Integrating computational thinking into K-12 education: Bridging the gap between theories and practices

Over-investigation of tools

One of the most common and important gaps between the theory and practices for the integration of CT into K-12 STEM classrooms is the overemphasis on tool usage for both teachers and students. For instance, many professional development programs focus primarily on the teaching of tools, without sufficient emphasis on the underlying concepts of CT.^[6] Similarly, many school CT initiatives only work with students to learn how to use various computing devices.^[26] This approach could lead to the misinterpretation of CT as merely a set of skills that can be acquired and applied to specific tools, rather than a deeper understanding of CT as a problem-solving approach.

However, the underlying challenge is that although the theory asserts that the tool is not central to CT as thinking skills, it is crucial to recognize that the ability to use tools is an essential initial step in learning CT.^[12,19] Tools such as blocked-based programming, syntax-based programming, or unplugged computing tools should only serve as a means of expressing CT concepts in a tangible and practical way. For example, Saxena et al. demonstrated that even young children at the age of 4–6, could learn basic pattern recognition and algorithmic design by engaging with computing tools.^[27] By engaging with programming languages, students can learn to think algorithmically, break down complex problems into smaller parts, and develop logical reasoning skills. To achieve that, sufficient learning experiences with the tools per so is inevitable.

As a result, the practice of CT needs to start, but also often ends with tools even if the tools should not be the sole focus of instruction. In the literature on teacher professional development in CT, a considerable amount of learning experiences is found to increase teachers’ confidence and the adoption of computing tools.^[6] However, even for professional development opportunities that focus on moving teachers beyond the scope of tool adoption, many teachers need to spend a significant amount of time at the beginning of professional learning to familiarize themselves with the tools before starting to develop an understanding of CT. For instance, Kayumova et al. explored the 16 teachers’ CT professional development engagement and reluctance. They have identified that at the beginning of professional learning, teachers are most challenged by new tools, such as learning to code and setting up devices. Almost all teachers had to experience technical and logistics barriers because they felt they needed to first figure out how to use the tools and develop the ability to solve technical problems students might encounter in the classroom.^[28] This pattern echoes the research in promoting teachers’ technology-integrated practices: Teachers are often not confident with the tools to begin with, which tremendously prevents them from exploring further the deeper implications of CT.^[29]

The integration of CT into K-12 STEM classrooms requires a balance between teaching the use of tools and developing a deeper understanding of the underlying CT concepts. Professional development programs should strive to emphasize the conceptual understanding of CT, rather than the mere use of tools. By doing so, teachers can better facilitate the integration of CT into their classrooms and provide students with a solid foundation in CT skills.

Trivialization and essentializing of CT

As described previously, CT is not a clearly defined subject, and the integrated view of CT may have different levels from a theoretical perspective. However, such theoretical complication is not accessible to practitioners, which often results in problems of trivialization and essentialization (see discussions around culturally responsive pedagogy^[30]). While these mindsets may represent an intermediate stage of CT learning, they create a significant gap between theory and practice.

Trivialization occurs when CT is reduced to a set of technical skills and concepts related to how to use computing tools, without considering its broader implications and applications. For instance, some teachers may focus solely on teaching coding syntax and manipulation of computing tools without exploring the creative and problem-solving aspects of CT. This approach may lead students to view CT as a mere tool for building apps or games, rather than a way of thinking and reasoning about complex problems. Only experiencing CT learning within one single context (e.g., games) may significantly limit students’ ability to transfer the learned skills to other tasks.^[31]

Essentialization, on the other hand, refers to the oversimplification and stereotyping of CT, assuming that CT skills, essentially, are just a simple combination of cognitive skills such as algorithmic thinking, abstraction, and pattern recognition.^[2] For example, participants in professional development may demonstrate engagement patterns of paraphrasing and referring to keywords of the CT concepts and it is often an oversimplification of the content.^[32] This essentialization is sometimes also encouraged by the instructors. For instance, in one of the professional development efforts I am involved in, teachers often were confused with the idea of abstraction in CT and asked for clarification. The facilitator could not provide a straightforward answer and had to essentialize abstraction as identifying the most important thing. As a result, the teachers interpreted understanding the problem components, which is problem decomposition, as abstraction as well. Moreover, teachers implemented the misconception with all student levels assuming that CT should be taught in the same way across all subjects and grade levels. This approach may lead to a lack of attention to the contexts of learning, such as students’ backgrounds, interests, and learning needs, and may reinforce stereotypes and biases related to gender, race, and socioeconomic status.^[1]

It is important to recognize that trivialization and essentialization are not necessarily permanent mindsets, but rather an intermediate stage of CT learning. Longitudinal evidence from both teachers’ CT learning and students’ CT learning revealed that they can move beyond these limiting perspectives and develop a more nuanced and comprehensive understanding of CT.^[9,33] However, if we do not address these issues and move beyond this intermediate stage of CT learning, there are potential harms. Trivialization may lead to a lack of creativity and innovation in CT applications, limiting students’ ability to solve complex problems in different contexts. Essentialization may perpetuate stereotypes and biases, leading to inequitable access and outcomes in CT education. Therefore, it is crucial to provide teachers and students with the necessary tools and resources to develop a deeper and more meaningful understanding of CT, one that goes beyond technical skills and concepts and embraces the broader implications and applications of CT thinking.

One-size-fits-all solution to CT learning

Another significant issue in the integration of CT into K-12 STEM classrooms is the gap between the theory that CT may look different in each context and the practice of designing a one-size-fits-all tool or curriculum. CT is not a fixed set of skills or concepts that can be easily taught or learned. Rather, it is a flexible problem-solving approach that can be applied in diverse contexts and for different purposes.^[1] However, in many cases, both teachers and students ask for simplified curricula or tools that can be easily taught or learned. Pre-packaged and pre-scripted curricula are also easy to promote.^[34] All these reasons can lead to a misinterpretation of CT as a set of fixed skills that can be acquired and applied in a particular context.

The situated framing of learning implies that CT may look different in each context, depending on the context, scope, and purpose because of the uniqueness of each community of practice.^[35] For example, CT in students’ STEM learning would look different from CT in a computer science class, as the shared goals of two different communities may differ. Weintrop et al. developed a taxonomy for CT practices in STEM learning and identified the commonly shared practices between CT and STEM learning.^[23] Although the taxonomy is decontextualized, it provides the teachers a great tool to recognize CT practices within their context and implement CT in their classrooms in a flexible way depending on the needs. The integration of CT into K-12 STEM classrooms requires a flexible and adaptable approach to CT instruction, tailored to the specific context and audience. CT is not a fixed set of skills or concepts that can be easily taught or learned, but a flexible problem-solving approach that can be applied in diverse contexts and for different purposes. Therefore, designing CT curricula that are tailored to the specific context and audience is essential to effectively bring CT into STEM classrooms.^[36]

However, the practice of designing CT curricula often ends up being a one-size-fits-all approach, which may start with recognizing the contextualized meaning of CT learning within local communities but then stop evolving once the resources are developed. Many curricula and tools are designed to be easily taught or learned, with a focus on specific skills or concepts, rather than the broader problem-solving approach of CT.^[9,37] This gap between theory and practice in CT curricula can lead to a persistence challenge in effectively bringing CT into STEM classrooms.

Short-term intervention

The final gap I will discuss is that many programs about CT integration are short-term initiatives while impactful system change needs sustainable engagement. The lack of sustainable models for building a learning community for students and teachers is a significant gap in integrating CT into K-12 education. While various initiatives have been implemented to integrate CT into K-12 education, they have been short-term interventions with limited lasting impact.^[6] The issue is not unique to CT integration,^[38] but it is particularly important in this case, as CT also interconnects with many other important disciplines.

One-shot learning experiences for teachers and students are common in K-12 STEM classrooms. This approach involves a single session or workshop in which teachers or students are introduced to CT concepts and tools. While these one-shot learning experiences can be beneficial in raising awareness of CT and introducing basic concepts, they are often not effective in developing a deep and meaningful understanding of CT.

At the teachers’ end, many professional development are organized in the form of one-day workshops or summer institutes.^[6] Although these short professional development opportunities are, arguably, the most practical and straightforward way of introducing new ideas to teachers, past studies on teacher professional development have already illustrated that this one-shot model is not effective in bringing system changes.^[38] Desimone outlined the levels of impact on teachers’ professional development and urged the research community to move away from immediate evaluations of professional development outcomes.^[39] In many cases, teachers may not have enough time to fully understand the implications of integrating CT into their teaching practices, resulting in surface-level implementation. Additionally, teachers may not have the necessary support to continue integrating CT into their teaching practices after the short-term intervention ends.^[40] Without sustained engagement and support, teachers may not be able to fully integrate CT into their teaching practices, resulting in a limited impact on student learning.

The same problem also reflects on the students’ end: Students often experience the CT integration learning experience as a temporary introduction to the tools and some big ideas. In very few cases, we see students’ longitudinal engagement with CT learning and STEM integration.^[33] As a result, oftentimes, the short-term effect of CT learning or students’ enthusiasm can be confounded with the novelty effect. And once the intervention from the research team is removed, the computing tools will stay on the shelf and gather dust.^[41] While they may be enthusiastic about learning CT at first, this initial excitement may not be sustained without ongoing engagement and support. In some cases, students may not have the opportunity to continue practicing CT skills after the short-term intervention ends. As a result, the impact of CT learning on student learning may be limited.

The lack of a sustainable model for building a learning community for both teachers and students in integrating CT with STEM highlights the need for a long-term approach to CT integration. This approach should aim to create a culture of CT and STEM integration in schools, which will help to sustain CT practices in STEM classrooms. Teachers and students need ongoing engagement and support to fully integrate CT into their teaching and learning practices. There is a need for a long-term approach that creates a culture of CT and STEM integration in schools, which will help to sustain CT practices in STEM classrooms. Without such an approach, the impact of CT integration on student learning will remain limited.

This gap between commonly practiced one-shot learning opportunities and the prolonged engagement of CT learning can lead to a persistent challenge in effectively bringing CT into K-12 STEM classrooms. Teachers and students may not develop a deep and meaningful understanding of CT, which can limit their ability to use CT skills to tackle real-world problems and contexts.^[6]

Interconnections in the gaps and the conceptual model

Before discussing the recommendations, it is worth noting that the gaps discussed in this article are also interconnected. When the focus is heavily placed on learning to use CT tools like block-based programming languages, the deeper conceptual understanding of CT tends to get overlooked. Teachers and students start viewing CT as simplistic skills and procedures for manipulating these tools, rather than as a complex problem-solving process, which could lead to trivialization and essentialization.^[6,29] When CT is reduced to step-by-step skills (i.e., trivialization and essentialization), it becomes easy to design generic curricula, tools, and strategies that would be uniformly applied across different contexts.^[1] However, this contradicts the situated nature of CT, which necessitates adapting CT integration based on the specific goals, cultures, and practices of each learning community. In turn, with a belief that a generic curriculum or tool can effectively teach CT to all students, there is less motivation to sustain long-term, customized integration tailored to local contexts.^[36] Short teacher workshops or student summer camps seem sufficient as ways to deliver standardized CT content to everyone. Just like the complicated nature of the gaps, the potential solutions to bridge the gap would also be multileveled.

To address the challenges brought situated in the interconnected gaps, the proposed conceptual model (Figure 1) draws on existing theories related to the design of effective professional learning, with a focus on practical implementation and collaborative participation.^[39,42] These theories provide insights into how to effectively bring about practice change and involve various stakeholders as a community, recognizing the long-term impact of such changes.^[43,44] Therefore, the model presents a multi-leveled and interconnected way to address the challenges. It discusses potential approaches at three levels: the conceptual level, the practice level, and the community level. These levels recognize that addressing the gaps requires a comprehensive and systematic solution that encompasses different aspects of CT integration in STEM classrooms. It is important to note that the recommendations provided within each level are not exhaustive or prescriptive. They are meant to inspire further exploration and serve as a starting point for researchers and practitioners to consider a systematic approach to bridging the gaps.

Building technology self-efficacy

While an over-emphasis on tools should be avoided, building basic technology self-efficacy is an important first step, especially for teachers and students new to CT. At the initial stages of engaging with new tools like block-based programming platforms, a sense of confidence in one’s ability to use the technology is needed before deeper CT skills can be developed.^[45] Teachers need to feel comfortable navigating and troubleshooting the tools before they can shift focus to the conceptual teaching of CT.^[46]

Similarly, students need a level of fluency with the technology in order to then apply it for creative problem solving versus getting stuck on syntax errors.^[47] Building self-efficacy can facilitate initial engagement with novel technologies when integrating CT. It is important to provide learners hands-on learning opportunities that enable them to explore CT concepts using computing tools. For instance, past research has demonstrated that partial pair programming could help students develop their CT and self-efficacy.^[48]

However, the end goal should be to use these tools as vehicles for CT skill development, not mastery of the tools themselves.^[18] Once basic fluency is achieved, the educational focus should move towards conceptual understanding and connecting CT across disciplines. Scaffolded, sustained engagement can gradually transition learners from tool fluency to deeper competencies.

Developing a nuanced understanding of CT

Given the diversity in definitions and endpoints of CT learning, it is important to engage learners in discussing and defining CT within their own contexts. This avoids essentialization or trivialization of CT as just a set of predefined practices. Teachers should be encouraged to explore pedagogical approaches that emphasize conceptual understanding of CT, rather than mere tool use. For example, Ketelhut et al. found that integrating CT into the curriculum through project-based learning helped students develop a deeper, more nuanced understanding of CT concepts and improved their pedagogical content knowledge around how to integrate CT into STEM learning.^[9]

Rather than treating CT as fixed skills to be acquired, teachers can engage students in defining what CT means in their learning community and empower them to apply CT as a way of thinking to solve authentic, situated problems.^[1] Learners can be guided to move beyond viewing CT as simplified cognitive skills and instead explore various views and conceptualizations of CT. With continued, contextualized engagement, the nuances of CT and its possibilities within different subjects can emerge over time. The goal is for both teachers and students to see CT as an evolving set of context-dependent practices, not a generic checklist of competencies.

In addition, learners could be encouraged to engage in reflective practice, which involves reflecting on their own experiences and beliefs about CT and how they can integrate CT into their teaching practices in a way that is meaningful and effective for their students. Researchers have demonstrated that reflective experiences could be an important way for teachers to develop nuanced CT understanding over time.^[9,10,40] This way, teachers can develop a more contextualized understanding of CT and avoid essentializing or trivializing CT concepts and practices.

Co-design and development approach

One of the ways to achieve a contextualized understanding and implementation of CT learning is to engage teachers through co-design and development activities. Teachers are not provided with pre-packaged curricula and the learning community works together collaboratively to design CT curricula and select tools that are appropriate for the local contexts, needs, and cultures. The co-design process emphasizes participatory practices where educators and learners are actively involved in creating customized CT learning experiences, rather than being passive recipients of pre-packaged, generic instructional materials.^[49]

Co-design activities that involve teachers in the design of CT curricula and activities can help to foster a deeper and more contextualized understanding of CT. Teachers could be encouraged to collaborate with CT experts, other educators, and students to design CT learning activities that are relevant, engaging, and accessible to all learners. This way, teachers can develop a better understanding of the diverse needs and interests of their students and design CT learning experiences that are tailored to the specific context and audience. For instance, researchers have developed co-design workshops with teachers. After the co-design experience, participants were able to integrate multiple computing tools that engage students in various CT-STEM practices.^[50]

Moving towards a sustainable learning model

Finally, CT learning needs to become a prolonged process both for teachers and students. The learning experience should be sustained by the entire community of practice, not necessarily just the teachers and the students, but also the administrators, information technology (IT) or pedagogical supports, and parents.^[40] Community engagement and commitment will be able to sustain the practice of CT learning. To that end, the gap also calls for a sustainable learning model that takes a community perspective. A sustainable learning model can help all the members of the learning community develop their CT knowledge and skills over time, integrate CT into their learning practices, and collaborate with other educators and stakeholders to improve CT learning.

The past literature on teacher professional development has highlighted the importance of engaging various stakeholders in the learning community. For instance, Simmonds et al. highlighted how support from administrators could be crucial in building a long-term CT program.^[51] In addition to the existing stakeholders, the CT learning community could engage learners outside of the existing community and move the existing members forward by transforming their roles.^[52] In this way, the entire community can collaborate to create a CT-rich learning environment that could be sustained for a long time.^[52]

Finally, as discussed previously, the recommendations to address the gap are going to be multileveled because of the complicated nature of the problem. It is worth noting that the recommendations do not necessarily follow a linear pattern. Although the conceptual level recommendations (i.e., building self-efficacy and nuanced understanding of CT) provide the foundations for the practice level recommendation (i.e., co-design and co-development). The practice-level recommendation also could be an important step to take at the community level. However, the community level changes would also impact the co-design and co-developed practices. Through these practical changes, self-efficacy and nuanced understanding of CT may also emerge.