Promoting STEM Education of Future Chemistry Teachers with an Engineering Approach Involving Single-Board Computers

Featured Application

A project using a single board computer offers at a great learning experience, and a good project context for future chemistry teachers to implement in their future teaching careers. Its versatility is also a great challenge that needs addressing.

Abstract

We describe a master’s level chemistry education course that was designed to support STEM education by strengthening the E component with an engineering approach. Engineering approach is a method of conducting projects systematically similar to professional engineers. In the course, the future chemistry teachers were given the task of building a measurement instrument using a single-board computer (SBC). In addition to course description, we present a pilot study, the aim of which was to explore the opportunities and challenges the engineering approach initiates with pre-service chemistry teachers trying to accomplish a SBC-based open engineering project. The study employed a qualitative research approach, using the course as the data collection platform. The collected data was analyzed using an inductive content analysis. The data analysis shows that an open SBC project is a good platform for learning and teaching future chemistry teachers about chemistry-driven STEM education, but it is very challenging to conduct. The main conclusion is that the engineering approach is a practical solution for strengthening the engineering in STEM education. To generalize these findings to a wider context, we suggest further research to improve the course using this study’s results and re-evaluate the approach in a new instance of the course.

1. Introduction

Single-board computers (SBC), as an affordable and open connected platform, revolutionized the areas of measurement and automation soon after their appearance in 2012 [1]. Experimental work in chemistry education was no exception, and there are many articles describing SBC-based devices that were built to support laboratory instruction [2,3,4,5]. The establishment of the Raspberry Pi foundation, as a charity promoting the educational value of their products, helped spread the idea of using a small, inexpensive, connected computer in various scientific and educational setups [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23]. Several authors have researched and recognized the potential of using SBCs in chemistry-driven integrated Science, Technology, Engineering, and Mathematics (STEM) education. For example, Chng and Patuwo [24] had undergraduate students build a UV−vis spectrophotometer and found a significant improvement in the students’ knowledge of spectrophotometers compared to the control group. Bougot-Robin et al. [25] managed to reduce the black-box effect of using chemistry-laboratory instruments by engaging chemistry students in designing and optimizing an absorbance spectrometer. They argue that building a chemistry device offers multiple learning possibilities. For example, it is a good context to teach analytics skills, such as calibration and determination of the limits of linearity and dynamic range, and, at the same time, demonstrate the iterative nature of scientific research and the instrumental side of chemistry. In addition, to improved content knowledge, this kind of integrated STEM project has the potential to support learning motivation, and the development of versatile problem-solving skills by introducing engineering-type thinking. Other research identified the challenges that educators face when trying to engage students in integrated STEM projects, such as an open SBC project. The most important challenge is related to the diversity of the required knowledge both from teachers and students in multidisciplinary learning environments and the limited resources available in real-life teaching situations [26,27]. To implement engineering design in chemistry-driven STEM education, teachers need an understanding of engineering processes and adequate resources for building SBC-based setups. Another challenge is pedagogical—how to secure an active role for students and maintain it during the whole course? [28,29,30].

A possible solution to these challenges is to apply the so-called engineering approach to designing educational devices and experiments. This approach is a systematic way to design, test, and build devices and experiments [31,32]. It makes it possible to control complex design settings and involves students in the design process as active stakeholders. From several design methods of engineering and design cycles, we chose the function-and-functionality matrix method [32]. The method was originally designed for the field of engineering, but there is no doubt that many fields can benefit from a systematic design framework. By using the engineering approach, a chemistry educator, who is not entirely familiar with engineering, can incorporate engineering and design thinking into STEM education in a coherent manner that has the potential to support learning. Therefore, the aim of this research is to explore what kind of opportunities and challenges the engineering approach initiates when it is implemented in chemistry teacher education in the SBC context. The purpose of the paper is to present good practice examples for all educators interested in applying SBC to strengthen the engineering in chemistry driven STEM education. Analyzing the opportunities and challenges helps address the support needs of the future chemistry teachers developing an SBC-based experiment using an engineering approach.

To match this aim, the research is based on two research questions:

• What kind of opportunities does an SBC project with an engineering approach offer to STEM education? (RQ1)
• What kind of challenges does an SBC project with an engineering approach initiate with future chemistry teachers? (RQ2)

To answer these questions, we introduced the engineering approach in a master’s level chemistry education course and used this as a research platform. As part of the course, the future chemistry teachers were assigned the task of engineering an application of a SBC in chemistry. To understand the possibilities and limitations of the research context, we built a theoretical framework by first justifying the SBCs through experiment automation, then we addressed the STEM research conducted in the context of chemistry-teacher education, and finally, we presented the engineering approach as a tool for strengthening the engineering in STEM.

2. Materials and Methods

2.1. Single-Board Computer as a Tool for Experiment Automation

Experiment automation in its most basic form utilizes sensors to measure physical and chemical quantities, and actuators to adjust the experiment’s input parameters based on the measured values. The two most obvious advantages of having an experiment automated are increasing its repeatability, and the clarity of its results. This is achieved by reducing the human intervention required to conduct the experiment, thus limiting the influence of errors that this inevitably brings with it. Experiment automation is a clear use case for SBCs in chemistry education. For example, Soong et al. [5] built an SBC-based automated burette that can be controlled over the internet to support distance learning during the COVID-19 pandemic, and Sun et al. [4] developed an automatic droplet-analysis device.

An SBC is a suitable tool for experiment automation for several reasons. First, being a fully equipped computer running a standard operating system, it provides the possibility to run a very large range of software to accomplish various tasks. Users can opt to use the ready-made solutions or develop their own. With a lot of the software being open source, it is also possible to adapt previously developed solutions for individual needs. Second, most SBCs provide a wide selection of standard hardware interfaces for connecting the sensors and actuators, together with the software for reading and controlling them. The final, often decisive, advantage of SBCs is their affordability in terms of purchase price and energy consumption.

In this research, we used Raspberry Pi SBCs. The main reason for this choice is the level of support provided by the manufacturer and the user community [33], and their dedication to popularizing STEM education [34].

2.2. STEM in Chemistry Teacher Education

In chemistry education, the STEM approach is often justified by the possibility to link chemistry knowledge to more comprehensive knowledge systems, which can involve, e.g., inquiry- or problem-based learning in authentic multi- or inter-disciplinary project contexts [29]. According to Aydın-Günbatar et al. [35] integrated STEM teacher education should emphasize the integration of different STEM disciplines, but also teach future teachers engineering-based design processes. They suggest, along with many other researchers [36,37,38,39], that engineering could be the connecting component in STEM education, making it meaningful at a practical level. In fact, Moore et al. have defined Integrated STEM education as “an effort by educators to have students participate in engineering design as a means to develop technologies that require meaningful learning and the application of mathematics and/or science” [36]. This can support the relevance of education and provide a less-fragmented learning experience [40]. In addition, many scholars argue that STEM offers possibilities to engage and motivate students [41]. In sum, STEM education has an established status in many countries around the world. Therefore, we argue that it must be addressed in chemistry teacher education, at least to some extent [35].

In general, pre-service chemistry teachers have a positive attitude toward STEM education. In addition to covering content knowledge and improving hands-on skills, it is seen as a way to teach 21st century skills to learners. However, it is also true that STEM projects can be time consuming and costly [42]. For example, projects might demand a variety of materials, tools, and other instruments [40], and organizing STEM projects often involves collaboration between teachers and other stakeholders to cover all the multidisciplinary content demands [43]. As a result, designing STEM projects is challenging. A possible solution to tackle this challenge is to use already published examples and contexts from the research literature [3,4,5,24,25,29]. Another major challenge for STEM in chemistry teacher education is that engineering is not commonly taught in STEM education programs [35]. In this article, we present a solution to this challenge by introducing the Engineering Approach for the engineering method of STEM projects.

2.3. The Engineering Approach and Using It to Prepare and Realize a Sound Concept

Synthesizing a concept for an educational experiment, bringing it to the point of realization, building it, and successfully integrating it into the teaching process requires a systematic approach. This differs from basic exploratory learning often referred to as ‘tinkering’, which de-prioritizes clear objectives in favor of providing free hands-on experience. The target-oriented approach of utilizing commonly known, simple methods of concept synthesis and evaluation to bring a design from idea to product, is often called the Engineering Approach. It is not only useful in engineering but can be applied in all fields of STEM education and beyond. Its principal feature is the design process, which, according to Howard et al. [44], can be linked to creative processes, bringing benefit to the end result. In its most basic form, the design process consists of the following activities:

• Concept synthesis;
• Concept evaluation;
• Materialization of the concept with the highest evaluation result.

The first two activities can be done recursively using a so-called design loop, to optimize the output before engaging in the activity of materializing the final concept, as shown in Figure 1.

The concept-synthesis tool presented to, and used by, the attendees on the course was a function-and-functionality matrix (also called a morphological matrix) [32]. It is represented by a two-dimensional table, where the first column (designated by numerals from 1 onwards) contains the functional break-down of the project idea, and the first row (designated by letters from A onwards) contains the possible implementations for each of the functions. The table cells are filled according to the available function implementations. An example is shown in Figure 2. The concepts are derived from the function-and-functionality matrix by permutating the cells in the table rows and columns. The subset of the concepts selected for evaluation usually consists of those that are realizable with the available materials and skills. The concept synthesis of an educational experiment would typically yield 4–8 concepts that are ready for evaluation.

Once the concept synthesis is complete, the synthesized concepts need to be evaluated in order to be able to select the most appropriate one for production. There are many available methods for concept evaluation [31]. One of the simplest, yet most efficient, is the multi-criteria analysis. It is based on selecting n arbitrary criteria, K_i for evaluating the qualities of a concept, assigning each of them a weight w_i according to their importance to the outcome, and grading them with grades g_i on an arbitrary grade scale. This assigns a real-number value q to each of the concepts as per Equation (1):

$$q={{\displaystyle \sum }}_{i=1}^n{g}_i·w_i$$

(1)

After the value of q has been calculated for each of the concepts, the values can be arranged from the smallest to the largest, effectively ranking the concepts from the least to the most appropriate.

A multicriteria analysis can only yield objective results if selecting the criteria, their weights, and grading is based on objective, quantifiable data, keeping the subjective influence to a minimum. Therefore, a special effort should always be put into defining the criteria and their weights. Subjective influences on grading can be reduced by letting the multicriteria analysis be conducted by several well-informed experts. For engineering projects and their deliverables, two contradictory aspects of their value are usually evaluated: the technical value, and the economic value. The best concepts are those that score the highest in both categories and have a good balance between them, as shown in Figure 3.

2.4. Research Context

The research was conducted on a 5 ECTS (European Credit Transfer and Accumulation System points) advanced master’s level chemistry education course in the spring semester of 2022. One ECTS equals to 27 working hours. Due to the COVID-19 pandemic, the course was designed as a hybrid one, combining the conventional approach with inquiry-based learning and face-to-face education with online sessions. The 5 ECTS course included three parts: 1 ECTS for the Project-based learning Massive Open Online Course (MOOC), including the theoretical framework behind problem-based integrated chemistry education; 2 ECTS for the SBC project at the theoretical and practical levels; and 2 ECTS for reporting the project via a seminar and a short article. The 2 ECTS SBC project part of the course was further divided into three parts (see Figure 4). In the first part of the course, the basic concepts of experiment automation and measurement were presented to the students. This included the fundamentals of the operating principles for sensors and actuators, their interconnection, examples of application, and a brief presentation of SBCs, with an emphasis on how to utilize them in experimental work. At the end of the first part, the students were presented with examples of already-realized educational experiment automation and given an assignment to create their own experiments following the supplied guidelines. In the second part, the students learned about the engineering approach to solving problems and the tools to facilitate the process, such as concept synthesis using function-and-functionality matrices, and concept evaluation using multicriteria analysis. The third part consisted of a practical, semi-guided, 3D-modelling tutorial using FreeCAD [45]. Each of those parts was carried out in a self-contained interactive session. Due to the COVID-19 pandemic restrictions, valid at the time, the sessions were carried out online using teleconferencing software.

After the initial part of the course, the students were given an assignment to synthesize and evaluate the concepts for their proposed experiments. The assignment required them to research the available components, their interconnectivity and compatibility, carry out concept synthesis and evaluation, and select a concept for the realization of their experiment prototype. The project work on the realization of the physical prototypes was conducted during subsequent course sessions, which were held face-to-face in the laboratory.

2.5. Data Collection

The research data were collected during the entire duration of the course. Prior to collecting any data, all data-protection requirements were taken care of by filling out a data-protection declaration with the host university’s legal department. Informed consent was obtained from the students by asking them to complete the web form on the course web site. The host university’s chemistry teacher program is small and therefore, the course was also a small, advanced level activity that included four students. The characteristics of each of the four students were different in several aspects. Their motivation levels ranged from very high to passive, some had a clear and well-defined vision and objectives, while others were struggling to define their objectives throughout the project. Their course attendance records were also different, with some of them present and actively participating in all the course sessions, and others being largely passive or even absent from most of the face-to-face sessions. This led to different levels of project realization, from a fully working experiment prototype to partly working ones at different stages of completion.

The data-collection process was as follows:

• Field notes were written during the course by the course instructors. These included remarks and observations about all the course sessions and all the additional communications between the students and the instructors (e-mails, posts on course Q and A board, and oral communications outside the course);
• At the end of the course, semi-structured interviews [46] with the participants were performed, recorded, and transcribed. The interviews were conducted as part of the workflow presented in the Supplementary Materials;
• As the final assignment of the course, the students wrote and published a project-description article in an open-access journal for teachers-in-service training purposes. Articles are licensed under the Creative Commons attribution license (CC-BY), which enabled us to include them in our data. Three out of the four articles by the participating students were peer-reviewed, accepted, and published.

The course finished with zero attrition rate. Four students participated and were able to present a project deliverable at the end. Of those, two students agreed to participate in the interview as part of the data collection. The small sample size is not problematic in this case, because our research is piloting the research design, and research questions demand a qualitative research strategy. To provide a comprehensive description of the studied phenomenon and support the reliability of the research, we used data and method triangulation by applying versatile methods (interview and observation) on different data sources (field notes, documents, and interview transcriptions) [46]. Overall, the number of participants is small, which is why this research is not aiming to provide a comprehensive qualitative understanding by means of data saturation. The aim of the research is to report a good practice example by presenting an analysis of the data gathered during an authentic course implementation.

2.6. Data Analysis

First, all the collected data were processed into a suitable electronic document format. Then, the analysis was carried out using an inductive, qualitative content analysis [46]. To ensure that the analysis provided answers to the set research questions, it was guided with the themes raised from the research questions (RQs):

• RQ1: Opportunities of SBCs for STEM education;
• RQ2: Challenges that the engineering approach initiates.

The researchers read the raw data and extracted all the sentences and paragraphs that could potentially fall under the individual RQ themes. Then, the extracted analysis units were simplified for easier processing. Next, the simplified expressions were processed as clear observations that provide answers to the RQs. Finally, the processed observations were classified into categories under each research question. This procedure was used to extract the meaning from all the parts of the collected data. An analysis example is presented in

Table 1. Inductive contents analysis.

Data Source	Raw Data	Simplified Expression	Observation/ Solution	Category	Research Question
Article A3	“This device can be used in chemistry education as a demonstration tool or as part of a project-based learning lesson plan centered around laboratory experiments, aimed at figuring out what affects the quality of air.”	The result is a diverse tool for education	SBC project produced a working device with multiple educational opportunities	Future teaching (5)	Opportunities (RQ1)
Interview I1	“So, it involves a lot of scientific knowledge. Uh, and problem solving. And quite meticulously. One would have to, at least from my viewpoint, just find the allocated time and have a very large knowledge base to actually execute these projects.”	Planning takes time	Planning needs more guidance	Planning (4)	Challenges (RQ2)

, the semi-structured interview schema can be found in Appendix A, and all the relevant parts of the dataset are available in the Supplementary Materials.

3. Results and Discussion

When the learning opportunities, challenges, and support needs are considered, it is important to keep in mind that, in addition to the general course objectives, the students also set their own learning goals. For example, in this research the participants had different kinds of expectations for the course:

• “Build something that will work rather than just something on paper”.
• “Try to make something simple that will work”.
• “Just pass the course that’s still missing before my graduation”.

These answers reflect the different sets of objectives by different students and indicate the requirements for a different approach to each of them by the course’s instructors. The different objectives are attributed to the characteristic student categories and provide the diversity required for an unbiased data analysis with the small data sample size. With this in mind, in the next subsections, we will report the results according to the research questions (RQ1 and RQ2) based on the data from the interviews, field notes, and the final published articles written by the students. The relevant excerpts from the raw data are available in the Supplementary Materials.

3.1. Opportunities That an SBC Project Offers for STEM Education (RQ1)

The data analysis produced five different categories for opportunities. Three categories—lecture contents (1), planning (2) and supply (3)—addressed the engineering approach and two other categories were about learning (4) and future teaching (5). According to the interviews, the lecture contents should address the possibilities of working examples through already-built devices introduced in the literature., e.g., [23,25]. The function-and-functionality matrix was found to be helpful for the systematic planning of objectives, components, and budget. However, the component selection should be based on the needs derived from measuring the selected chemical phenomenon, which will determine the necessary functions of the device [31,32]. This is a crucial aspect to consider in the planning phase, where the concept is evaluated. In addition, the supply phase made it possible to learn new know-how about purchasing components. From the learning and the future teaching perspectives, pre-service chemistry teachers felt that an SBC project offers a good context to implement learning-by-doing through project-based learning. It stimulates the recalling of old knowledge, upon which the new knowledge is built. This offers the possibility to integrate multiple STEM disciplines into the project [35]. In addition, the SBC project was experienced as an important learning opportunity for instrumentation in science, from both the teachers’ and the learners’ perspectives. This aspect can be used to show the nature of chemistry as an instrumental science. Interview 2 revealed meta-level opportunities. First, managing to build something that works improved the feeling of self-efficacy. Interviewee 2 worked as a substitute teacher during the course, and when she showed the result to her pupils, she felt that she had the opportunity to be a positive role model for life-long learning and support of gender equality in science. The last notion of learning is about the opportunities of the seminar. It enabled peer support, learning from others, and provided some healthy competition with peers. The summary of the observations and the identified solutions for presenting the opportunities are shown, grouped by category, in

Table 2. Categorized observations associated with the opportunities of SBC for STEM education (RQ1).

Category	Observations/Solutions
Lecture contents (1)	Presentation of opportunities early on, based on working examples. Learn the function-and-functionality matrix for systematic planning of objectives, components, and budget.
Planning (2)	Component selection should be based on the needs first
Supply (3)	Opportunity of acquiring new know-how about buying components.
Learning (4)	Opportunity of emphasizing and affirmation of the concept of lifelong learning. Seminar enables learning from others. Opportunity of learning-by-doing experience. Understanding instruments in science education. Opportunity of healthy competition with peers. Opportunity of satisfaction in reaching a goal. Opportunity to build on old knowledge in a multidisciplinary context Opportunity to learn the connection between chemical phenomena
Future teaching (5)	Context for project-based learning Show the instrumentation of science to students SBC project produced a working device with multiple educational opportunities. Opportunity of providing a positive example and a role model. Opportunity of emphasizing and affirmation of gender equality.

.

3.2. Challenges That the Engineering Approach Initiates (RQ2)

The data analysis identified five categories of challenges that an SBC project can present to the students and the course instructors. The first two categories include communication (1) and guidance (2) challenges, the third category, learning (3), includes a broad set of challenges connected to lecture content and timing, the fourth category, planning (4), includes challenges associated with planning and logistics, and the fifth category addresses collaboration (5). Note, that we also discuss potential solutions and support that may address and solve the identified challenges.

The data collected from the interviews and from the student articles shows that the main challenges in the communication category are conveying the message about the instructor’s readiness to help and presenting the opportunities of the technology applications. The instructors should carefully present the available communication channels at the very beginning of the course, explicitly inviting the students to use them, and provide periodic reminders to do so during the entire course. It is clear that having access to various ways of quality and timely communication is valued highly by the students. These responses show that the respondents find the guidance useful in general. The wish that was expressed most clearly and most often was the desire for practical guidance during the building phase. It is also interesting to note that, in retrospect, the students appreciate the balanced and friendly pressure by instructors that can steer them into the right direction.

Addressing the challenges in the guidance (2) category is a more difficult task. Depending on the skills and willingness of the students, it is best addressed by presenting them with a set of application examples, and a suitable presentation of pointers to the literature on previous research. The interview respondents also identified the need for better guidance through the initial phases of their projects. This calls for the instructors to carefully monitor the students’ progress during the entire course, catch any unwanted deviations from the optimal path, and provide discreet and decisive guidance in a timely manner. The latter can pose an additional challenge to the instructors if the course students-per-instructor ratio is too high.

The next category (3) includes challenges related to the learning process itself. The data analysis reveals that the interview responses and the students’ articles identify a broad spectrum of those challenges. Most of the identified challenges are related to difficulties with connecting the theory to practice, and bridging the gap between “theory” and the “real world”. Addressing challenges of this type can be demanding for an instructor and the students, as it depends a great deal on the previous skills, the students’ abilities, and networks. The subsequent data analyses show that most of these challenges are best addressed by the efficient use of communication channels and providing individual support in form of scaffolding instructions [30]. These responses clearly reflect the need for regular and frequent personal contact and the constant availability of the instructor, but there is also a strongly expressed wish for hands-on experience, a visual explanation of concepts, and a more in-depth presentation of the project-organization skills. The responses also contain ideas for solving some of the learning challenges, such as providing a concrete form for how to proceed in the project-related problem solving. Further responses indicate the difficulty with identifying the goals set during the course, and show that the students find following the procedure of concept synthesis difficult. Even when they do master it, some of them still have difficulties executing the appropriate number of the design-loop iterations. A negative consequence of this is a failure to produce a clear concept idea, and thus a failure to properly document and execute the selected concept. Again, this calls for a more careful presentation of the project’s goals and the tools available to achieve them. Both are also associated with providing the students with a clearly conveyed opportunity to give feedback and react to it. As with the guidance challenges above, the demands on the instructor here grow with the students-per-instructor ratio.

The fourth category of challenges, (4), contains the ones related to planning. These include the challenges imposed by planning the design process in the broader sense and those related to material supply and logistics. For example, the planning of the course content and scheduling, planning and coordination of the activity paths, and the planning of logistics and supply operations. These responses clearly show that the students need careful mentoring during the planning phase, which will increase the probability of achieving the goal of the project. Additionally, their progress must be constantly monitored. The challenge of proper planning is considered a complex one in general engineering and becomes even more so if the project is being conducted by students who lack previous experience with project work. Therefore, it is essential that the planning tools and the skills to apply them are included in the course very early on. Combined with the logistical and supply challenges, especially in the current uncertain conditions, this also poses a significant timing challenge. In an ideal case, all the components would be obtained from a single supplier, which is rarely possible with real-world projects. Therefore, the support to the students with logistics planning presents the challenge of finding a delicate balance between the suitability and the availability of individual components, which must not be left to the students alone.

The last category of responses, (5), was associated with different types of collaboration. These responses express the need for at least two types of collaboration: one between teachers from different fields, thus striving for interdisciplinarity, and the other between peers. Both types are important and require careful planning, execution, and support from the instructors.

A summary of the observations and the identified possible solutions to the challenges by category are presented in

Table 3. Categorized observations associated with the challenges that the engineering approach initiates (RQ2).

Category	Observations/Solutions
Communication (1)	Help and interaction possibilities should be offered more clearly. Expressing a clear intent to provide support. Presentation of opportunities early on, based on working examples. Realistic feedback on the projects’ challenges and opportunities. Providing means to discuss the issues. Providing help to students who find themselves in frustrating situations is essential.
Guidance (2)	An open engineering task is challenging. Monitoring of students’ notes is required. Providing support by documenting the project’s progress. Spare parts available if something goes wrong in the assembly. Workshops for all basic engineering skills are important (e.g., soldering, coding, 3D printing, etc.). Providing a means to reach consensus when discussing open issues. Providing well-argumented and friendly pressure to help solve problems.
Learning (3)	Adapt theory to previous knowledge and skills. Clear communication of goals is needed. Students lack of experience, demanded content knowledge, tacit and meta level knowledge. Steps in the design process must be explained. Providing pointers to relevant literature is needed. Lectures has to be introduced and timed in accordance with previous knowledge. Help with coding is required from instructors. More training on project management. Frequent live meetings to support progress. Live classes with hands-on experience increase understanding and improve synthesized concepts. Students need a clear working schedule about when instructors are available, and the range of communications channels. Step-by-step procedure for problem solving, e.g., a form. Providing basic terminology is needed. Providing help with mastering the basics of OS navigation helps reduce students’ frustration.
Planning (4)	Planning phase needs more attention and guidance because poor planning makes finishing harder. Project is challenging and demands careful planning from instructors, such as curation of potential projects. Prices and availability must be verified before starting the project. Project planning to minimize costs is essential. More careful planning of component supply is required. Cataloguing the components simplifies the planning of activities. Make sure everything is available at one provider or the laboratory inventory. Students need help with the identification of compatibility. Questions about hardware specifics must be answered in time to continue the project. Realistic feedback from the plans. More support to component selections. Directions on application must be passed on earlier in the project.
Collaboration (5)	Support when searching for potential collaboration partners. More peer support. Peer meetings can be therapeutic.

.

4. Conclusions

This qualitative pilot study has shown that an open SBC project is a good platform for learning and teaching future chemistry teachers about chemistry-driven STEM education, but, at the same time, it is very challenging to conduct. Altogether, the qualitative content analysis of the observations, interviews, and student articles revealed five categories for opportunities (RQ1) and five for challenges (RQ2) (see

Table 4. A summary of the found opportunities and challenges.

	Opportunities (RQ1)	Challenges (RQ2)
Categories	Lecture contents (1) Planning (2) Supply (3) Learning (4) Future teaching (5)	Communication (1) Guidance (2) Learning (3) Planning (4) Collaboration (5)

).

An SBC project’s strength is its versatility. For the pre-service chemistry teachers, it offers at the same time a great learning experience (e.g., engineering thinking, project work, and applying content knowledge), and a good project context for them to implement in their future teaching careers. The project deliverables are tangible and durable. If designed thoughtfully, they are adaptable, expandable and implementable in various teaching and scientific activities. However, the versatility is also a great challenge, because of the multidisciplinary content knowledge demands. For example, students needed to learn new skills (3D modelling, 3D printing, soldering, use of a command-line interface, basic level programming, etc.) from scratch, and apply them in the chemistry-instrument design context. It is obvious that those skills, in addition to the pedagogical and organizational ones, are also required in the team of the course instructors. Projects of this kind can be conducted in a 5 ECTS pre-service chemistry-teacher education course, but the time allocated to the project should be much greater than in this pilot study where it was 2 ECTS. The time challenge can be addressed by conducting the project in small teams or pairs within a larger group.

The planning is one of the most vital aspect of the project. This presents a serious challenge for the course instructors and has a series of practical implications. According to our analysis, it is very difficult to achieve smooth progress if the plans are not realistic or accurate. The challenge is that the students lack the experience and knowledge to make a sound project plan. This can be supported by more concrete and prompter feedback from the instructors, and by providing students with appropriate scaffolding instructions. In addition, students could also reach out to external experts for feedback about the feasibility of their plans. Solid plans will reduce the practical problems during the course, and should be highly prioritized when planning a replication of a similar course. In our case, changing the objectives affected the requirements relating to components. The components need to be ordered, which will take time from a few days to several weeks, but a university course has a strict schedule and a limited duration. It would be best to order all the components at once to ensure that all the students have all the parts to proceed with their projects. The other way to address the supply and lead-time issue would be to build a large enough repository of components beforehand and require the students to only use those components in their projects.

Concerning the feedback, it is important to consider what kind of feedback is suitable for different types of students. In this study, four different types of students participated on the course. One student found that open, reflective feedback inspired them to work harder, but others would have needed more concrete, straightforward help, such as “this will work/this will not work because …” Everything considered, the course needs clear communication channels, and students need to know what kind of questions they can ask and what they are expected to be able to solve by themselves. This is not an easy issue to resolve, but improving the communications can begin with developing a culture of regular discussion. Even though the project’s nature is independent, both the students and the instructors would benefit from regular meetings to make certain that all the students proceed well.

As discussed earlier, this study indicates that an SBC project conducted with an engineering approach offers many opportunities for learning, but also includes many challenges. The identified opportunities and challenges are in accordance with the findings of previous research on project-based learning and integrated STEM education [30,47,48,49]. Based on this result, we suggest that the engineering approach is a valid one to concretize the E-component in STEM education, which often remains abstract for students. This is the main take home message for the students because it is an authentic method for ensuring solid project plans, which is the most challenging part of the entire process. By implementing this method in their future profession, they will transfer authentic engineering practices to their future teaching. The results can be considered reliable in this study context because we triangulated the phenomena using multiple data sources on a diversified participant sample and conducted the analysis carefully. However, it is important to keep in mind that the course has been organized for the first time and during the peak of the COVID-19 pandemic. Thus, our pilot study was carried out including only four students finishing in the course, and two of them participating in the interview as part of the data collection. Therefore, the results of the study cannot be generalized to other contexts without further investigation, which is a clear limitation. Then again, the objective of this pilot research was not to provide an in-depth analysis of the opportunities and challenges via saturation, but rather to serve as a good practice example of what one should consider when starting to conduct a SBC related project with future chemistry teachers. For future work, we suggest a redesign of the course with an improved feedback strategy, communication channels, an iteration of the lecture content to support the central challenges, then a re-evaluation of the course with more students participating.