorting objects based on patterns.

We present an example from Teacher No. 42. The excerpt was first coded as an algorithm since students independently created a step-by-step procedure for planting seeds. During a sensitivity analysis, we re-examined the coding to see whether alternative interpretations could apply. Upon review, it was also interpreted as inference, since students had to reason about how to organise the procedure to ensure successful plant growth. Both codes were retained in the final analysis to reflect this nuanced interpretation, demonstrating that coding decisions were tested and adjusted to ensure robustness and transparency:

“We discussed what plants need to grow (food, water, air, sunlight, suitable space…). The students received a worksheet on planting seeds. Using the pictures on the worksheet, they independently created an algorithm for planting the seeds (code: algorithm). The seeds varied in size and planting depth, so students had to decide how to organise the procedure to ensure the plants would grow successfully (new code: inference). The students then exchanged their algorithms and checked whether their classmate’s algorithm made sense.”

Teachers most frequently focused on algorithms and programming in their ratings and qualitative responses; accordingly, we analyse this area in greater detail than those that appeared less often in the data.

Algorithms and Programming

The most frequently coded concept within algorithms and programming was the concept of algorithms, often introduced through practical examples. For instance, Teacher No. 4 described how students created and performed dance routines using sequenced picture cards: “First, students are introduced to the concept of an algorithm and provided examples from everyday life. They watch the teacher perform a simple dance routine of basic movements. Students imitate the teacher’s dance using picture cards arranged in the correct sequence and a list of steps projected on the board. Then, in small groups, they create their own dance algorithms based on the projected model.”

Teachers commonly drew on everyday tasks, such as making tea, baking pancakes, brushing teeth, or doing laundry, to help students understand algorithms as sequences of steps to reach a goal. Algorithms were also developed within subject-specific contexts, including solving word problems in mathematics, constructing narratives in language lessons, and sequencing biological processes in science. Robotics further supported algorithms, as students created and modified basic programs consisting of step-by-step instructions for the robots to follow. Similarly, without robots, students composed simple instruction sequences, such as directional commands to navigate each other. As Teacher No. 45 noted: “Using arrows, students wrote a program on a blank sheet of paper representing a path from the starting point to a location marked on the map.”

Control was also frequently addressed. A commonly used approach involved students giving and following instructions, often on a grid. In several lessons, students issued directional commands (e.g., “five steps forward, two to the right”) to guide a classmate to a target, such as a picture or a word, and then switched roles. Teachers also described activities in which students followed verbal instructions to draw specific objects, with some noting the discrepancies between intended and actual outcomes. As Teacher No. 57 explained: “In the drawing activity, students learned how difficult it is to give clear instructions and how amusing the outcomes can be when the instructions are vague or imprecise.”

In digital contexts, control structures, such as conditionals and loops, were introduced through the programming environment Scratch. For instance, Teacher No. 128 reported: “Using conditional statements and loops in Scratch, the children programmed the movement of a character in accordance with traffic rules.”

Furthermore, programming and program development were most frequently mentioned in connection with Scratch and robotic kits. While most teachers did not elaborate on whether the concept of program development was explicitly explained to students, many used expressions such as “the children programmed,” as seen in the previous example from Teacher No. 128. However, Teacher No. 43 provided a more detailed account, describing how students moved “from the idea to the program,” writing step-by-step instructions for constructing a traditional Kurent mask, which were then exchanged and followed by other groups.

Notably, no teacher explicitly referred to the use of variables, and only one mentioned modularity, indicating limited emphasis or awareness of these two concepts in the reported classroom activities.

Networks and the Internet

All teachers who addressed networks and the internet content area focused specifically on cybersecurity, with no references to network communication or organisation. Activities commonly involved basic cryptographic systems, and frequently used methods included Morse code, Caesar cypher, Huffman trees, and symbol substitution. For instance, teacher No. 14 described: “I placed encrypted words around the classroom. Students tried to guess their meaning, then used a worksheet to decrypt them using symbols. The decoded message revealed the location of a hidden treasure.”

Impacts of Computing

Teacher reflections primarily emphasised online safety and responsible digital behaviour. While teachers frequently mentioned collaboration and teamwork among students, they did not clarify whether these interactions were contextualised within broader discussions about how computing enables new forms of social connection.

In the context of online safety, teachers reported discussing the importance of using strong passwords and protecting personal data, including names and family photos. Some extended these conversations beyond the classroom by involving parents. As Teacher No. 62 described: “We invited parents to participate in a digital safety activity lesson. They received a link to a microlearning resource about sharing children’s photos and videos, reposting content, and managing screen time.”

In addition to online safety, some teachers addressed broader aspects of digital literacy. For example, Teacher No. 41 showed students AI-generated images containing subtle mistakes, which led to a discussion about misinformation and the importance of critically evaluating online content.

However, some concepts, such as culture and digital citizenship, were missing from the implemented lessons. Students did not learn about the legal and ethical issues that shape computing practices, how computing influences culture, and how culture affects engagement and access to computing. These gaps will be addressed in the project’s second year.

Computing Systems

Teachers presented various activities within the computing systems content area, and their frequency was comparable to other core content areas. However, these activities often focused on specific concepts, providing less comprehensive coverage of how devices function, including software and hardware components and their interconnections.

Some teachers’ reflections touched on the aspect of troubleshooting. For example, Teacher No. 50 described an interdisciplinary activity in physical education: “In PE class, we pretended our bodies were computers. We ‘turned them on’ with an imaginary button, ‘loaded programs’ for walking, running, and jumping, and practised natural movements. We debugged the program to ensure that the movements matched the expected ones. In the final game, students created sequences using image cards and numbers, and classmates tested and corrected them.” While this activity could also be seen as debugging within the algorithms and programming domain, we coded it as troubleshooting because it extends beyond algorithmic correction to encompass broader aspects of how devices and systems operate.

Several teachers explained how information is represented digitally with pixel-based tasks, such as colouring grid cells to form images: “Once coloured, the pieces were cut out and assembled into a collective image. We discussed how the picture looked from close and far away (Teacher No. 124)”. Others introduced binary and ASCII systems. For example, through encoding exercises: “We introduced binary and ASCII codes by having students find the binary version of their names and those of family members (Teacher No. 28).” In higher grades, few teachers introduced more advanced concepts such as ‘deadlock’. As Teacher No. 145 explained: “During a school event, students participated in a game that simulated a ‘deadlock’, a situation where no process can move forward because everyone is waiting on someone else. Students had different goals that could only be completed through strategic collaboration. Parents were also involved as guests, while students took on the role of explaining the rules and background of the task.”

Data and Analysis

This content area includes how data is collected, represented, transformed, and used to make inferences or decisions. For example: “Students collected data about their favourite and least favourite sports and presented the results using a bar chart (Teacher No. 140).”

Some teachers also incorporated optimisation problems into their teaching. As reported, Teacher No. 15 explained: “Students were given a task with multiple valid solutions and had to find the optimal one. They were introduced to a graph as an abstract way of representing data.”

Other examples included teaching how data is stored on computers, extending sequences based on patterns, and making inferences from given data or information. Overall, all concepts from this content area appeared in the reflections, although they were mentioned relatively infrequently.

In general, the qualitative coding supports the quantitative responses on the Likert scale, with algorithms and programming emerging as the most frequently addressed content area. Other areas were covered less extensively, and several concepts from different content areas—such as network communication, variables, ethics, and culture—were absent from the implemented activities.

4.1.2. Cognitive and Thinking Skills

Teachers frequently mentioned cognitive and thinking skills development in their reflections; however, detailed justifications or concrete examples were often limited. Teachers frequently mentioned that students developed problem-solving skills, CT, and spatial reasoning. Less often, they noted creative thinking and reading literacy (see

Table 5. Emergence of cognitive and thinking skills during lessons.

Category	Codes	Frequency
Problem solving	problem solving, logical thinking	34
Computational thinking	computational thinking, algorithmic thinking	24
Spatial reasoning	spatial reasoning, orientation	18
Creative thinking	creativity	4
Reading literacy	reading literacy, vocabulary development	2

). Given the overlap between the CSI concepts and components of CT—specifically, the concept of an algorithm and algorithmic thinking—we coded teacher reflections according to their focus. If a reflection addressed the concept itself, we coded it under concepts; if it addressed the thinking dimension, we coded it under practices. When excerpts included both aspects, we coded them in both categories.

Teachers frequently referred to problem solving and logical reasoning, typically related to structured tasks, games, or simulations that required strategic thinking. CT was also frequently mentioned, though often in general terms, with limited elaboration on specific components such as decomposition, abstraction, or pattern recognition. For example, one reflection described students “baking bagels and setting the table for dinner following a given procedure (Teacher no. 97)” as an activity that developed CT. Although the task aligns with algorithmic thinking, the teacher did not specify which CT components were addressed.

Further, spatial reasoning emerged more often than anticipated, despite not being initially expected. It was commonly associated with navigation tasks, grid-based orientation, and the use of robots. Finally, fewer references were made to enhancing creative thinking and reading literacy.

4.1.3. School Subjects and Curriculum Areas

Teachers most frequently integrated CSI concepts into STEM subjects, particularly mathematics, followed by social sciences and languages (see

Table 6. Integration of CSI content into school subjects and curriculum areas.

Category	Codes	Frequency
STEM	mathematics, natural Science and technology	35
Social sciences	environmental studies, social studies	25
Languages	English, Slovenian language	17
Physical education	physical education, dance	14
Art education	music education, visual arts	9

). Integration into physical education and art education was also notable but less common.

A deeper analysis of teacher reflections revealed that interdisciplinary integration of CSI occurred in three main ways: (1) embedding CSI within disciplinary instruction, (2) using subject matter as a contextual framework, and (3) reinforcing subject-specific knowledge.

In many cases, CSI was embedded directly into the core disciplinary task. For example, Teacher No. 39 explained how algorithms were central across three lessons in different subjects: “In math, students created algorithms to solve word problems. In language class, they developed story algorithms. In environmental studies, they used algorithms to model the development of living organisms and daily routines.” In these examples, algorithms were the central concept being taught, with the subject tasks serving as vehicles to practice the concept.

Subject matter also often served as a thematic context for introducing computing concepts. This was particularly common in social sciences, where computing was used to explore local culture and history. As Teacher No. 141 described: “We created Scratch animations featuring local landmarks such as Ptuj Castle and the Drava River. Students acted as instructors and demonstrated Scratch to their peers while developing storytelling and programming skills.”

In other instances, CSI was used to reinforce or practice core subject knowledge. For example, Teacher No. 38 used the Photon robot to support written division, with students programming step-by-step instructions that mirrored the mathematical procedure. In a Slovenian language lesson, a movement-based activity was designed where students practised antonyms while navigating a floor grid using algorithmic commands (Teacher No. 126). Similarly, in English lessons, vocabulary development was supported through a coding-based maze game (Teacher No. 111).

4.1.4. Interfaces and Modalities

Regarding interfaces and modalities, unplugged approaches emerged as the predominant mode of instruction for introducing CSI concepts in the primary classroom context (see

Table 7. Frequency of Instructional Modalities and Interfaces in CSI Activities.

Category	Codes	Frequency
CSI unplugged	movement, paper-based task, game, floor grid, unplugged	121
Robotic kits	Lego robotics, Photon, Vex123, robotic kit	13
Online block-based programming	Scratch, Blockly games	9
Microcontroller-based computing	Micro:bit	1

). In many cases, teachers explicitly mentioned unplugged methods; in others, their use was evident from the context. These included movement-based tasks, paper-and-pencil games, and logic puzzles. Robotic kits, including LEGO Education sets, Photon, and VEX123, were used less frequently.

In one example, students used the VEX123 robot in a custom-designed board game that combined movement through a maze with visual arts-related questions (Teacher No. 79). In another case, students assembled and programmed LEGO robots, modifying basic algorithms to explore different movement outcomes (Teacher No. 138). Some teachers employed online block-based environments such as Scratch and Blockly to introduce programming concepts in a visual format. Microcontroller-based computing, such as the use of Micro:bit (Teacher No. 6), was mentioned only once across all reflections.

4.2. Students’ Experience of CSI Activities

In addition to teacher reflections on lesson implementation, data were collected on students’ experiences with CSI activities. It should be noted that the number of responses in this section exceeds that of the previous subsection by three, as some teachers provided input on students’ experiences even if they had not completed the earlier question about the content area. Consequently, teacher identification numbers may not align consistently across both sections.

The analysis is organised into two categories: (1) affective and emotional response, and (2) barriers and difficulties encountered during the activities.

4.2.1. Students’ Affective and Emotional Responses to CSI Activities

As shown in

Table 8. Teachers’ Ratings of Students’ Affective and Emotional Responses to CSI Activities.

Dimension	N	Mean	Std. Deviation	Median	IQR
Motivation	151	4.71	0.50	5	1
Engagement	151	4.75	0.53	5	0
Enjoyment	151	4.62	0.61	5	1
Curiosity	151	4.54	0.67	5	1
Enthusiasm	149	4.54	0.64	5	1
Surprise	149	4.11	0.91	4	2

, teachers reported high levels of student engagement (M = 4.75, SD = 0.53) and motivation (M = 4.71, SD = 0.50). Enjoyment (M = 4.62, SD = 0.61), curiosity (M = 4.54, SD = 0.67), and enthusiasm (M = 4.54, SD = 0.64) were also highly rated. The average rating for surprise was slightly lower (M = 4.11, SD = 0.91), indicating a more varied response.

We conducted Spearman correlations between the six dimensions. Results showed that all dimensions were significantly correlated (p < 0.01), indicating that students who were more engaged also tended to be more motivated, enthusiastic, curious, and to experience enjoyment and surprise. Similarly, higher scores on any one dimension were positively associated with higher scores on the others.

To complement the quantitative data, qualitative coding was conducted to identify reflections related to students’ affective and emotional responses. As shown in

Table 9. Frequencies of Codes Related to Student Affective and Emotional Response in Teacher Reflections.

Code	Frequency
Engagement	35
Motivation	35
Enjoyment	25
Peer support and collaboration	24
Enthusiasm	19
Interest	18
Independence	7
Satisfaction	6
Curiosity	5

, the most frequently mentioned codes were engagement (N = 35) and motivation (N = 35), followed by enjoyment (N = 25), peer support and collaboration (N = 24), and enthusiasm (N = 19). Other affective responses, such as interest (N = 18), independence (N = 7), satisfaction (N = 6), and curiosity (N = 5), were also noted, though less frequently.

A comparison of the quantitative and qualitative data reveals an alignment in several areas. Engagement and motivation were both rated highest in the Likert-scale responses. They were also the most frequently mentioned in qualitative reflections (N = 35 for each), suggesting consistent perceptions of students being actively involved in the lessons. Enjoyment and enthusiasm also matched high average ratings and notable frequencies in qualitative coding.

However, despite receiving a relatively high average score (M = 4.54), curiosity was mentioned less often (N = 5), indicating that while students may have experienced curiosity, it was not consistently highlighted in teacher reflections. On the other hand, interest, peer collaboration, independence, and satisfaction appeared solely in the qualitative data, suggesting the Likert scale may have underrepresented specific social and emotional dynamics.

Teachers consistently reported that students were motivated, engaged, and enjoyed the activities; however, these responses were largely observational rather than deeply reflective. For example, one teacher noted that “the students were very motivated and had fun while learning through play” (Teacher No. 8), while another observed that students “quickly figured out the system and were excited by the tasks” (Teacher No. 25). Other teachers highlighted students’ independence and enthusiasm, noting that they completed tasks without additional explanation and found the activities both fun and educational (Teachers No. 48, 84, 119).

Overall, the data indicate that CSI activities fostered student engagement, motivation, and enjoyment, with qualitative evidence pointing to positive emotional and social dynamics within the classroom.

4.2.2. Barriers and Difficulties

Teachers also reflected on potential barriers and challenges observed during CSI activities. Quantitative ratings indicate that exclusion of certain students was rated very low (M = 1.64, SD = 1.00), suggesting that most students could participate. Similarly, stressfulness (M = 2.08, SD = 1.13) and dominance of individual students (M = 2.61, SD = 1.21) were rated low, indicating that these issues were not prevalent (see

Table 10. Students’ Experience of Exclusion, Dominance, and Stress During CSI Activities.

Dimension	N	Mean	Std. Deviation	Median	IQR
Exclusion of certain students	151	1.64	1.00	1	1
Dominance of individual students	151	2.61	1.21	3	1
Stressfulness	150	2.08	1.13	2	2

).

We conducted Spearman correlations between the three dimensions. Results showed that all three were significantly correlated (p < 0.01), indicating that when students experienced more exclusion, they also reported higher levels of stress and dominance of individual students.

Open-ended reflections provided a more detailed insight into students’ challenges (see

Table 11. Qualitative Coding of Reported Barriers and Difficulties During CSI Activities.

Code	Frequency
Difficulties in understanding and following instructions	16
Difficulties in task solving	12
Lack of time	6
Lack of concentration	4
Lack of interest	1
Impulsiveness	1
Difficulty expressing oneself	1

). Teachers noted students had difficulties in understanding instructions (N = 16) and solving tasks (N = 12). Teachers also mentioned issues such as lack of time (N = 6) and limited concentration (N = 4). Isolated low interest, impulsiveness, and difficulty expressing oneself (each N = 1) were also noted.

Teachers reported that students encountered various difficulties, most often misunderstanding or following instructions—meaning students had trouble comprehending, remembering, or correctly executing the given instructions, and difficulties in solving tasks—meaning students had trouble working through the problem itself or grasping the underlying concept, even when they understood the instructions. For example, Teacher No. 143 noted that students did not know how to colour the squares (pixels). When learning control concepts, some students gave instructions too quickly or provided imprecise directions (Teacher No. 97). Two teachers observed that students struggled with spacing in letters during cryptography exercises. Teacher No. 28 also highlighted that students initially struggled to understand the connection between ASCII and binary code. These challenges were most often reported for 1st or 2nd-grade students.

Some teachers attributed students’ difficulties to reading challenges, which hindered their understanding of the tasks. Others noted that students did not carefully read or listen to instructions. For instance, Teacher No. 88 remarked, “Instead of giving instructions, they just took hold of the student and tried to guide them.” Teachers further observed that even after a task was demonstrated, some students were unsure how to begin. Further analysis indicated that the greatest challenges appeared in cryptography, control, and binary systems. Additionally, some teachers reported time constraints in preparing and implementing project activities.

No significant correlations were found (all p > 0.05) between any of the negative dimensions (exclusion, dominance, stress) and any of the positive dimensions (engagement, motivation, enjoyment, curiosity, enthusiasm, surprise), suggesting that students’ positive affective responses were largely independent of the presence of challenges.

In addition, Mann–Whitney U tests were conducted to examine potential differences between younger students (Grades 1–2) and older students (Grades 3–5) across all nine dimensions. No statistically significant differences were found (all p > 0.05); levels of motivation, engagement, enjoyment, curiosity, enthusiasm, and surprise, as well as perceptions of exclusion, dominance, and stressfulness, were comparable across grade levels. However, open-ended responses revealed that teachers reported challenges with younger students more frequently.

In summary, the analysis of student experiences indicates a positive response to CSI activities across affective and emotional domains. While a limited number of challenges were identified, primarily related to individual differences in attention, understanding, or task complexity, these were infrequent and did not appear to compromise the overall accessibility or educational value of the activities.

5. Discussion

This study examined two central questions: (1) which core content areas from the K–12 Computer Science Framework were most and least frequently addressed by primary school teachers in the UTRINKI project, and (2) how students in the UTRINKI project engaged with and responded to CSI activities based on teachers’ reflections.

5.1. Interpreting the Patterns in the Integration of CSI in Primary Education

In this section, we discuss the CSI concepts and practices introduced in the lessons, the subject areas in which the CSI concepts were integrated, and the modality of the CSI activities.

5.1.1. Addressing Core CSI Content Areas

Participants most frequently addressed algorithms and programming, confirming that algorithms are central not only in curriculum design [22,65] but also in classroom practice, as reflected in our findings. To understand the concept of algorithms, teachers often incorporate algorithms from everyday life in their practice, such as brushing teeth or making pancakes. They also included tasks such as following a sequence of steps on the grids, with students often taking turns to control or navigate each other’s movements.

In contrast, the computing systems, impacts of computing, networks and the internet, and data and analysis were addressed less often in ratings and reflections. This may be due to an unbalanced PD programme that placed greater emphasis on the algorithms and programming domain. Although teachers were provided with activities across all five core content areas, the two-day workshop focused primarily on unplugged methods, with algorithms being the most prevalent subdomain. Teachers may also perceive algorithms as more concrete and age-appropriate for primary school students, as algorithms can be introduced through familiar, everyday tasks such as sequencing routines or designing simple step-by-step instructions [66]. By contrast, concepts from other domains—such as software and hardware systems, networking, or computing ethics—are more abstract and often require specific tools, or guided discussion that teachers may feel less prepared or resourced to facilitate. Moreover, unplugged algorithmic activities demand minimal technological infrastructure, making them easier to implement across diverse classroom contexts. Together, these factors likely contributed to the predominance of algorithm-focused teaching observed in our data.

Although teachers rated data and analysis as frequently addressed, these concepts were seldom elaborated upon in their reflections, suggesting possible misconceptions about data-related instruction and a potential overestimation in self-reported data. Spearman correlations showed that data and analysis was the only domain significantly associated with algorithms and programming, which may reflect the natural connection between these domains—a finding partially supported by additional reflexive coding. Therefore, the brevity of reflections and the subjective nature of coding should also be considered when interpreting the results.

Given the growing importance of CSI education in fostering digital citizenship [17,18], more attention should be given to domains such as data ethics and online communication, which remain insufficiently addressed despite their relevance to students’ everyday digital experiences. Equity in CSI education [29,30] has also recently gained increased attention, yet our PD programme did not explicitly address these issues. Ensuring that CSI activities are inclusive is critical, as research shows that addressing systemic barriers, such as racial and ethnic inequities, can shape students’ engagement, confidence, and future interest in computing [31]. Future work in Slovenia should explore ethnic and other dimensions (e.g., gender gaps, socio-economic disparities, and rural access to resources) once a solid foundation for CSI education has been established.

5.1.2. Emergence of Cognitive and Thinking Skills

The Computer Science Framework [22] distinguishes between concepts—such as algorithms, variables, and control—and computational practices, which are the behaviours and thinking skills that emerge while students are computing. Regarding these practices, teachers most frequently highlighted the development of students’ problem-solving and CT, which are at the heart of computational practices [22], followed by spatial reasoning. In contrast, creative thinking and reading literacy were only marginally mentioned. The latter three are not recognised as computational practices [22] but emerged in the context of the activities.

CSI activities provide a natural context for the development of CT [11,42,67], which is regarded as a foundational tool for introducing problem-solving and cognitive processes [68,69]. Unexpectedly frequent mentions of spatial reasoning, often associated with navigation tasks and robotic activities, suggest that lessons may provide a natural context for developing spatial skills. This aligns with emerging research that includes spatial reasoning as one of the emerging components of CT [70,71], although the area remains insufficiently explored in current studies. While teachers frequently referred to CT and problem solving, their reflections rarely explored how these skills were fostered during instruction, suggesting that deeper pedagogical reflection on CSI practices should be encouraged in the second year of our PD.

5.1.3. Integrating CSI in School Subjects

Interdisciplinary integration of CSI occurred in (1) embedding CSI within disciplinary instruction, (2) using subject matter as a contextual framework, and (3) reinforcing subject-specific knowledge. CSI content was primarily integrated into STEM subjects. For example, in science, students designed an algorithm to seed plants, or in mathematics, students practised written division with the Photon robot, programming step-by-step instructions that mirrored the mathematical procedure. This trend is consistent with prior research showing that CSI is mainly incorporated within STEM subjects, where it aligns most readily with the subject matter and instructional practices [11,45,72]. Teachers also explored connections beyond STEM, embedding CSI in physical education, the arts, and language learning, meaning other areas can also benefit from applying and recognising CSI concepts [52,73]. Examples such as dance algorithms or navigating antonyms on a floor grid illustrate this versatility. One teacher, for instance, used Scratch to animate a local story, combining narrative sequencing with basic programming.

Such findings support broader calls to integrate computational concepts and practices across subject areas [11,43,53,72], thereby reducing cognitive load and easing implementation for in-service primary teachers in comparison with implementing a stand-alone subject [54]. Given the challenges of meeting today’s curricular demands, connecting computational ideas to what teachers already do in their classrooms may be the best approach [24].

5.1.4. Modality of the Activities

Unplugged activities dominate lesson implementation when teachers are given free choice of modality [55], a pattern also evident in our study. The frequent use of movement-based games, paper tasks, and floor grids reflects a preference for embodied, low-tech formats that support the basis for cooperation and teamwork [14]. One of the key advantages of the unplugged method is the active engagement of students. These activities are often game-based [74] and designed to encourage physical interaction and collaboration, motivating children to solve problems together [75]. This pattern was also prevalent in our study and is further discussed in the following section.

Reflections rarely mentioned modularity or variable use, suggesting that instruction sometimes remained surface-level even when programming tools were used. However, the prevalent use of unplugged methods alongside algorithms suggests that the interface type may influence how CSI content is taught and which practices are prioritised. These patterns indicate the need for PD initiatives that move beyond familiar formats, helping teachers connect diverse tools with broader conceptual and cognitive goals [76].

The limited proportion of plugged activities, such as Scratch, aligns with the commonly expressed reluctance to introduce screens in the early years of education [55,77]. However, to promote balanced development, PD should guide teachers from unplugged activities toward richer, technology-enhanced environments. Previous research shows that by beginning with unplugged activities and then moving to plugged, teachers had the opportunity to gradually increase their comfort with CSI integration [54]. Engaging with plugged approaches is particularly important for understanding automation processes and the capabilities of computers as tools. Consequently, incorporating plugged activities in higher grades is essential for students to fully grasp the core concepts and practices of computing [78,79].

5.2. Reflections on Students’ Experience of CSI Activities

Teachers observed that students were motivated, engaged, curious, enthusiastic, and enjoyed CSI activities. Those findings are consistent with the findings of Ozturk et al. [43], who noted that during the implementation of CSI in primary education, students were engaged in project activities, which also provided opportunities for peer collaboration.

Teachers in our study noted that some challenges, particularly task complexity, attention span, and difficulties in following instructions, were more common among younger learners and in the more technical concepts, such as cryptography and binary systems. Teachers also mentioned occasional structural barriers, including a lack of time and insufficient supporting instruction, which may be linked to differences in prior experience and available resources.

Similar issues are reflected in previous research, where teachers cited limited content knowledge and a lack of time in planning and implementing project activities [43,48]. Despite these challenges, most students in our study were able to participate, with teachers reporting low levels of exclusion, stress, or dominance by individual students. This suggests that the activities were generally accessible and appropriate for the primary level, supporting inclusive participation.

To address the reported barriers and further strengthen classroom implementation, future efforts should include sustained, hands-on PD that is aligned with teachers’ prior knowledge, contexts, and needs [48,80].

6. Limitations and Future Research

While the findings provide preliminary insights into how a structured PD programme for teachers can support interdisciplinary approaches in CSI education, several limitations should be acknowledged. First, the findings rely primarily on self-reported teacher reflections, which are inherently subjective and may overstate student engagement or the depth of CSI content integration. Although triangulated with quantitative data, the qualitative insights—derived through inductive coding and thematic analysis—remain shaped by participants’ reporting styles and researchers’ interpretations. Notably, the study lacks direct assessment of student learning and engagement, such as classroom observations and pre-/post-test designs, which would offer more robust validation of lesson effectiveness.

The pilot design did not include baseline measures or a comparison group (e.g., non-PD teachers), which limits inferences about the impact of the PD programme. Additionally, the sample—47 volunteer teachers from seven schools across two Slovenian regions—introduces potential self-selection bias, as these teachers are likely highly motivated and received substantial content and pedagogical support. Participating schools also received additional funding and equipment, which may have influenced outcomes. The absence of demographic data, such as teaching experience and school socioeconomic status, further limits the ability to assess sample representativeness. Together, these factors restrict the generalisability of the findings to the broader population of Slovenian primary teachers and to international contexts.

Future research should address these limitations by incorporating objective measures such as classroom observations, student work analysis, and longitudinal tracking of learning outcomes to better understand the depth and sustainability of CSI integration. Collecting teacher data (e.g., years of experience) would enable more nuanced analyses of how contextual factors influence implementation. Future studies with larger and more diverse samples are needed to confirm and extend these findings. Comparative studies across national contexts could also shed light on the role of systemic and cultural factors. Finally, given the underrepresentation of specific content areas (e.g., data and analysis, and impacts of computing), future PD initiatives should explore ways to better support teachers in implementing more abstract and complex dimensions of CSI education.

7. Conclusions

The findings underscore the importance of sustained, hands-on PD in supporting teachers in translating CSI concepts into interdisciplinary classroom practice. By offering structured support and practical resources, the PD programme enabled teachers to implement interdisciplinary lessons. The diverse ways teachers incorporated CSI into subjects beyond STEM, including language, art, and physical education, illustrate the potential of cross-curricular approaches to make CSI concepts and practices more accessible and relevant. These insights contribute to the broader European discourse on CSI as a foundational discipline that can be embedded across primary-level subjects.

A novel contribution of this study is its focus on which core CSI content areas teachers choose to integrate when given the freedom to select the topics within the framework. Our analysis reveals clear patterns of instructional preference: algorithms and programming dominate, while computing systems, networks and the internet, impacts of computing, and data and analysis remain underrepresented. These blind spots likely reflect both the PD programme’s emphasis on algorithms and teachers’ greater confidence in teaching them. Recognising these comfort zones provides practical guidance for future PD design: programmes must balance teacher autonomy with targeted scaffolding to strengthen less frequently addressed domains and ensure more comprehensive classroom integration.

Furthermore, teachers in our study reflected on student emotional responses. Teachers reported that students were motivated, engaged, and enthusiastic during CSI activities. These findings help address a gap in the literature, as this area of CSI integration has been underexplored in previous research [21].

While this study focused on implementation within the Slovenian context, the findings align with broader international efforts to embed CSI principles into national curricula, providing all learners opportunities to develop CSI knowledge and practices, thereby promoting quality, inclusive education and social sustainability.