Engaging solution-based design process for integrated STEM program development: an exploratory study through autoethnographic design practice

The introduction of integrated STEM education is challenging for many education systems around the globe. In response to this, the National Research Council’s (National Research Council, 2011) report Successful K-12 STEM Education: Identifying Effective Approaches in Science, Technology, Engineering, and Mathematics examines research on STEM-focused schools and STEM education practices. Few studies specifically address integrated STEM—simultaneously achieving transdisciplinary integration, authentic contexts, and design problem-solving—through educational program development, not to mention program application in school classrooms. For instance, while the NRC’s report contains overall strategies, it lacks descriptions of how to structure the teaching and learning in integrated STEM education. Similarly, although Australian Curriculum, Assessment and Reporting Authority (ACARA) has released a series of initiative policies to promote integrated STEM education, there have been few practical guidelines for the development of integrated STEM programs in Australia. One of the few exceptions is ACARA’s STEM Connections Workbook (Australian Curriculum, Assessment and Reporting Authority, 2016). This document represents the current best instructional process for developing integrated STEM programs within the Australian curriculum. The suggested process demonstrates a structure that is comprised of four stages. In this study, we focus on the first and third stages (construction of integrated STEM tasks, formulation of implementation approaches) because these represent the principal sections in the program development. However, it is not evident to STEM teachers how they might use the suggested process and how the developed program connected with syllabus requirements. Stage steps do not guarantee that teachers will develop a topic that is appropriate for integrated STEM. Even if they formulate a concept with transdisciplinary integration, they must also evaluate the inclusion of curriculum knowledge. If their chosen idea does not target the Australian curriculum content, the development process will be terminated. ACARA instructional process and its stages have clear limitations in establishing integrated STEM topics and targeting curriculum content knowledge.

As highlighted by Li et al. (2019), “design and design thinking are vital to creativity and innovation, and have become increasingly important in the current movement of developing and implementing integrated STEM education”. For example, English (2019) successfully ran a primary classroom design-based problem-solving activity that integrated the four STEM disciplines. When proposing design as a pedagogical framework, many researchers suggest employing design process models (e.g., Engineering Design Cycle, Design Thinking Process) for STEM education. Teachers can use a specific design process model as the educational design ladder (Wrigley & Straker, 2017) to develop and implement STEM programs, in which the process model will also help students conceptualise abstract concepts for knowledge learning and construction (Li et al., 2019). Along the process of design development, students become more aware of the STEM knowledge they utilise or need in the relevant educational program and can make knowledge-based judgments and explanations (English, 2019). The employment of design process models for integrated STEM education represents an attempt to adopt DBP (Design-based Pedagogy), namely “an educational environment with instructional scaffolds that allow students to solve problems through the practice of design” (Royalty, 2018: p. 138).

It is rational to adopt DBP in developing integrated STEM programs because it allows students to explore and expand the boundaries of creativity through design practice (Royalty, 2018). DBP is also highly contextualised, as students work on real-world problems (Royalty, 2018). For educational purposes, design and Design Thinking are often used as synonyms to describe non-designers’ learning through design practice (Johansson-Sköldberg et al., 2013). Morehen et al. (2013: p. 57) view design as “a vehicle for facilitating higher-order thinking and complex problem-solving abilities”. Together, design and DBP incorporate the development of integrated STEM programs into Design-led Educational Innovation (Zhou et al., 2020), which requires three teacher and student capabilities: (i) Design processes, (ii) Design skills, and (iii) Design mindsets (Wright & Wrigley, 2019). In an adaptation of Goldman et al. (2012), Wright et al. (2018) summarise 15 design skills and four mindsets. A related glossary of terms is outlined in Appendix 1.

Zhou et al. (2020) propose identifying the specific design processes, skills, and mindsets necessary to create the instructional scaffolds for the implementation of DBP. The three capabilities can prepare students for problem-solving through design practice (Royalty, 2018), provide both instructional content and approach, and enable the development of integrated STEM programs. Middleton (2005) also acknowledges that the promise of design in education lies in its process, so a specific design process may inform related design skills and mindsets. Accordingly, Zhou et al. (2020) demonstrates the systematic relationship between DBP and the capabilities of Design-led Educational Innovation. DBP—with its design process, skills, and mindsets—needs to be incorporated into instructional processes of integrated STEM programs, so these components constitute a system for relevant program development (see Fig. 1). Therefore, the system model provides an opportunity to explore the construction of integrated STEM tasks and the formulation of implementation approaches.

The design process model Engineering Design Cycle has been broadly used to construct integrated STEM tasks, in an effort to engage students in relevant instructional content. Its five iterative phases are Problem Scoping, Idea Creation, Designing & Constructing, Assessing Design, and Redesigning & Reconstructing (English et al., 2017). A considerable number of researchers (e.g., Billiar et al., 2014; Fan & Yu, 2017; Shirey, 2018; Tillman et al., 2014) recommend the Engineering Design Cycle as a construction model for STEM tasks. However, through the lens of the three integrated STEM attributes, Zhou et al. (2020) evaluated relevant tasks and then acknowledged that the Engineering Design Cycle might not be fully capable of constructing integrated STEM tasks. Related learning tasks engage students in solving design problems, and most tasks integrate STEM curriculum knowledge. Although teachers can input design topics with designated knowledge, due to a lack of student engagement in problem definition, authentic contexts do not relate to their lived experience in school, community, or work (Committee on STEM Education of the National Science & Technological Council, 2018). This limitation may have negative impacts on students motivation and interest in STEM-related knowledge and activities (Behizadeh & Fink, 2015).

In addition to the Engineering Design Cycle, the Design Thinking Process (Emphasis, Define, Ideate, Prototype, and Test) proposed by Stanford’s d.school (2007) has also been used to construct and implement STEM tasks (Carroll et al., 2010). By employing the three attributes of integrated STEM education to evaluate Design Thinking Process-based STEM tasks, Zhou et al. (2020) indicate that the Empathise and Define phases may allow students to adopt their own design problems. Namely, the problems are self-identified by students, thus addressing the issue of authentic context in the Engineering Design Cycle-based STEM tasks. However, students in related STEM tasks may define problems based on their current knowledge base rather than the knowledge that has not yet been learned, not to mention the integration across disciplines. Students are not constrained to applying content knowledge from the designated curriculum so that teachers in specialised areas are incapable of supervision. This limitation means that the Design Thinking Process is also unsuited to the development of integrated STEM tasks.

Although the Engineering Design Cycle and the Design Thinking Process are inapplicable, they share a manipulation phase—Prototype (or Designing & Constructing). Another common phase—Test (or Assessing Design)—leads to an iterative process that ensures opportunities for improvement. This pair of phases imply the use of related design skills (e.g., prototyping, evaluation, reflection) and mindsets (e.g., experimental mindset) to implement the approach by using prototyping tools to make design models. Design models often serve as an intermediary between design ideas and the physical world, and allow feedback for further design improvement (Greenhalgh, 2016; Sass & Oxman, 2006). Thus, ‘designing, prototyping, and testing’ leads to a design skill and mindset-based approach for implementing integrated STEM tasks. This approach appears to allow students to utilise digital fabrication, which represents a broad suite of computer programming and toolkits technologies as prototyping equipment (Blikstein, 2013).

Evaluation results by Zhou et al. (2020) reveal the limitations in the two recommended design processes: the use of the Engineering Design Cycle may subordinate students’ authentic design practice to the acquisition of discipline knowledge, while the Design Thinking Process appears to overshadow curriculum content knowledge if it takes precedence. These potential impacts echo Carroll et al. and’s (2010: p. 50) argument that “creating a classroom design project that integrates academic standards, content learning and design thinking is a challenging process”. It is likely because applicable design processes that can reconcile authentic design practice and integrated STEM curriculum knowledge remain largely unexplored. This knowledge gap leads to a lack of understanding of (i) The design skills and mindsets of DBP for integrated STEM education, and (ii) The instructional processes for constructing and implementing integrated STEM tasks that involve design models. This knowledge gap, and its corresponding research problem, needs to be adequately addressed to inform the development of integrated STEM programs.

Dorst’s (2011) theory of design reasoning patterns can explain the limitations of the two inapplicable design processes (Engineering Design Cycle and Design Thinking Process) and help identify applicable DBP process models. As illustrated in Fig. 2, Dorst’s theory uses the Basic Model to demonstrate that three specific components—the ‘what’, the ‘how’, and the ‘value’—comprise a fundamental design practice (Dorst, 2011). The ‘what’ represents a designed artefact; the ‘how’ relates to a knowledge-based working principle; and the ‘value’ can be any desired result (e.g., solving a specific problem). Achievement of a desirable benefit requires a design solution that progresses through these working principles. Because these three components are already prepared, Engineering Design Cycle-based STEM tasks seldom provide students with the freedom to define authentic problems. By contrast, the Design Thinking Process leading a complete free problem finding and solving aligns with Abduction-2; that is, the relevant tasks allow students to define the ‘value’ according to personal understanding or preference. Subsequently, each student might use a widely differing ‘how’ to design their ‘what’. For instance, in the case of redesigning urban space (Smith et al., 2015), some students created an elevated cycle path, while others produced a bicycle hedge mount. Thus, it is impossible to designate curriculum content knowledge that will be certainly targeted in that approach.

Supposedly, Dorst’s Abduction-1 where both the ‘how’ and ‘value’ are clear demonstrates the suitability of DBP process models for integrated STEM education. This is because its mechanisms strike a balance between authentic design practice and integrated curriculum knowledge in learning tasks. In Abduction-1, a designer already knows the ‘how’ as working principles; this knowledge helps define a specific ‘value’ and achieve it by solving problems. What the designer does not know is the ‘what’, namely, the solution that needs to be designed. Because the designer can define the ‘value’ that is relevant to him or her, the knowledge principles of the ‘how’ will be designated in authentic design practice. For example, after learning knowledge about an ultrasonic detector, two designers could understand the working principles related to the measurement of distances to nearby objects. Designer A might adopt this ‘how’ to design a parking sensor for warning drivers, while designer B may aim at another ‘value’ by designing a rangefinder as the ‘what’. In these cases, both design actions reconcile fixed knowledge and flexible process. The designers also solve different problems in their respective authentic contexts. In other words, two integrated STEM attributes—design problem-solving and authentic context—have been achieved. Provided that the involved working principle signifies integrated knowledge, transdisciplinary integration—the third STEM attribute—will also occur.

In this study, we have a tentative plan to involve SBDP (Solution-based Design Process) in the development of integrated STEM programs. SBDP represents one of the few applicable design processes that align with Abduction-1. This process model originates from the area of biomimetic design. The Centre for Biologically Inspired Design at Georgia Institute of Technology is the first to acknowledge that relevant cases “began with a biological solution, extracted a deep principle, and then found problems to which the principle could be applied” (Helms et al., 2009: p. 616). A typical example is Velcro invented by Swiss engineer George de Mestral (Ekdahl, 2017) (see Fig. 3), who discovered burrs stuck on dog fur. The burrs gave him the idea of the Velcro strap as a zipper without moving parts. His design exploration from the initial idea to finished product demonstrates the SBDP which follows the steps below:

In the context of integrated STEM education, Steps 1 to 3 represent the procedure of clarifying knowledge principles and represent the ‘how’ in Abduction-1. In this study, we would like to propose that irrespective of whether a natural creature or artificial product serves as a ‘solution’, all bring working principles with specific knowledge. Thus, STEM teachers need to select a transdisciplinary solution that helps to incorporate curriculum content knowledge into a specific task. In Steps 4 to 6, students can then—in a designerly manner—identify problems as the abduction ‘value’. Step 7 allows the students to ground the knowledge necessary to solve the self-defined problem (the abduction ‘what’). Although the Engineering Design Cycle and the Design Thinking Process are inapplicable, they share three rational phases: Ideate (or Idea Creation), Prototype (or Designing & Constructing), and Test (or Assessing Design). It is proposed, therefore, that the addition of Step 7 to these approaches would ensure an iterative procedure to enact SBDP through design model building and prototyping. A prototype model serves as an intermediary between design ideas and the physical world, and then allows feedback for further improvement (Greenhalgh, 2016; Sass & Oxman, 2006). On this basis, we optimise the current SBDP into a nine-step process (see Fig. 4).

The research problem in this study is novel, and this guides us to take an exploratory approach. Addressing the problem and answering the research question (see the Introduction) provide a directive as a research objective: to lay the foundations for methodologically rigorous and evidence-based studies on the development of design-led integrated STEM programs. In seeking to achieve the objective, this study adopts an autoethnographic methodology recommended by Munro (2011) to address a systematic understanding of SBDP’s usefulness in developing integrated STEM programs. Autoethnography humanises research by focusing on experience as ‘lived through’ in its complexities, engaging researchers in rigorous self-reflection to articulate insider knowledge of their experiences (Adams et al., 2017). It is also a method relevant to STEM education, having been adopted extensively by Science educators in recent decades (Roth, 2005). According to Munro (2011), the methodology of autoethnography specifies a system that is an effective research strategy for generating new knowledge, as it brings an applicable approach for evidence gathering and evidence interpretation that is embedded in the temporality of a specific design process. In the research documented here, the ‘auto’ firstly locates us (the authors of this article) as the designers centrally in the creative exploration with our subjectivity and experience. Secondly, the ‘ethno’ (culture) settles our design practice in the culture of SBDP exploration, where the related interrogation is used as part of the critical reflective moment. The ‘graphy’ finally suggests systems of capturing and documenting SBDP-related raw data in the form of the visual language of designing and/or written language of reporting. In other words, the autoethnographic methodology can promote our first-person exploration, in which SBDP itself generates knowledge about the informed design skills and mindsets and the instructional processes for integrated STEM program development.

This exploratory study through autoethnographic design practice is advised by our reflective practice (also referred to as critical self-reflection), a cognitive exercise whereby professionals learn from experience in order to understand and improve their practice (Jasper, 2013). Due to our background as industrial designers, design educators, and educational researchers, we are experienced in design processes, skills, mindsets, instructional content, and approach. The task of this design exploration is to understand the working principles of a selected product, and to apply these to finding and solving an authentic problem. To this end, we created a working prototype and tested it in an extension of the optimised SBDP. Reflective practice allows for the consideration of both subjective feelings and objective events. Because such a type of first-person study (Marshall et al., 2005) is usually based on interpretive theory, McDonald (2012) suggests timely reflections in order to minimise a reflective practice recall bias. Thus, for the most part, we adopted Schön’s (1983) ‘reflection-in-action’, the process of reflecting during practice. ‘Reflection-on-action’, the retrospective contemplation of activity, was also undertaken.

In each step of the exploratory study, we collected data through reflective journals. Such a narrative can capture the events that occur during design practice (Munro, 2011). Corresponding to the nine phases of the optimised SBDP, the whole task was divided into sequential sub-tasks. As each sub-task was completed, we first referred to Helms et al. (2009) for a general review of the step’s practice. Based on this information, we then further reflected on the ‘techniques’ used to achieve the step’s goal. Meanwhile, because design models serve as an intermediary between mental concepts and the physical world, and allow feedback for further design improvement (Greenhalgh, 2016; Sass & Oxman, 2006), the operation skills of design model building and prototyping were also reflected on. This reflection produced data related to “Digital fabrication technologies” section. Consequently, the overall strategy provides a general-to-specific framework that allows for sufficient reflection.

Furthermore, we used a qualitative content analysis method to examine the gathered text data. In responding to the research question, the data was examined corresponding to three clusters: (i) Design skills and mindsets, (ii) Digital fabrication technologies, and (iii) Sub-task contents and approaches to completion. First, we coded the data relating to a general review and ‘techniques’ in NVivo 12 (a qualitative data analysis program), according to Goldman et al.’s (2012) categories of design skills and mindsets (outlined in Appendix 1). Thus, we extracted the skills and mindsets of each SBDP step and then identified the features of the results. Secondly, we sorted the data on “Digital fabrication technologies” section into hardware and software categories. The outcomes presented the educational technologies demanded in design model building. Finally, we summarised all data for each step to describe ‘what each sub-task was’ and ‘how it was completed’.

Our reflective journals captured primary data through the design exploration led by the optimised SBDP (see Appendix 2). The journal content of each step is summarised below:

This exploratory study demonstrates that the optimised SBDP is an applicable DBP process model for integrated STEM education. SBDP and its informed design skills and mindsets, materialises the DBP system model as outlined in Fig. 1. As found in related results, the Solution-based DBP involves conscious mental processes of thinking and reasoning. Because design skill ‘adaptability’ and ‘metacognitive’ and ‘human-centred’ mindsets were widely distributed in step practice (see Figs. 6 and 7), the use of this specific DBP requires complex cognitive strategies and skills. As all process steps demand design skill ‘process language’, the DBP has a highly logical feature. This cognitive and logical DBP implies that STEM teachers should master the optimised SBDP and its critical design skills and mindsets before starting to develop integrated STEM programs. Therefore, we suggest that teachers be equipped with the three DBP capabilities (design processes, skills, and mindsets) through design immersion professional development (PD). The exposure to the optimised SBDP and design exploration will promote the growth of teachers’ relevant expertise.

Although digital fabrication technologies were mainly required in Steps 8 and 9 of the optimised SBDP (see Table 2), they have a significant impact on the completion of design-led integrated STEM programs. The fundamental role of 3D printing aligns with Loy’s (2019) argument that this technology grounds authentic learning in a real world for engagement with complex problems across subject boundaries. However, unskilled operation of 3D printing-based digital fabrication may cause frustration, or even lead to the termination of STEM design projects (Nemorin, 2017). As accepted disciplinary knowledge, digital fabrication has expanded to include programming, engineering, and design (Astrachan et al., 2009; Yasar & Landau, 2003). Thus, teachers in relevant subjects (e.g., Digital Technologies, Design and Technologies in the Australian Curriculum) should be capable of operating various hardware and software to facilitate their students’ building of working prototypes. Students also need skills training to prepare them for the process of learning by making. Supposedly, the skilled operation of 3D printing-based digital fabrication is necessary for both teaching and learning in design-led integrated STEM education.

The summative outcomes in Table 3 indicate the instructional processes necessary for the development of design-led integrated STEM programs. To construct integrated STEM tasks, ‘sub-task content’ first provides teachers with the opportunity to extend the detailed description of each task. Because the sub-task of Step 1 involves overarching planning, STEM teachers need to address the curriculum content targeted by the selected solution. For the other eight steps of the optimised SBDP, teachers can refine corresponding student sub-tasks and outline the exact part of the curriculum content embedded in each step. The last element that should be considered in constructing integrated STEM tasks is the development of achievement standards to evaluate the quality of student learning in Steps 2–9.

Moreover, ‘approaches to completion’ (in Table 3) demonstrates the procedural knowledge needed to formulate implementation approaches to integrated STEM tasks. The procedural knowledge of Step 1 is ready for STEM teachers to use directly. To prepare for the implementation of the other eight sub-tasks, teachers should extend corresponding procedural knowledge to key teaching and learning experiences. These actions will ground their teaching strategies and skills in the context of STEM education by transferring design process steps to the teaching and learning process. Additionally, I identified the specific design skills and mindsets extracted from each ‘approaches to completion’ procedure by referring to NVivo nodes. Teachers may use these prompts to facilitate sub-task implementation.

The outcomes of these two stages of the instructional process need to be converted into a detailed unit plan for classrooms—the final artefact in developing design-led integrated STEM programs. I recommend the 4C/ID model (Four-component Instructional Design model) (Van Merriënboer et al., 2002) for this purpose as it is particularly capable of dealing with complex learning such as integrated STEM education. The four components of this model include learning tasks, supportive information, procedural information, and part-task practice. Therefore, three PD workshops—associated with integrated STEM tasks, implementation approaches, and unit plans, respectively—will likely enable STEM teachers to learn achieving corresponding sections of the above blueprint for developing relevant educational programs.

There are two main limitations to this study. First, despite the rationale of autoethnographic methodology, we were only able to produce limited data through the exploratory study at this preliminary stage. This is because of the conceivable difficulty of getting a research team like the authors of this study at the preliminary stage, whose backgrounds integrate industrial design, design education, and educational research. Future studies may employ the same methodology to involve different designer participants in the design exploration led by SBDP. Then, multiple data sources and triangulation can enrich qualitative datasets for a refinement of current data analysis outcomes, thus developing deeper insights into the research question of the study documented here. The second limitation lies in the research scope, which is not associated with the impact of engaging SBDP in the DBP system for integrated STEM program development. Until now, this topic is still under-researched in the field of design-led STEM, while the influence of other design process models has been explored. For example, National Research Council (2012) states that the Engineering Design Cycle brings ideal access points to integrate design practices into STEM education. Bybee (2010) also points out its essential role, namely embedding problem-solving steps into the implementation process. Future studies need to adequately examine the blueprint proposed earlier, particularly the perceived benefits and challenges for teachers in using Solution-based DBP to develop integrated STEM programs. Research finding are likely to propose guidelines for relevant educational program development.