Shifting Students’ Perceptions About Homelessness: Quantitative Assessment of a Project-Based Approach

Abstract

Although engineering institutional bodies uphold public welfare and the impact of engineering on people and society, engineering curricula rarely scaffold students to connect their technical learning with sociotechnical perspectives. This paper describes a project-based learning approach where engineering students engaged with issues faced by people experiencing homelessness to better understand the sociotechnical nature of effective, user-centered, engineering design. We conducted a quantitative assessment to determine how well and in what ways the project-based learning curriculum shifted students’ perceptions about homelessness. We collected pre-/post-survey data from students on 21 statements about their perceptions and attitudes about homelessness prior to and after an engineering project with a focus on homelessness in San Diego, CA, USA. The study aimed to measure the effectiveness of the course/project on shifting students’ perceptions from myths about homelessness towards reality, which supported the course objectives regarding diversity, inclusion, and social justice. We found that, from data from 166 students over 8 semesters, students’ perceptions had statistically significant (p < 0.05) shifts in five survey statements, which regarded beliefs about the personal choices or perceived moral decisions of those experiencing homelessness, and that students were able to more strongly identify with an engineer’s duty to care for those experiencing homelessness.

1. Introduction

Two institutional bodies that drive engineering morals, values, and change include engineering accreditation, such as the Accreditation Board for Engineering and Technology (ABET), and engineering professional societies, such as the American Society of Mechanical Engineers among many others. Annually, ABET releases a set of student outcomes (SOs) that it uses to assess engineering students and the programs from which they graduate (Criteria for Accrediting Engineering Programs, 2025–2026). In its latest iteration (established 2019), three SOs in particular have led to some consternation across engineering faculty. These SOs state that upon graduation students should have the following:

• SO2: an ability to apply engineering design to produce solutions that meet specified needs with consideration of public health, safety, and welfare, as well as global, cultural, social, environmental, and economic factors.
• SO4: an ability to recognize ethical and professional responsibilities in engineering situations and make informed judgments, which must consider the impact of engineering solutions in global, economic, environmental, and societal contexts.
• SO5: an ability to function effectively on a team whose members … create a collaborative and inclusive environment….

On the other hand, engineering professional societies each have their own Codes of Ethics, which are used to guide engineers’ behavior as they produce work that is used by people and affects society. Across several of the largest disciplines’ societies, the American Society of Mechanical Engineers (ASME.org, n.d.; Criteria for Accrediting Engineering Programs, 2025–2026), the Biomedical Engineering Society (2024 BMES Annual Meeting Policies & Code of Conduct—Biomedical Engineering Society, n.d.), the American Society of Civil Engineers (Code of Ethics, n.d.-a), the Institute of Electrical and Electronics Engineers (IEEE Code of Ethics, n.d.), and the National Society of Professional Engineers (Code of Ethics, n.d.-b), the first tenet of each Code of Ethics is for a professional engineer to use their knowledge and skills for the enhancement of human welfare, and/or to hold paramount the safety, health, and welfare of the public.

The ABET outcomes described above emphasize the complexity of real-world engineering and the impact it has on people more explicitly and in a more integrated way than previous iterations have. Similarly, the Codes of Ethics all emphasize human welfare/welfare of the public. It is evident that both institutional bodies prioritize engineering’s impact on people; however, the way engineering is taught rarely discusses the impact of engineering technologies on society (Mejia et al., 2018). One study found that, despite many students entering engineering with a commitment to public welfare, engineering’s “culture of disengagement” leads students to be disenfranchised when they graduate (Cech, 2014). One reason for this decline in student concern for public welfare may be the lack of exposure and practice students receive around how human life and decision-making are involved in engineering (Mejia et al., 2021). Another reason is the framing of engineering as apolitical and decontextualized. Several scholars have studied the role of objectivity in engineering culture that has led to engineering’s problem with decontextualization (Chen & Wodin-Schwartz, 2019; Downey, 2015; Lucena & Leydens, 2015; Moseley, 2017). Others have also found that a sociotechnical lens can help students tie their technical skills to real-world problems. For example, Chen et al. (2023) found that framing engineering content as sociotechnical increased student engagement and helped students to think more deeply about their own goals as future engineers. By directly relating engineering to its sociotechnical and public welfare applications, we hope to combat the culture of disengagement in our own curricula.

Project-based learning (PBL) courses are especially well suited for the integration of sociotechnical context. We consider ‘sociotechnical’ to indicate the complex ways in which the social and technical aspects of these issues are interconnected. Successful PBL courses often base a project on a meaningful problem and operationalize the project with an open-ended, engaging question (de Graaff & Kolmos, 2007). In addition, sustained inquiry, where students find and use resources of their own over time; authenticity of a real-world context, process, and impact that connects to students’ own concerns, interests, and identities; and student voice and choice in how they use their time and the products they create are key elements to creating an engaging project (Buck Institute for Education, 2015). Engineering design projects, particularly those that take user contexts into account, lend themselves especially well to PBL when students are asked to consider sociotechnical aspects that influence their technical design solutions. This study uses PBL to integrate local issues faced by people experiencing homelessness into an engineering design course to garner student engagement. Homelessness is an escalating global issue that has worsened over the last few years due to economic instability, housing shortages, and other systemic factors. Homelessness has become a growing crisis in recent years, particularly in the United States and in cities like San Diego, CA, where rising housing costs and insufficient shelter capacity have contributed to an increasing unhoused population. As the crisis grows, it is increasingly becoming an engineering challenge that requires innovative solutions in areas such as sustainable housing, infrastructure design, and resource accessibility. Given this perspective, our research explores how project-based learning can equip future engineers with the skills and social awareness needed to tackle complex societal issues like homelessness.

Existing literature has explored homelessness from social, economic, and public health perspectives. However, there is a gap in research that explicitly connects engineering education—particularly project-based learning—with addressing homelessness as both a technical and social challenge. Our study fills this gap by examining how project-based learning can cultivate engineers who are not only technically proficient but also socially responsible and equipped to develop empathy toward these issues. Recent comments from technocrats claiming that empathy has been the demise of the Western civilization demonstrate the idea that technical and social work must be taught separately. This rhetoric can be dangerous for engineering education because it reinforces an outdated, technocratic view of engineering that prioritizes efficiency, logic, and innovation at the expense of human-centered considerations. This perspective risks devaluing the ethical and social dimensions of engineering, which are essential for addressing complex global challenges such as homelessness.

It is important to note that the existing engineering education framework in the U.S. continues to largely prioritize Western, White, and male epistemologies (Leydens & Lucena, 2018; Momo et al., 2020; Riley, 2008). These dominant perspectives shape the way engineering students engage with community-based projects, often privileging top-down, technocratic solutions rather than collaborative, culturally responsive approaches. By interrogating these underlying tendencies to see empathy as an interfering force to conducting transformative engineering work, our research seeks to challenge traditional engineering education paradigms and reimagine project-based learning as a tool for ethical, community-centered reflection. Through a localized case study on homelessness, we aim to demonstrate how engineering students make sense of social issues. This work contributes to a growing body of scholarship advocating for a more inclusive, justice-oriented engineering education that critically examines its own epistemological foundations and redefines its role in addressing systemic inequities.

1.1. Course Context and Institutional Resources

The study described in this paper is based within an undergraduate engineering course, User-Centered Design, that is required of students in all engineering majors at the University of San Diego (USD). The course is typically taken in a student’s second or third semester in the engineering curriculum. As a result of the campus environment, the User-Centered Design course evolved over time to be unique in its objective to cover topics including justice, power, intersectionality, and privilege and oppression (Mejia et al., 2018).

The unique nature of this User-Centered Design course grew out of an opportunity in the School of Engineering to align engineering requirements with the institution’s mission and initiatives. The University of San Diego is a primarily undergraduate, liberal arts, contemporary Catholic institution that embraces the Catholic moral and social tradition in its mission and values, which emphasizes the importance of sustainability and social justice. USD is also designated as an Ashoka U “Changemaker Campus”, a recognition of our university’s commitment to finding sustainable solutions to the world’s most pressing problems. Furthermore, the university’s last strategic plan, “Because the World Needs Changemakers”, aligned university-wide curricula with the values of the institution. In keeping with the mission, all undergraduate students are required to complete two courses with the diversity, inclusion, and social justice (DISJ) Core flag, which seeks to help students develop critical self-reflection and the ability to analyze the complexities of social constructions in everyday life.

In Spring 2017, the engineering faculty body voted to designate the User-Centered Design course as a lower level DISJ course. While the impetus for the change was to ensure that students were seeing the relevance of DISJ topics within an engineering context, the benefit was two-fold, as it also led to one fewer additional (i.e., external to engineering) graduation requirement for the already lengthy engineering curricula. The integration of the DISJ topics with engineering design proved to be challenging, and the progression of the course is detailed elsewhere for the reader (Chen et al., 2019, 2020; Lord et al., 2018; Mejia et al., 2018). In Fall 2019, after reviewing the efficacy of the course, the primary instructors decided to shift the final design project (which was previously situated around building solar water heaters with a context divorced from the local area) towards homelessness (Chen et al., 2020; Mejia et al., 2021). Not only does San Diego rank as the sixth highest for rates of individuals experiencing homelessness in the United States (Johnson et al., 2025), providing a pressing context for a locally visible issue, but the context allowed for topics around social justice to be better tied to course content. In addition, USD’s Changemaker Hub—which is described as both a place and a process that highlights the interconnections of people, activities, and organizations that constitute changemaking—launched an Urgent Challenges Collective initiative in 2019 to study and address homelessness and food insecurity in our local community. The timing of the homelessness initiative aligned well with the redevelopment of the course to provide activities, events, and resources on campus to help students engage with and learn about the issues faced by individuals experiencing homelessness.

Based on the required learning objectives for the DISJ designation, the full list of course objectives are presented below. By the end of the course, students will have achieved the following:

• Have reflected on and be able to communicate about their own identity and personal experiences (i.e., privilege and disadvantage) in relation to others.
• Have demonstrated empathy for users by describing how users’ experiences may be influenced by societal norms around the intersectionality of issues such as race, ethnicity, gender, age, physical ability, immigration status, literacy, and language.
• Be familiar with qualitative research methodologies (e.g., interviews, observation, and immersion) to engage users and identify user issues related to intersectionality.
• Be able to demonstrate empathy and other mindsets that support a user-centered approach to engineering design.
• Be able to analyze and design consumer products by applying principles of design.
• Be able to translate customer needs to product specifications.
• Be able to develop a plan to complete a design task.
• Be able to use prototyping techniques and iteration to develop design ideas.
• Be able to elicit feedback from users to improve designs.
• Be able to describe and practice attributes of effective teams and team members.
• Be able to collaborate with people, especially users, throughout a design process to develop user-oriented concepts, products, or services.
• Be able to communicate design solutions to various stakeholders.

1.2. Collaboration with Think Dignity

As a part of the course redesign, the primary instructors cultivated a relationship with a local non-profit organization, Think Dignity, which is a legal advocacy group that strives to provide services and programs that focus on advancing the basic dignity for those living on the streets (Think Dignity, n.d.). One of the goals of the partnership was to provide a more realistic situation for students to engage with the DISJ content, local to San Diego. As a part of the collaboration, a staff member from Think Dignity was invited every semester to give a guest lecture to the engineering students. The talk included statistics about homelessness both nationally and locally for the county and presented different myths regarding homelessness that were often surprising to students. For example, four myths and their realities are provided below (Watanabe, 2019):

• Myth: A large majority of homeless individuals have issues with substance abuse that prevents them from maintaining stability in their lives. (Reality: Only 3% of homeless individuals have these issues).
• Myth: A large majority of homeless individuals have a severe mental illness. (Reality: Only 4% of homeless individuals have these issues).
• Myth: A large majority of homeless individuals abuse free services or welfare. (Reality: Less than 20% of homeless individuals know about/are qualified for free services or welfare).
• Myth: Homeless people are criminals. (Reality: Homeless people are more likely to be victims of crime).

The presentation also often included topics such as compounding legal ramifications that individuals experiencing homelessness often faced and were left to navigate on their own, as well as presenting resources that these individuals could engage with, which often inspired students’ project topics.

This paper describes a four-year study that investigates, “How well and in what ways did the project-based learning curriculum contribute to a shift in students’ perceptions about homelessness?” While this paper focuses on quantitative data, the reader is directed to Chen et al. (2020) for a discussion of the qualitative data collected that describes how the project-based approach affected students’ learning outcomes around the idea that engineering alone cannot solve complex socio-political problems. Together, our overall goal was to use the context of homelessness embedded into PBL to help students to meet course objectives around developing empathy and learning about social justice topics within an engineering context. Our hope with this project-based approach and community collaboration was that we could dispel myths that students might hold about homelessness in order to encourage more students to become more compassionate engineers. The context of homelessness helped to better integrate the diversity, inclusion, and social justice course elements with the engineering design concepts (whereas previous iterations of the course had modules on topics such as race and privilege separated from the engineering design project), serving as a conduit for achieving the course objectives and reinforcing the idea that engineering is a sociotechnical endeavor.

2. Methods

At the start of each semester, as a part of the first homework assignment, students in the class were provided with a pre-survey to gather data on their preconceptions about homelessness prior to the information discussed throughout the course. The survey included 21 statements about homelessness and individuals experiencing homelessness, where students were asked to indicate the option that most closely represented their views using a Likert scale with five options, ranging from strongly disagree to strongly agree. The prompt reassured students that there were no right or wrong answers, and that they did not need to spend long on each statement, as often their first response would be the most accurate representation of their perceptions. At the conclusion of the survey, students were asked to type the last four digits of their cell phone number so that future responses (identical post-survey) could be linked anonymously.

The 21 statements about homelessness were adapted from the Health Professional Attitudes Towards the Homeless Inventory (HPATHI)1, developed by Buck et al. for use in the medical community (2005). The HPATHI targets three subscales: (1) Personal Advocacy, which reflects a personal commitment to work with people experiencing homelessness, (2) Social Advocacy, which reflects society’s responsibility to care for the homeless population, and (3) Cynicism, which reflects a negative attitude and sense of futility in working with the homeless (Crow, 2013). The HPATHI is a validated instrument that measures medical providers’ attitudes toward the homeless in the hopes that results can point towards the design and implementation of educational activities that foster more compassionate homeless health care (Buck et al., 2005).

Medicine, in particular, is a profession that attempts to develop an ethic of service to the underserved in its students; however, studies have indicated that progression through the medical curriculum is linked with a rise in cynicism, a decrease in empathy, and a decreased interest in caring for the poor (Fine et al., 2013). The undergraduate engineering curriculum presents a similar conundrum: while engineering, broadly, upholds and values the welfare of the public and the impact of engineering and technology on people and society, the traditional engineering curriculum does not tend to scaffold students in learning and applying these values. In fact, Cech (2014) found that the culture of engineering education causes students to become more disengaged with the public good over time.

While the original HPATHI instrument presents 19 statements, the survey we used for this project presented students with 21 statements that were reworded for an engineering context. Of the original HPATHI statements, we kept and/or adapted 9 and added 12 more.

Table 1. Statements used in pre-/post-survey with themes labeled.

#	Statement Text	Theme
		T1	T2	T3	T4
Q1	Homeless people do not choose to be homeless.		X
Q2	Nearly all homeless people are drug addicts.		X
Q3	You only need to learn about homelessness if you want to be a social worker.			X
Q4	Homeless people are victims.		X
Q5	Homeless people are rude.		X
Q6	Engineers have a duty to care for the homeless.			X
Q7	Homeless people are aggressive.		X
Q8	Homelessness is a major problem in our society.	X
Q9	Homelessness is a self-inflicted state.		X
Q10	Homelessness is not a health issue.	X
Q11	The government should not waste its resources on the homeless.				X
Q12	I entered engineering because I want to help those in need.			X
Q13	People make themselves homeless to get government benefits.		X
Q14	No one in this country has to ‘sleep rough’.	X
Q15	The government should spend more money on providing housing.				X
Q16	Alcoholism is a personal weakness.		X
Q17	I entered engineering because I was good at math and/or science.			X
Q18	Homelessness is not a significant problem in San Diego.	X
Q19	Engineers should address technical and social problems.			X
Q20	The government should spend more money on the care of the homeless.				X
Q21	Homelessness can be solved using technology.			X

shows the 21 statements used in our engineering study. In particular, our survey targeted four subscales (themes):

• T1: Beliefs about homelessness being a real (as opposed to trivial or imaginary) issue.
• T2: Beliefs about individuals experiencing homelessness, reflecting a bias against personal choices or moral deficiencies.
• T3: Relevance of homelessness for engineers/engineering.
• T4: Government responsibility towards addressing homelessness.

Gjersing et al. (2010) make the case that previously validated instruments are not necessarily valid in another time, culture, or context. Given that the original HPATHI instrument was written specifically for medical professionals, it was not only appropriate but necessary for us to adapt the statements to our own discipline. By adapting the instrument, the study preserves the integrity of the original subscales—Personal Advocacy, Social Advocacy, and Cynicism—while making them more suitable for the target engineering population. This adaptation seeks to retain the validated structure and intent of the HPATHI while adjusting the language and context of the statements to align with an engineering context. Since the HPATHI has already demonstrated reliability in measuring attitudes toward homelessness, modifying it for a different but related audience allows for meaningful comparisons while maintaining the rigor of the original instrument. However, we did conduct a reliability check and determined that the Cronbach’s alpha = 0.553, indicating some internal consistency. While this indicates medium internal consistency, the instrument was retained due to its conceptual relevance and prior validation in other domains. The medium reliability may be attributed to contextual differences between disciplines, potentially affecting how participants interpreted certain items. This limitation is acknowledged, and future work should consider refining or revalidating the instrument for use with engineering populations. Furthermore, the adaptation supports the goal of assessing attitudes to inform educational interventions, ensuring that findings can still contribute to strategies for fostering empathy and advocacy in homelessness-related work.

While the pre-/post-survey was identical for all instructors that participated, every instructor inevitably has their own teaching style and may have delivered course content differently. Over the course of 9 semesters, 7 instructors (named A–F below) had their students participate in the study. Course enrollments are typically limited to 24 students total per section.

Table 2. Student participation rate by semester.

Semester	Total Enrollment by Instructor	Total Student Participation (S)	Usable Responses (N)	Success Rate
Fall 2019	Instructor A—24 Instructor B—24 Instructor C—11	59	38	64%
Spring 2020 *	Instructor B—14 Instructor C—17 Instructor D—7 Instructor D—21	59	38	64%
Fall 2020 **	Instructor B—9	9	6	67%
Spring 2021 **	Instructor B—24 Instructor C—21	45	18	40%
Fall 2021 ***	Instructor A—20 Instructor B—22	0	0	0%
Spring 2022	Instructor A—22 Instructor A—20	42	24	57%
Fall 2022	Instructor E—10	10	7	70%
Spring 2023	Instructor B—11 Instructor F—17	28	18	64%
Fall 2023	Instructor A—24	24	17	71%
	Total	S = 276	N = 166	60%

* Spring 2020 Switched to Emergency Remote Teaching halfway through the semester. ** Fall 2020 and Spring 2021 Entirely online due to global pandemic. *** Fall 2021 Did not collect data.

shows the total student enrollment in each section with the participating instructors, alongside the resulting survey success rate. Data collection was concluded after the Fall 2023 semester due to the course developers no longer being consistently assigned to teach the course.

A total of 318 students were enrolled in the version of the course that explored homelessness (there were some instructors in the school who offered different versions of the course). However, the instructors opted not to collect data in Fall 2021, resulting in only 276 students who were instructed to complete the pre-/post-survey. Of the 276 students, only 166 adequately completed the pre-/post-surveys, resulting in a total completion rate of 60%. The survey was conducted in Qualtrics, which enabled us to match the unique ID numbers of each entry (students’ self-reported last four digits of their cell phone number). To condition the data down to the final N = 166, incomplete responses were removed, and only unique pairs were used (i.e., one unique ID appeared only twice in one semester—once for pre and once for post). In cases where students mistakenly completed a pre- or post-survey multiple times, the later entry was used. In later semesters, several instructors moved towards making the post-survey an extra credit opportunity to reduce excess assignments at the end of the semester during final exams. However, it does not appear that this shift in requirements played a significant role (either positively or negatively) in the success rate of the post-survey.

There are also changes between each semester that are important to note, which may have affected students’ learning outcomes and shifts in their perceptions of homelessness. First, the scope of the project changed slightly from semester to semester, as described in

Table 3. Description of final project prompt by semester.

Semester	Final Project Description
Fall 2019	This project will be aiming to assist Think Dignity with their Fresh Start Showers Program. Design a solar water heater that interfaces with Think Dignity’s mobile shower unit.
Spring 2020 *	Design a prototype or process aimed at an unmet need associated with COVID-19 and homelessness.
Fall 2020 **
Spring 2021 **
Fall 2021 ***	Investigate the issues facing our unhoused neighbors and create a design that addresses an unmet need in their community. (open-ended)
Spring 2022	Investigate the issues facing our unhoused neighbors and create a design that addresses an unmet need in their community. Choose one aspect of hygiene to tackle: bathing, laundry, toilet access, feminine care, other. Note: all of these depend on access to fresh clean water. What is an engineering solution you could create to help with this?
Fall 2022	Investigate the issues facing our unhoused neighbors and create a design that addresses an unmet need in their community. (open-ended)
Spring 2023	Investigate the issues facing our unhoused neighbors and create a design that addresses an unmet need in their community. Choose one aspect of hygiene to tackle: bathing, laundry, toilet access, feminine care, other. Note: all of these depend on access to fresh clean water. What is an engineering solution you could create to help with this?
Fall 2023	Investigate the issues facing our unhoused neighbors and create a design that addresses an unmet need in their community. San Diego instated an “Unsafe Camping Ban”, which prohibits outdoor sleeping in any open areas or near schools and transit hubs. In conjunction, two new “Safe Sleeping Sites” have been opened as temporary “housing” for those experiencing homelessness. What is an engineering solution you could create to help with these two new initiatives?

* Spring 2020 Switched to Emergency Remote Teaching halfway through the semester. ** Fall 2020 and Spring 2021 Entirely online due to global pandemic. *** Fall 2021 Did not collect data.

.

Additionally, the participation of our invited speaker from Think Dignity varied from semester to semester. We believe this guest talk played a major role in shifting students’ perceptions about homelessness, so the format and requirement of attending the talk may have influenced the results of our study. For the most part, the guest talk was held over Zoom during “dead hours”, which spans 12:15–2:20 p.m. on Tuesdays and Thursdays during which no classes are scheduled (intending to give students time for extracurricular activities such as meetings through student organizations). The primary reason for this was that there were multiple sections of students, and scheduling the talk during these hours reduced the coordination and time commitment from the guest speaker. However, using this time block also meant that not all students could or did attend the talk, due to other engagements. In every semester, the talk was recorded and all students were required to submit a written reflection about the talk, but those that did not attend were not able to ask questions and engage in exploring potential project topics more deeply with the speaker.

Table 4. Variations in Guest Speaker Presentation.

Semester	In-Person vs. Online Talk	Dead Hours vs. In-Class
Fall 2019	In-Person	Dead Hours
Spring 2020 *	Online	Dead Hours
Fall 2020 **	Online	In-Class
Spring 2021 **	Online	Dead Hours
Fall 2021 ***	Online	Dead Hours
Spring 2022	Online	Dead Hours
Fall 2022	In-Person	In-Class
Spring 2023	In-Person	Dead Hours
Fall 2023	In-Person	In-Class

* Spring 2020 Switched to Emergency Remote Teaching halfway through the semester. ** Fall 2020 and Spring 2021 Entirely online due to global pandemic. *** Fall 2021 Did not collect data.

shows the semesters in which the guest speaker presented in-person on campus versus online via Zoom, and if the presentation was during class time when all students were present or during dead hours, so that multiple sections could attend at the same time.

Notably, in Fall 2020, Fall 2022, and Fall 2023, there was only one section of this version of the course offered; in these cases, the guest speaker was invited to give her talk within the boundaries of the class meeting time. (In Fall 2020, this meant still on Zoom due to completely remote teaching due to the global pandemic, but in Fall 2022 and Fall 2023, the guest speaker was invited to attend class in person).

3. Results

We performed statistical analysis on the students’ responses on the pre-/post-survey by adapting the Likert scale to be a numerical scale, with 1 corresponding with ‘strongly disagree’ and 5 with ‘strongly agree’. To determine whether or not we could reasonably group all survey data collected across eight different semesters (data was not collected in Fall 2021), we conducted a one-way ANOVA based on semester. The Tukey post hoc analysis found no significant differences in 33 out of 42 prompts (the 21 statements were each presented twice, once for pre and once for post) between semesters. Of the nine comparisons that did show significance (Q8 pre and post, Q11 pre, Q15 post, Q17 pre, Q19 post, Q20 post, and Q21 post), only minor differences were found between a few semesters, and no systematic patterns were revealed across the prompts (i.e., certain semesters were not driving the identified differences). Accordingly, due to the limited inter-semester differences, all pre- and post-data across semesters for each prompt are pooled and analyzed for aggregate differences.

Table 5. Descriptive group statistics for each prompt across 8 semesters, with paired t-test sample statistics. 1 = strongly disagree and 5 = strongly agree. Statistical significance was set to p ≤ 0.05.

Statement Text		Mean	Std. Deviation	Std. Error Mean	p Value
Q1: Homeless people do not choose to be homeless.	Pre	3.63	0.987	0.077	0.006
	Post	3.86	0.974	0.076
Q2: Nearly all homeless people are drug addicts.	Pre	2.36	0.915	0.071	0.000
	Post	2.01	0.809	0.063
Q3: You only need to learn about homelessness if you want to be a social worker.	Pre	1.80	0.840	0.065	0.896
	Post	1.79	0.907	0.070
Q4: Homeless people are victims.	Pre	3.28	0.837	0.065	0.073
	Post	3.41	0.860	0.067
Q5: Homeless people are rude.	Pre	2.40	0.794	0.062	0.455
	Post	2.36	0.747	0.058
Q6: Engineers have a duty to care for the homeless.	Pre	3.40	0.927	0.072	0.031
	Post	3.55	0.931	0.072
Q7: Homeless people are aggressive.	Pre	2.46	0.799	0.062	0.367
	Post	2.40	0.755	0.059
Q8: Homelessness is a major problem in our society.	Pre	4.52	0.620	0.048	0.231
	Post	4.45	0.743	0.058
Q9: Homelessness is a self-inflicted state.	Pre	2.64	0.908	0.070	0.000
	Post	2.43	0.812	0.063
Q10: Homelessness is not a health issue.	Pre	2.16	0.978	0.076	0.820
	Post	2.17	0.947	0.073
Q11: The government should not waste its resources on the homeless.	Pre	1.81	0.808	0.063	0.100
	Post	1.72	0.769	0.060
Q12: I entered engineering because I want to help those in need.	Pre	3.73	1.068	0.083	0.419
	Post	3.79	1.072	0.083
Q13: People make themselves homeless to get government benefits.	Pre	1.88	0.785	0.061	0.239
	Post	1.81	0.814	0.063
Q14: No one in this country has to ‘sleep rough’.	Pre	2.73	1.257	0.098	0.312
	Post	2.66	1.352	0.105
Q15: The government should spend more money on providing housing.	Pre	3.60	0.978	0.076	0.399
	Post	3.65	0.953	0.074
Q16: Alcoholism is a personal weakness.	Pre	2.89	1.246	0.097	0.037
	Post	2.71	1.096	0.085
Q17: I entered engineering because I was good at math and/or science.	Pre	3.78	0.987	0.077	0.465
	Post	3.82	1.023	0.079
Q18: Homelessness is not a significant problem in San Diego.	Pre	1.68	0.860	0.067	0.373
	Post	1.61	0.807	0.063
Q19: Engineers should address technical not social problems.	Pre	2.16	0.867	0.067	0.683
	Post	2.19	0.850	0.066
Q20: The government should spend more money on the care of the homeless.	Pre	3.69	0.887	0.069	0.579
	Post	3.72	0.919	0.071
Q21: Homelessness can be solved using technology.	Pre	3.38	0.798	0.062	0.191
	Post	3.29	0.902	0.070

shows the group descriptive statistics for each prompt for the data pooled across semesters. Figure 1 shows a diverging stacked bar chart to visualize the main patterns and statistical differences in the data. Notably, a paired t-test revealed significant shifts in student perspectives in Statements 1, 2, 6, 9, and 16 (Table 5, Figure 1), with effect sizes (i.e., Cohen’s d) of −0.22, 0.40, −0.17, 0.29, and 0.16, respectively.

4. Discussion

Our goal with this study was to evaluate how the PBL course elements contributed to shifts in students’ perceptions about homelessness in order to analyze the impact of the course in meeting its objectives related to diversity, inclusion, and social justice. Of the statements provided in the pre-/post-survey, the results of the paired t-test revealed statistically significant shifts in student responses from before and after the course interventions in the following five (of twenty-one) statements, suggesting that the intervention had a measurable impact on certain attitudes or beliefs.

• Q1: Homeless people do not choose to be homeless.
• Q2: Nearly all homeless people are drug addicts.
• Q6: Engineers have a duty to care for the homeless.
• Q9: Homelessness is a self-inflicted state.
• Q16: Alcoholism is a personal weakness.

Notably, Statement 2 showed a moderate positive effect size (Cohen’s d = 0.40), indicating a meaningful shift in student responses. Statement 9 also showed a small-to-moderate effect (d = 0.29), while the remaining significant statements reflected smaller shifts. The negative effect sizes for Statements 1 and 6 (d = −0.22 and −0.17, respectively) may point to nuanced changes in student interpretation or unexpected reactions that warrant further exploration. Overall, these findings suggest that while the instrument had limitations in internal consistency, it was still sensitive enough to detect changes in specific areas of student thinking. This supports the potential value of the intervention, though further refinement of the instrument for use in engineering contexts is recommended to strengthen future interpretations.

Interestingly, Q1, Q6, Q9, and Q16 were all related to Theme 2: Beliefs about individuals experiencing homelessness. These questions captured preconceived notions and stereotypes, such as the idea that homelessness is primarily the result of poor life choices or a lack of moral responsibility. In all four cases, students’ post-survey responses shifted to a more sympathetic perspective following the class project. We believe these shifts in the perceptions of homelessness were tied to the project activities, such as the guest talk that debunked the myths regarding homelessness and the additional research students were asked to conduct. The requirement for students to conduct in-depth research about the specific need they identified ensured that their designs were user-centered, encouraging them to consider the lived experiences of individuals facing homelessness. These activities compelled students to move beyond surface-level assumptions and engage critically with the intersectionality of privilege, disadvantage, and social inequities. All of these shifts towards a more sympathetic perspective were positively correlated with our course learning objectives (e.g., demonstrating empathy, identifying user issues related to intersectionality, reflecting on how their own personal experiences may be different than others with different privileges and disadvantages).

The last statement in which we observed a shift in student perceptions was Q6: Engineers have a duty to care for the homeless, which was related to Theme 3: Relevance of homelessness for engineers/engineering. This outcome aligned with the overarching goal of the class to help students see that engineering as a whole is a sociotechnical endeavor. By framing engineering as a discipline with responsibilities that extend beyond technical problem-solving, the course encouraged students to see themselves as agents of positive societal change. Furthermore, although this course does not present ABET outcomes or professional societies’ codes of ethics to the students, this shift in perception also aligned with these institutional bodies’ goals for students to have “an ability to apply engineering design to produce solutions that meet specified needs with consideration of public health, safety, and welfare”, to have “an ability to recognize ethical and professional responsibilities in engineering situations and make informed judgments, which must consider the impact of engineering solutions in … societal contexts”, and to “use their knowledge and skill for the enhancement of human welfare.” This shift toward a more socially conscious view of engineering highlights what the course sought to do in terms of fostering a sense of purpose and responsibility among students. The results suggest that students are beginning to see themselves not only as problem-solvers but also as contributors to addressing pressing societal challenges, such as homelessness. This evolution in their perspectives underscores the importance of incorporating sociotechnical considerations into engineering education to prepare students for the complex challenges they will encounter in their professional lives.

Other prompts in Theme 3 related to the relevance of homelessness to engineering (Q3, Q12, Q17, Q19, and Q21) did not reveal significant shifts in student perceptions. The low pre-survey mean for Q3 (1.8), and above-neutral means for Q6 (3.4) and Q12 (3.73) indicated that students entered the course with attitudes already aligned with our desired student outcomes—possibly due to the student population that self-selected into our university, based on the values of the institution. If anything, the lack of shift in these prompts can be perceived as a positive outcome, suggesting that this course did not contribute to the “culture of disengagement” described by Cech (2014). We are hopeful that our unique take on user-centered design helped students to see why it is important for them to continue caring about social issues as engineers.

Two prompts in which we may have expected significant shifts but did not were Q19 and Q21. For a course that emphasizes the role of engineering as a sociotechnical endeavor, it was surprising that we observed a significant shift in Q6, which specifically tied engineering to “a duty to care for the homeless”, but not significant shifts in engineering’s relation to broader sociotechnical issues (outside of homelessness). This observation suggests an opportunity for the course to expand its focus and more effectively showcase the breadth of sociotechnical issues that engineers can address through their work. Additionally, the course could place greater emphasis on the importance of interdisciplinary collaboration, particularly with social scientists, to achieve truly user-centered design. Highlighting the value of diverse expertise and perspectives can help students appreciate that engineering solutions are most effective when informed by a deep understanding of social contexts and human experiences. This approach not only enriches the learning experience but also fosters empathy, empowering students to become more thoughtful and inclusive engineers.

Notably, we did not observe significant shifts in either Theme 1: Beliefs about homelessness being a real issue, or Theme 4: Government responsibility towards addressing homelessness. However, these themes were not explicitly addressed through course content, so the lack of shift is not surprising. Nonetheless, it is important to acknowledge that some of the observed shifts in student responses may not be solely attributable to the intervention itself. For instance, students may have had prior exposure to similar content through coursework, lived experiences, or extracurricular experiences, which could have influenced their perspectives independently of the study. Additionally, variations in student motivation or engagement levels—either at the time of the intervention or during the assessment—may have affected how thoughtfully or critically they responded to the instrument. These factors could contribute to both positive and negative shifts in responses, and highlight the importance of accounting for contextual variables when interpreting changes in attitudes or perceptions. Future studies should consider collecting additional background data, increasing the reliability of the instrument, or implementing control conditions to better isolate the effects of the intervention.

5. Conclusions

This paper describes a study conducted in an engineering design course that implemented a project-based learning approach to help students meet course objectives around developing empathy and learning about social justice topics in a way that demonstrates that engineering is a sociotechnical endeavor. To integrate social issues with engineering, the instructors chose to use the context of homelessness to achieve said objectives. The intervention involved a collaboration with a local non-profit organization and a final course project that asked students to identify a need related to homelessness and design a solution to alleviate hardships. This study investigated, “How well and in what ways did the project-based learning curriculum contribute to a shift in students’ perceptions about homelessness?” The study explored the applicability of an established instrument—originally developed in the medical sciences—for assessing changes in student perspectives regarding homelessness within an engineering education context. The results revealed significant shifts in student responses on select items, indicating that the intervention may have positively influenced certain dimensions of student thinking—primarily shifting biases away from believing individuals experiencing homelessness suffer due to their own personal choices or moral deficiencies and towards more sympathetic and compassionate beliefs following the course intervention. However, the internal consistency of the instrument in this context was relatively low (Cronbach’s alpha = 0.553), suggesting that while some items may resonate with engineering students, others may not fully align with their disciplinary context or experience. This finding underscores the need for future research where cultural or disciplinary norms differ.

The positive results of the study suggest that students began to see themselves as agents of positive change in addressing homelessness. The shift in student perspectives regarding homelessness highlights the importance of integrating sociotechnical considerations into engineering curricula through project-based learning, rather than leaving social issues (and the impact engineering and technology has on these issues) to the humanities. The shifts towards more compassionate perspectives on the human experience aligned with the course’s objectives.

However, the study is not without limitations. First, student perceptions prior to the course may have aligned with our study’s end goals, based on students’ self-selection into our university which upholds values regarding compassion and social issues in the mission statement. Students may also have had previous courses that exposed them to similar content. Second, we did not observe significant shifts in student perceptions of engineers’ duty to broader sociotechnical issues outside of homelessness, pointing to an area for improvement for the course. While the project may center on homelessness, future iterations of the course should highlight homelessness as just one example of a social issue that engineering solutions intersect with. Moreover, there are several issues that future studies could consider addressing. First, adapting and revalidating the instrument for use in engineering contexts could enhance measurement accuracy. Second, implementing a mixed-methods approach—including interviews or open-ended surveys—may help uncover how students interpret key items and provide richer insight into the observed changes. Finally, future studies should consider larger, more diverse samples and the inclusion of control groups to better isolate intervention effects.

Overall, while preliminary, the findings suggest promise in adapting reflective or attitudinal assessments across disciplines while emphasizing the impact of critical reflection among engineering students. We encourage engineering educators to continue our work in combatting a technocratic view of engineering that prioritizes efficiency, logic, and innovation at the expense of human well-being. By emphasizing the compassion and empathy needed both for good engineering designs and for addressing social issues, engineering educators can use project-based learning as one approach for highlighting the sociotechnical nature of engineering work. This study’s insights contribute to the broader conversation about the development of critical consciousness among future engineers, and the importance of measuring shifts in attitudes toward complex problems in society. We hope our work has impacts beyond just the students in the study and contributes to on-going scholarly efforts that seek to transform engineering education into a more inclusive, justice-oriented field and redefines its role in addressing systemic inequities.