Engineering design has been widely implemented in K-12 curricula to cultivate future workforce. In this study, seventh-grade students (
To date, adopting renewable energy has been an environmentally sustainable solution to mitigating the adverse effect of climate change and natural resource depletion. For this vision, younger generations of citizens are expected to become the agent of addressing the issue of climate change (Tayne et al., 2021). Providing young generations with appropriate learning opportunities to understand renewable resources and make conscious decisions about energy consumption is thus important.
An action-oriented science curriculum engages students in meaningful, real-world activities that promote civic participation and responsible citizenship (Cordero et al., 2008;Hampshire et al., 2021) and allows them to witness how their actions can drive societal change (Hodson, 2014;Holloway et al., 1998). Research has shown that science curriculum on climate change needs to empower youths with hope and encourage individual actions towards promoting environmental sustainability, beyond narratives that focus on fear or alarming stories about the climate crisis (Busch et al., 2019;Li & Monroe, 2019;Tayne et al., 2021). Involving youth participation in this effort necessitates an action-oriented science curriculum that models authentic practices of combatting climate change.
Particularly, integrating engineering design offers a practical approach to modeling authentic practices and fostering youth's skills and competencies crucial for taking meaningful actions (Aydoğan & Çakıroğlu, 2022;Magana et al., 2021;Wendell & Rogers, 2013). Engineering design is an iterative decision-making process of developing solutions to fulfilling design objectives, usually under a set of constraints (Xie et al., 2018). Engineering design provides rich learning experiences and engages students in applying science knowledge and practices in addressing tasks with a design objective in an authentic setting (Mumba et al., 2024). Students are enabled to navigate trade-offs between criteria and constraints within authentic problem-solving contexts (Kelly & Cunningham, 2019). The literature has demonstrated that students should be guided to carefully consider contexts and constraints in developing or refining design solutions (Anderson et al., 2010). While the literature recognized the importance of empowering students to balance criteria and constraints, students often lacked guidance and opportunities for practicing it. This gap called for a nuanced understanding of how students approach engineering design processes and balance criteria and constraints, which is the focus of this study.
With the integration of epistemic tools in engineering design processes, students are allowed to take ownership of knowledge building and inquiry (Magana et al., 2021). Epistemic tools are broadly defined as strategies or resources that facilitate individual or collective construction of knowledge (Kelly & Cunningham, 2019). The benefit of using epistemic tools, such as computer-based simulations, to advance knowledge construction and develop student agency has been documented. For instance, computer-based simulations offer the opportunity for youths to explore the interplay between constraints and design objectives in engineering design processes, which further facilitates their translation of knowledge into actions in authentic settings (Chao et al., 2017;Kelly & Cunningham, 2019).
Therefore, the purpose of this study was twofold. First, the study integrated engineering design processes into an action-oriented seventh-grade science education curriculum, Solarizing Your School, on an epistemic computer-aided design (CAD) tool, Aladdin (Xie et al., 2023). Second, the study aimed to investigate in what ways and to what extent the action-oriented curriculum impacted seventh-grade students' actions towards adopting renewable resources. The findings of this study provide significant implications on designing epistemic practices and action-oriented learning opportunities for youths to participate in the endeavor of combating climate change and natural resource depletion.
Constructionism, developed by Seymour Papert, focuses on knowledge reconstruction by students using meaningful artifacts (Papert, 1980). Rooted in constructivist theories with a focus on mental construction of knowledge (Piaget, 1967(Piaget, , 1973)), constructionism emphasizes that knowledge is actively constructed by students rather than passively transmitted from teachers (Kafai & Resnick, 1996). This theory posits that students construct knowledge by actively engaging in the creation of epistemic artifacts (Harel & Papert, 1991;Papert, 1980), in which students can thereby facilitate knowledge construction and develop conceptual understanding. Central to this theory is the idea that learning occurs when students design and build personally meaningful artifacts. Personally meaningful artifacts serve as "objects-to-think-with" (Papert, 1980, p. 11) for students, which externalizes their ideas and prompts reflectionin-action and iterative design processes to reinforce their understanding (Blikstein & Wilensky, 2010;Kafai & Burke, 2015). Another key tenet of constructionism is the focus on socially and culturally situated knowledge construction, where students collaborate to create shared artifacts (Kafai & Burke, 2015;Kafai & Resnick, 1996). These artifacts provide opportunities for students to interact with peers, construct knowledge together, and refine their understanding collaboratively (Blikstein & Wilensky, 2010;Cano et al., 2015). Moreover, constructionism emphasizes publicly shareable artifacts that foster authentic dialogue and feedback within a learning community (Kafai & Burke, 2015).
Constructionism has been increasingly applied to frame environmental science education programs to foster student knowledge construction and promote action-oriented agency towards combating climate change (Jose et al., 2017;Karahan & Roehring, 2015;Parth et al., 2024). Karahan and Roehrig (2015) demonstrated that students who developed multimedia artifacts addressing personally relevant environmental concerns exhibited increased environmental awareness and agency. Similarly, Jose et al. (2017) implemented constructionist principles into outdoor field trips in which secondary students engaged in hands-on learning and iterative knowledge construction focused on local ecological systems, improving their environmental science knowledge. Furthermore, Parth et al. (2024) conducted a design-based research project in which teenagers participated as researchers in a climate change education program. Through this process, students constructed meaningful scientific knowledge and demonstrated agency towards contributing to broader environmental discourse.
Informed by the framework of constructionism, this study allowed students to create authentic design artifacts by following engineering design processes on an epistemic tool. Engineering design facilitates the creation of artifacts, especially when these epistemic tools enable students to reflect and think critically (Atman et al., 2007;Harel & Papert, 1991). In addition, engineering design processes engage students in iterative analyses of design tasks and contextual factors (Akins & Burghardt, 2006;Atman et al., 2007). This allowed students to align their design objectives with practical criteria and consider physical constraints, ensuring that their design artifacts are meaningful (Baze et al., 2023;Hod & Sagy, 2019).
Moreover, this study addressed the critical need to empower students to develop socially and culturally relevant knowledge (Kafai & Burke, 2015;Kafai & Resnick, 1996) on renewable energy in an authentic, action-oriented curriculum (Cordero et al., 2008;Hampshire et al., 2021). This approach can help students connect themselves to real-world challenges and drive their action towards effective ways of balancing design criteria and constraints, which are essential for developing solutions aimed at adopting renewable energy practices.
Renewable energy is strategically important to economic development worldwide, especially when the emerging AI industry is driving up the cost of electricity significantly. Accordingly, citizens should be aware of the urgency of adopting renewable energy but, on the contrary, their opposition to proposed deployments sets significant obstacles to achieving the goal of clean energy. Therefore, it is crucial to educate citizens about renewable energy and involve each citizen in the vision of establishing an environmentally friendly society.
Empowering the younger generation to embrace renewable energy emerges as a pivotal strategy. The Youth Strike for Climate movement started in 2018 has instilled in youth with a renewed sense of urgency around climate action and has pushed the issue to the forefront of global political and social discourse (Lowan-Trudeau & Fowler, 2022). Noteworthily, young generations, given their age and relatively limited knowledge on climate change, can often feel overwhelmed and discouraged by the disparity of the crisis, especially when exposed to frequent coverage of alarming stories on this topic (Busch et al., 2019;Ojala, 2012). A curriculum that emphasizes fear about the environmental crisis may exacerbate youth's feelings of grief and hopelessness (Hicks & Bord, 2001;Ojala, 2012). Research has indicated that those negative feelings over environmental issues can thwart their active participation in efforts to address the crisis (Li & Monroe, 2019;Sanson et al., 2019).
Alternatively, climate change education seeks to empower youth to take meaningful action in adopting renewable energy. From a positive psychology perspective, educators emphasize fostering "hope"-not only as a source of optimism but also as a motivator for young people to build agency in protecting the environment (Li & Monroe, 2019;Lowan-Trudeau & Fowler, 2022;Ojala, 2012;Tayne et al., 2021). For example, Tayne et al. (2021) demonstrated that co-produced films by students and mentors effectively instill hope and inspire others to make environmentally responsible choices. Li and Monroe (2019) examined youth's hope through the Reasonable Person Model and revealed that youth's belief in everyone's agency towards making meaningful change is a key factor in cultivating hope. Collectively, these studies highlight the importance of empowering students with environmental science knowledge in real-world contexts and encouraging individual actions that promote environmental sustainability.
A central tenet of the nationwide science education reforms in K-12 settings is to integrate engineering design processes into science instruction (National Research Council, 2012;Next Generation Science Standards, 2013). Engineering design processes include iteratively analyzing tasks and carrying out design details (Howard et al., 2008). When working on engineering design processes, expert engineers spend significantly more time than engineering students on problem scoping, which entails practices of framing design objectives based on criteria and constraints, collecting information, and articulating hypotheses (Atman et al., 2007). Typically, design objectives are situated in a real-world context, contain ill-structured problems, enable a wide range of design paths, and promote challenge-based learning (Jonassen et al., 2006). In Akins and Burghardt's (2006) study, students gained science and mathematical competency after designing a food dehydrator with mathematical reasoning. The design objective was set to promote two core disciplinary learning opportunities: the learning of math concepts, such as area and linear functions, and environmental science concepts including humidity and evaporation. In the design process, students learned design optimization and associated tradeoffs required to not just meet design criteria but also work with practical constraints. Similarly, Silk et al. (2009) provided an engineering design context in which students could reason scientifically. To enable scientific reasoning, students were required to first identify a personal need for an electrical alarm system. This approach held the promise of motivating students to take personal ownership of selfidentified problems.
In addition, the literature has shown that engineering design empowered younger students to consider the balance between criteria and constraints in solving authentic problems (NRC, 2012). For instance, Purzer et al. (2015) engaged students in a zero-energy challenge in which students designed energy-efficient houses with the capacity of producing more renewable energy than the energy required for heating and cooling over a year. In the zeroenergy challenge, students navigated design constraints by applying their interdisciplinary knowledge, such as using math skills to ensure their design stayed within budget limits and applying scientific principles to determine the size and placement of windows based on geographical constraints. In another study conducted by English et al. (2017), students followed engineering design processes to construct simulations of buildings with capacities of withstanding earthquake damage, through which students developed an in-depth understanding of earth science concepts and perceived the activity as authentic and engaging. From these studies, we recognize that "making trade-offs between criteria and constraints" (Cunningham & Kelly, 2017, p493) is a key element in the engineering design processes through which students learn about complex and rich earth science knowledge and practices.
Epistemic tools are "tools function in ongoing, concerted activity to produce knowledge" (Kelly & Cunningham, 2019, p. 1081). They can take on various forms, such as instructional strategies and processes, assessment rubrics, data visualization, models, simulations, and other concrete tools (Baze et al., 2023;Cunningham & Kelly, 2017;Kelly & Cunningham, 2019;Wendell & Rogers, 2013). Research has shown that epistemic tools are critical in science education as they enhance students' ability to conduct scientific inquiries, promote understanding of scientific concepts, and elicit new insights into scientific phenomena (Chao et al., 2017;Magana et al., 2021;Mumba et al., 2024).
Computer-based simulations, as a form of epistemic tools, offer to integrate engineering design processes in science teaching practices (McAlister et al., 2022). Aligned with the NGSS's vision (Next Generation Science Standards, 2013), computer-based simulations promote cross-disciplinary learning in science education (Magana et al., 2021). Through simulations, students engage in a constructionist learning environment, empowering them to construct knowledge by exploring, creating, evaluating, modifying, and disseminating engineering design solutions or artifacts (Chao et al., 2017;Magana et al., 2021). In addition, computer-based simulations can provide students with rapid iteration and prompt feedback while they are seeking to optimize engineering design solutions or artifacts (McAlister et al., 2022). For instance, students can formulate dynamic visualizations of scientific experimental results or subtle changes in scientific phenomena, which are often invisible or not readily available in real-life settings (Magana et al., 2021). In turn, the visualized results continue to facilitate interactions between students and the "object-to-think-with" (Papert, 1980, p. 11), providing students with tailored feedback to enhance the design artifacts and performance (Chao et al., 2017;McAlister et al., 2022). Through students' iterative interaction with design challenges and epistemic tools, computer-based simulations make it possible for students to balance criteria and constraints of the given challenges, linking learning with real-world contexts (Kelly & Cunningham, 2019).
RQ1. How do seventh-grade students' balance criteria and constraints in the action-oriented science curriculum? RQ2. How do seventh-grade students' engineering design processes relate to their design performance in the actionoriented science curriculum? RQ3. How do seventh-grade students describe their experience in the action-oriented science curriculum?
An institutional review board approval was granted before we recruited participants. A total of 38 seventh-grade students from a public middle school in the Northeastern United States voluntarily participated in the study, including 17 girls, 17 boys, and the other four participants who preferred not to reveal their gender. All the participants were from four different classes but were taught by the same teacher who had a well-established partnership with the research team. Course materials used in this study were developed by the research team. Before implementing the project, the teacher attended professional development workshops offered by the research team. Upon confirming the teacher's readiness for facilitating the curriculum and using relevant course materials, the research team took the position of participantobserver (Spradley, 1980). Our primary responsibilities were collecting data and giving feedback when needed.
The epistemic tool used in this study was Aladdin (https:// intof uture. org/ aladd in. html) developed by the Institute for Future Intelligence. Aladdin is an integrated CAD/CAE platform that can simulate renewable energy applications supported with computational models and analysis. Aladdin allows users to perform five categories of actions that represent critical stages of engineering design, including exploring (i.e., observation and formulation), constructing (i.e., building 3D models), analyzing (i.e., simulating energy use), optimizing (i.e., review and revision) and evaluating designs (i.e., analysis). For instance, in one scenario, a student first rotated the building (observation), then added a solar panel (construction) and then rotated the building again (observation). The engineering design processes of this scenario can be represented as (observation, construction, and observation).
Using built-in 3D models, users can build a house in various styles (e.g., colonial), install windows and doors, plant trees, and deploy solar panels. In addition, with the aid of computational algorithms, users can analyze the energy performance of the house referring to the percentage of the energy consumption that solar panels can offset. The tool provides computational algorithms with the capacity of integrating environmental factors, such as the location of the house (e.g., city, latitude) and the duration of sunlight hours for various seasons, into the analysis to granularly calculate the daily or annual production of electricity from the solar panels. For example, the annual production of electricity for the house (with the same solar panels) built in the northeastern United States is different from that in the southern regions as the amount of sunlight and the angle at which it hits the solar panels varies depending on the geographic location and the time of year. In general, areas closer to the equator receive more direct sunlight throughout the year, resulting in a higher potential for electricity production compared to areas further from the equator or with more cloudy weather patterns. Users can run the analysis periodically to monitor their progress toward their objectives, but they are also supposed to be aware of unforeseen constraints (e.g., each project has its budgets; solar panels cannot affect neighbors by creating excessive solar glare).
The action-oriented science curriculum, Solarize Your School, was designed to align with core tenets of constructionism (see Table 1). The objectives of this curriculum were to teach middle school students solar energy knowledge, facilitate knowledge transfer in solving an authentic problem, and foster their agentic actions towards adopting solar energy wisely and effectively.
Students were empowered to take ownership of their environmental impact and contribute to a more sustainable future. The final project was to design a plan for adopting solar energy to offset annual energy consumption in their school. The project was completed on the epistemic tool, Aladdin. Before working on the final project, participants attended the first three sessions focused on orientations (e.g., lectures about solar energy knowledge, the importance of building a sustainable future, and instructions on using the tool) and two simplified design tasks using the same tool. The curriculum lasted for five days, 50 min per day. Throughout the project, each participant worked independently on their own laptop.
The first day featured an overview of solar energy and an orientation about the epistemic tool, Aladdin. Participants first brainstormed daily objects (e.g., solar streetlights) powered by solar energy. Then the teacher explained how electricity is generated from solar energy. Furthermore, the teacher guided students to individually explore energy-efficient buildings in the epistemic tool, discuss the environmental impact of their daily actions, and think critically about ways to reduce their carbon footprint (Xie et al., 2018). Afterward, participants explored various functions of the tool, such as adding walls and windows, on their own.
On Days Two and Three, participants completed two simplified design projects [e.g., finding a solar "sweet spot" (i.e., an optimal spot where a solar panel generates maximal electricity throughout the year) and building an energy-efficient house] in which the participants manipulated one solar panel or several racks of solar panels to generate electricity for buildings. These simple design projects were designed to help participants get familiar with the tool and facilitate their decision-making process over energy consumption.
On Day Two, participants were guided on how to apply environmental science concepts to solve energy problems. Specifically, students were given a three-dimensional (3D) model (see Fig. 1; left), including a house surrounded Creating tangible artifacts Students created a concrete artifact of a three-dimensional design plan of solar panel array placement tailored for their school using an epistemic design tool Personal meaning-making
The final task required students to develop solar panel solutions for their own school, making learning highly relevant and meaningful to their lived experience Social construction of knowledge Although students worked independently, this curriculum fostered peer learning and communal knowledge growth through discussion and presentations on a topic closely relevant to their social contexts Reflection-in-action Following the engineering design process, students iteratively tested and refined their solutions while receiving immediate feedback via simulations Publicly shareable artifacts Final projects were presented to the class. Students were encouraged to articulate their design solutions and receive authentic feedback, fostering authentic dialogue within this learning community by trees and a single solar panel in the yard. The storyline of this task was that the homeowner hoped to supply power for their outdoor lighting using solar panels. The design problem was to help the homeowner increase the annual energy output of the solar panel. For this objective, participants were expected to modify the properties of the solar panel and change how each panel was positioned. Participants could also use the analysis function to track the progress toward their objectives. On Day Three, participants were introduced to a more complicated authentic problem in which they would help a client build an energy-efficient colonial house (see Fig. 1; right) with the capacity of producing as much renewable energy as possible to offset the energy consumed by the house. In addition, this design problem involved constraints such as a total budget of less than $220,000 for a house with a size of 120-160 square meters and a height of 8-10 m.
On Days Four and Five, participants completed a complex design task relevant to their own context-solarizing their school. The storyline of this design project was that the school board planned to replace non-sustainable fossil fuels with solar energy to power the school. Participants were asked to present a solar array design that could provide enough electricity to offset the school's annual energy consumption and had a payback period of fewer than ten years. For this plan, participants were asked to consider numerous factors that might affect their performance (including the cost of solar panels, the offset percentage of the energy generated by solar panels, and years of cost payback). In addition, participants should take account of constraints. For instance, participants should only install solar panels on the roof of buildings and the parking lots (which are known as disturbed spaces), instead of ground, to save space and minimize environmental impacts. Participants also needed to identify the best position of a solar panel with consideration of the angle between the solar panel's upper surface and sunlight for the city where the school was for that specific season). Furthermore, there should be room between solar panels for workers to provide maintenance services or repair these structures. A 3D model of the students' school (see Fig. 2) was provided. Then each participant worked on their own plan using the epistemic tool. In the end, each participant's plan was presented to the class.
We collected four types of data, including participants' design artifacts, log data of engineering design processes, and learning experience survey at the end of the project.
We collected participants' design artifacts in response to the design problem, using solar energy to provide the energy needs for their school. Design artifacts provided an unobtrusive measure of participants' engineering design processes and how they balanced criteria and constraints (Saldana & Omasta, 2017). Participants' design artifacts were gauged by two design criteria: increasing the energy offset and reducing the payback period. Fig. 2 A 3D model of the school campus for Solarizing Your School
The solution should produce as much electricity as possible to offset the annual energy consumed in the school. Additionally, the payback period of the project should be as short as possible. Formulas were provided for the participants to calculate the energy offset and payback years. In addition, participants were expected to take practical design constraints into consideration (e.g., budgets, campus size, feasibility), and they needed to recognize the constraints by themselves. For instance, participants should avoid obstructions such as chimneys, vents, and heating and cooling equipment when placing the solar panels. Design artifacts were collected to provide explanations of the engineering design processes.
The CAD tool logged participants' actions during engineering design processes while completing the final task of solarizing their school. Log data was collected to reconstruct each participant's design trajectories that led to their final artifacts. Sample actions included rotating cameras, adding solar panels, setting cell efficiency for solar panels, removing walls, and running annual energy analysis. In addition, the CAD tool logged the sequences of specific actions. For example, a specific action of "rotating cameras" recorded the timestamps wherein the camera was rotated to a specific direction. We collected log data for 37 out of 38 participants, as one log file was deleted by the participant accidentally.
An open-ended learning experience survey was administered at the end of the project. The survey was developed by our team with the focus on understanding students' engineering design processes and how they balanced criteria and constraints of the given challenges. The survey included three open-ended questions for the participants to articulate their experience in solarizing their school project and share their perception of those activities in this curriculum. A total of 37 participants responded to the survey.
To answer RQ1, we examined the participants' final artifacts to determine if and how they had addressed criteria and constraints during the engineering design processes. To ensure the reliability, two independent researchers with qualitative research expertise and engineering design research experience reviewed all the artifacts separately and then met to discuss any discrepancies. When reviewing artifacts, each researcher noted down if the participant addressed criteria and constraints and provided brief descriptions. For instance, Bryan's artifact (see Fig. 3) was coded by one researcher as "striving for offsetting energy use" and "overlooking design constraints" with a brief description of "Bryan strived for addressing the design criterion of offsetting energy use but overlooked another criterion of payback years as well as design constraints (e.g., budgets, feasibility). Bryan added extra solar panels to areas (e.g., those outside the campus border) unlikely to be installed in practice." Then participants' log data of engineering design processes were used to validate or modify the description. Log data provided an unobtrusive record of participants' actions (e.g., rotating cameras, adding solar panels), which provided additional information to reconstruct participants' engineering design processes. Upon individually reviewing all the design artifacts, the two researchers met to compare the codes and the notes to reach an agreement on clustering participants based on their engineering design patterns.
To answer RQ2, we calculated performance metrics (e.g., percentage of offset energy and years needed for payback) for each group. Note that all solar panels installed in areas deemed unsuitable were removed for analysis. Then separate analysis of variance (ANOVA) was performed to investigate how the pattern of engineering design processes impacted participants' design performance.
To answer RQ3, inductive analysis (Creswell & Creswell, 2017) on participants' responses to open-ended survey items was performed. Particularly, two cycles of coding were conducted. The first cycle of coding aimed to describe participants' experience with the curriculum, utilizing in vivo coding to generate initial codes (Saldaña, 2021). Particularly, in vivo coding sought to describe the phenomenon using participants' own voice (Saldaña, 2021). Sample in-vivo codes included "win-win," "recommending," "fun," and "personally relevant." To ensure the reliability of coding, two researchers individually completed the first cycle of coding and compared their initial codes. Once reaching an agreement on the initial codes, the two researchers scheduled meets to collaboratively complete the second cycle of coding. The second cycle, focusing on pattern coding, sought to elicit patterns (e.g., "building houses," "beyond reading"), categories (e.g., "hands-on," "usable knowledge"), and themes (Saldaña, 2021). The themes were iteratively developed and reviewed by the two researchers until mutual agreement was reached. In the end, a total of three themes were elicited, with direct quotes provided.
To answer this question, we sought to identify various patterns of participants' engineering design processes. By reviewing participants' design artifacts, we assigned participants into three groups based on how they addressed criteria and constraints, including design for performance (n = 19), design for irrelevant goals (n = 9), and design for practice (n = 9).
Design-for-performance students had clear goals of meeting the design criterion of increasing the energy output to offset energy use but overlook another criterion of reducing the payback period. Their solutions featured a large volume of solar panels, even including those placed outside the campus border. Looking at their log data, we found participants in this group frequently switched among observation, formulation, and revision. They recorded numerous actions of adding and editing solar panels for the purpose of increasing energy offset. For example, Bryan, a participant in this group, first turned on a heliodon (i.e.., a virtual device that allows students to see the sun path on different days of the year) to examine which side of the house would get the most sunshine (observation) and then added solar panels to roofs facing south (formulation). Afterward, he added solar panels on flat roofs (formulation) and revised the angles of solar panels to make them face south (revision). He also changed the default solar panel model to those with a high cell efficiency (revision). A panel with a high cell efficiency produces more electricity per panel but may cost more. After all the roofs were covered with solar panels, he created additional solar farms on the ground to further generate energy(formulation). Overall, Bryan's design solution prioritized energy production while having a long payback period (offset 71.53% of consumed energy with 24.93 years of payback). While Bryan's design could generate more energy than other students, it is worth noting that he did not consider some constraints (e.g., solar panels cannot be placed in neighbors' houses or yards) as he added an extra solar farm outside the school campus (see Fig. 3). Note that some participants in this group did not increase energy production despite deploying additional solar panels, because extra solar panels placed outside the designated areas were not calculated for energy production. Overall, the participants in this group placed a higher priority on energy production over constraints when devising solutions to offset their school's energy consumption.
In contrast, the design for practice group paid closer attention to design constraints and practical considerations (e.g., avoiding obstructions such as chimneys, vents, and heating and cooling equipment) leading to potentially more efficient design solutions. Their log data of engineering design processes showed frequent movements among observation, formulation, revision, and analysis. Taking Jennifer's design as an example, she started by adding solar panels on the roof of the main classroom building (formulation) and then ran analysis to check on annual energy production (analysis). Afterward, her engineering design processes mainly followed the routine of observing where to add solar panels (observation), adding solar panels (formulation), revising solar panel directions (revision), analyzing the amount of generated energy (analysis), and showing project cost breakdown (analysis). Particularly, Jennifer aimed to balance criteria and constraints in the process of engineering design. She avoided obstructions and chose suitable locations for solar panel installation. Unlike Bryan, Jennifer installed the solar panels on the roof, leaving pathways for workers to move around instead of covering the entire roof (see Table 2, Row 3). Another difference in their log data was that Jennifer frequently accessed the cost breakdown function. Since cost breakdown was strongly associated with the payback period, Jennifer's consistent actions of analysis (e.g., running energy analysis and showing project cost breakdown) were indicators of her effort to balance energy production and payback periods (offset 39.64% and payback 5.46 years).
The group designing for irrelevant goals prioritized a design solution that was playful, aesthetic, and personally meaningful, with less consideration for criteria or constraints. Consequently, their log data indicated frequent shifts between observation and other actions, with less emphasis on analysis or revision. For example, Carter added solar panels on other roofs (formulation) after randomly placing some solar panels on the roof of the main classroom building (formulation). Then she decided to design them for aesthetic purposes, selecting gray as the color of the solar panels and arranging them on the main building in an oval shape (as shown in Table 2, Row 4). Carter's intentions in making these choices might have been to enhance the visual appeal of the solar panels rather than prioritizing criteria or constraints. Consequently, her design solution unsurprisingly resulted in both long payback periods and lower energy production (offset 34.67% and payback 35.05 years).
Overall, the three groups approached design constraints and criteria in their own manner. Their performance was calculated in terms of the two design criteria (e.g., percentage of energy offset and the years of payback). The descriptive statistics for each group's performance was provided (see Table 3). Particularly, the design-for-practice group had the largest average offset and shortest payback period, but the design-for-performance group recorded the least energy offset and longest payback period.
Shapiro-Wilk test results showed the data for both variables met the condition of normal distribution (p >.05). ANOVA results showed that there were statistically significant differences between the three groups, F(2,34) = 7.89, p =.002, with a moderate effect size (η 2 =.317). The least significant difference (LSD) post hoc test revealed that the design-for-performance group (M = 29.66%, SD = 14.02%) recorded a significantly lower percentage of energy offset than the design-for-practice (M = 48.44%, SD = 10.33%, p <.001) and the design-for-irrelevant-goals group (M = 39.84%, SD = 6.58%, p =.04). However, there was no significant difference between the three groups in the number of years for payback, F(2,34) = 2.63, p =.084. The LSD post-hoc analysis showed that the design-for-practice group (M = 6.67, SD = 3.56) needed a significantly shorter payback period than the design-for-performance group (M = 18.35, SD = 12.56, p =.030).
Theme 1: Students had a Fun Experience with the Hands-on, Action-Oriented Curriculum This them described most of the seventh-grade students enjoyed this action-oriented curriculum. Two categories subsuming this theme were "fun" and "hands-on".
Fun. This category described that the participants would love to recommend this curriculum to others because it was fun. For instance, one participant described that attending this curriculum was "one of the most fun things" that they had done in a science class. Another participant indicated their enjoyment of this course and expressed a desire for others to have the same opportunity.
It is one of the most fun things I have done in a science class. It was useful for learning how to add solar panels to houses and building and it was useful for learning about solar energy. It was also fun to try to put solar panels on houses in the right places to give the houses energy.
I would recommend "Solarize Your World" because I thought the experiment was fun to do. I liked being able to engineer the building of the solar panels and I would like for other people to have fun with that.
Hands-on. This category indicated that a majority of the participants appreciated hands-on activities in this curriculum. One participant learning science by building their own house digitally was "a win-win." Participants also mentioned the advantages of learning about the critical topic of Students Related Science Learning to Real-World Contexts
This theme described the curriculum grounded in scientific inquiry in an authentic context in which students built personal connections and extended their learning beyond. Two categories included "personally meaningful" and "usable knowledge".
Personally meaningful. This category indicates that the curriculum offered students a personally meaningful context. One participant gained a novel perspective on their school and developed a keen interest in the project. Another found it enjoyable to work on a project related to their own school. Also, I liked to see my school from some different angles that I have not been able to before this project or activity. The last project was really fun, where we had to build solar panels on our school.
Usable knowledge. This category focused on the knowledge gained from participating in a curriculum centered on renewable energy adoption. One participant noted that the curriculum provided new insights into "how solar panels work and what solar workers do." This experience helped them connect their gained knowledge to the adoption of renewable energy. Additionally, students acquired knowledge about renewable energy that may not be easily obtained through reading or listening about it. I would recommend Solarize Your World to peers because it really gives me an understanding of how solar panels work and what solar workers do. I would recommend this to other students because it shows that there is more energy than just fossil fuels.
This theme described participants' challenges while attending the action-oriented curriculum. Two categories subsuming this theme were "challenges" and "unaware of contexts".
Challenges. This category denoted students' challenges in following the curriculum. For instance, participants found it challenging to fully benefit from the curriculum when they encountered unresolved issues in one activity, hindering their progress.
What I enjoyed least about this project is during the second activity when I could not get below zero no matter how much I met the requirements and tried.
Unaware of contexts. This category highlights students' limited awareness of the contexts of the given challenge. While the curriculum aimed to help students identify design constraints in the context, the lack of relevant scaffolds made it challenging for participants to be aware of these constraints or understand their significance revolving around constraints.
I didn't like when we had to do the challenge with a certain dimension restriction and height restriction.
This study implemented and evaluated the effectiveness of an action-oriented science curriculum, Solarizing Your School, in seventh-grade science classrooms. The curriculum engaged seventh-grade students in a series of engineering design processes to solve an authentic problem on renewable energy using an epistemic CAD tool, Aladdin (Xie et al., 2023). Particularly, criteria were presented to the students as rubrics to evaluate their solutions to the given problem, but constraints were not specified in this study but awaited to identify and work around those constraints. This study investigated students' profiles in terms of how they balanced the constraints and the criteria during engineering design processes, including design for performance, design for practice, and design for irrelevant goals. The relation between students' patterns of engineering design processes and their design performance was revealed, featuring the design for practice group who showed considerable effort in balancing criteria and constraints. In the end, qualitative inquiry provided insights on seventh-grade students' experience with this action-oriented curriculum.
For the students who designed for performance, they prioritized the goal of meeting the criteria set for this epistemic practice. This cohort of students carefully planned the installation of solar panels in accordance with the criteriaenergy offset and/or return on investment but overlooked the existing non-engineering constraints such as the size of the school campus and the fact that some areas (e.g., the area occupied) were not allowed to dispose of solar panels. Compared with constraints, criteria matter more in determining the level of their performance (Portillo & Dohr, 1994). In addition, given that constraints were not readily revealed to students, those students might just strive for a satisfying performance in line with the criteria (see Bryan's example). We speculate that students prioritized offsetting energy as the performance goal because identifying the payback period requires more advanced mathematical calculation (using the equation) while they could directly receive the energy offset percentage provided by the CAD tool. In contrast, students who designed for practice aimed to search for optimal options that reached a trade-off between criteria and constraints. This cohort of students carefully considered the practical constraints by iteratively undergoing movements among observation, formulation, revision, and analysis (see Jennifer's example). This group's design solutions produced the highest level of energy to offset the school's energy consumption in the shortest cycle of payback. This confirmed that the interplay between criteria and constraints created opportunities for students to take ownership of design processes (Cunningham & Kelly, 2017;Kelly & Cunningham, 2019), leading to potentially the most meaningful solutions. The last cohort of students who designed for irrelevant goals followed their own purposes to solve this problem. Those students seemingly did not address the interplay between constraints and criteria but focused more on their preferred solutions (e.g., designing a particular shape as Carter), which may not be an optimal solution to this problem.
This study also explored the impact of the action-oriented curriculum on middle school students' agentic actions of adopting renewable energy. The qualitative data gathered from students' learning experience surveys underscored the advantages of the action-oriented curriculum, particularly in fostering usable knowledge that students could apply to future efforts in renewable energy adoption (Busch et al., 2019;Li & Monroe, 2019;Tayne et al., 2021). This can be inferred from constructionist learning opportunities afforded in this action-oriented curriculum through hands-on activities and personally relevant contexts (Blikstein & Wilensky, 2010;Kafai & Burke, 2015;Kafai & Resnick, 1996;Kelly & Cunningham, 2019;Mumba et al., 2024). Guided by the constructionist root, hands-on activities allowed students to create meaningful design artifacts by actively engaging in engineering design processes using epistemic tools (Atman et al., 2007;Kafai & Burke, 2015). In addition, solving real-world problems in personally relevant contexts not only deepened students' sense of cross-disciplinary knowledge but also facilitated the development of socially and culturally relevant knowledge, which is at the core of constructionist learning (Blikstein & Wilensky, 2010;Kafai & Burke, 2015;Kafai & Resnick, 1996). By connecting crossdisciplinary knowledge to meaningful, real-world applications, the curriculum empowered students to take ownership of their learning while making a tangible impact on their environment.
The findings of this study have implications for environmental science education research. The results confirmed the importance of promoting the development of students' ability to navigate the trade-offs between criteria and constraints in engineering design practices (Akins & Burghardt, 2006;Atman et al., 2007). Furthermore, this study highlighted the importance of providing students with clear information about constraints to encourage them to consider a wider range of factors when designing solutions. By incorporating these factors into engineering design practices, researchers can help students develop a more holistic approach to problem-solving that takes into account the interconnectedness of environmental, social, and economic factors (Busch et al., 2019).
Moreover, this research provided practical implications on affording opportunities for students to develop secondary students' agentic actions towards renewable energy deployment via engineering design practices. First, K-12 educators and designers might consider embedding environmental science projects in an authentic setting so that students can leverage their own knowledge as a resource to solve the problem and then build a knowledge artifact that is meaningful to them. Second, K-12 educators might prompt students to further connect their epistemic practices to the local matters that they may contribute to engendering a positive change. This may allow students to invest in the effort to modify the structure that may constrain the actions. Third, K-12 educators may provide prompts that help students be aware of constraints, as identifying constraints is not a universal skill for students at a younger age. Therefore, pedagogical prompts may be needed for students to figure out and then work around the constraints when developing their solutions to practical problems.
While this study represents a promising step in the direction of advancing our understanding of agency in engineering design for science education, it is limited by its nature as a descriptive study. So, the current preliminary findings could not make a conclusive argument on how students balanced criteria and constraints in engineering design. In addition, the interpretation of the findings highly relied on our own subjective understanding of the context and the data, so further validation of the findings in a relatively larger dataset is needed. Moreover, while the curriculum was designed to align with constructionist principles, students' perceptions of the artifact's personal meaningfulness or their sense of epistemological agency were not collected, which limited the capacity of interpreting findings grounded in the constructionist framework.
Future research might consider adding additional complementary sources of data, such as interviews on learning experience and video recordings of students' engineering design sessions, to illustrate students' epistemic practices of climate-motivated learning. In addition, future research may include measures to capture students' reflections on the personal relevance of their design artifacts and their epistemic agency throughout the learning process, thereby strengthening the theoretical coherence and empirical grounding of constructionism-informed curricula. Moreover, future research may consider investigating constraints or factors that go beyond just technical or engineering constraints and include humanistic and ethical considerations that are crucial for developing environmentally responsible solutions.