Enhancing STEM Attitudes of Rural High School Students through Engineering Design-Based Learning

Locational Disparities

Rural areas in particular face significant marginalization in the realm of STEM education. This issue is not unique to the United States but is a global concern, with the gap between urban and rural areas widening (Harris & Hodges, 2018; Zarifa et al, 2019). Gonzalez and Kuenzi (2012) reported that student achievement in STEM fields in rural and urban areas consistently lags behind that of suburban areas. In rural areas, students encounter limited opportunities for STEM education, which subsequently reduces their prospects of securing STEM-related employment (Lavalley, 2018). The rural-urban disparity in STEM education is not solely a matter of differences in socioeconomic status; it is also influenced by broader factors encompassing culture, student identity, motivation, teacher recruitment and training, and the overall context of STEM education (Biddle & Azano, 2016). Haron et al. (2019) revealed that a significant number of rural students perceived a disconnection between the STEM education they receive and its relevance to their daily lives. Woolfolk (2019) noted that educational psychology indicates that context is a crucial developmental element that envelops and interacts with an individual’s thoughts, feelings, and actions, influencing their learning experiences.

Further exacerbating locational disparities, the issues related to gender and location are intertwined with regions, academic disciplines, prevalent stereotypes, and broader sociocultural factors. Makarova et al. (2019) discovered that secondary male and female students tended to perceive school STEM subjects as having dominantly masculine attributes. Importantly, their study underscored that these gender stereotypes play a significant role in dissuading female students from pursuing STEM majors at the university level. In a separate study conducted by Cardador et al. (2021), it was observed that even when female students harbor an interest in STEM fields, they are more inclined to opt for non-STEM disciplines when presented with alternative choices. Furthermore, it’s essential to recognize that students’ interest in STEM is closely linked to their subsequent career choices and academic achievement (Maltese & Tai, 2011; Simpson et al., 2021; van Tuijl & van der Molen, 2016; Wang, et al., 2017). Previous studies indicated that interest is not solely a matter of vocabulary or definition; it is intricately tied to when and how interest develops. For example, Simpson et al. (2021) determined that the development of an individual’s identity plays a pivotal role in shaping their path toward a STEM career, and this identity formation is significantly influenced by the context in which the student resides.

Potential of Engineering Design-Based Learning

Engineering design-based learning has been recognized as an effective strategy for STEM education (Shanta & Wells, 2020). This pedagogical approach employs engineering design to tackle engineering-related challenges while fostering systematic methods, systemic thinking, critical analysis, and collaborative skills (Gómez Puente, Van Eijck, & Jochems, 2014; Lin, et al., 2021). Although it is akin to project-based learning, design-based learning distinctly prioritizes the design process and its resulting artifacts. This pedagogical approach exemplifies integrative STEM education (Wells, 2013, 2016a), as it promotes transdisciplinary learning by addressing real-world challenges and devising solutions through design, thus leveraging integrative STEM education. The pedagogy of engineering design-based learning is rooted in constructivist theory, which highlights the active role of learners in constructing their knowledge through learning experiences and social interactions. Teaching STEM through engineering design-based instruction utilizes scientific inquiry, engineering design, problem-solving, collaboration, and cooperative learning to encourage self-directed learning of students (Kelley & Knowles, 2016). Furthermore, engineering design-based instruction intentionally combines technological design with scientific inquiry to tackle issues pertinent to students’ context. This approach underscores the significance of the social interaction and cultural context within which learning occurs (Dym et al., 2005). Previous studies indicated that design-based learning narrows the gender gaps and perceptions surrounding STEM by incorporating real-world challenges into the classroom (Phothong et al., 2023). When implementing engineering design-based instruction, the learning process unfolds within a real-world context or a similar setting, facilitated by collaboration within a community of practice, transcending the traditional boundaries of school subjects (Breiner et al., 2012).

The engineering design-based learning approach, which employs culturally relevant contexts, provides the opportunity for students to engage with real-world problems they encounter in their own lives and presents a viable pedagogical approach for positively influencing student attitudes toward STEM education. As a measure of this viability, this study employed engineering design-based instruction as an intervention to examine its impact on student attitudes toward STEM.

STEM Attitudes

Recent studies indicate that integrative STEM education plays a significant role in shaping students’ inclinations toward STEM, enhancing their academic performance in STEM subjects, and positively influencing their career decisions (Wells, 2016b; Haron et al., 2019; Maltese & Tai, 2011). Hall et al. (2011) revealed that students’ interest and persistence in STEM have significant relationships with their choices of STEM careers. Dou et al. (2019) identified that STEM identity in the early stage is a pivotal factor in their career choices. Their study found that the STEM identity of young students plays a significant element while controlling other factors in choosing a STEM career in college. Tseng et al. (2013) confirmed that students’ attitude toward STEM influences their career development and acquisition of knowledge and skills in STEM fields. Furthermore, Popa and Ciascai (2017) conducted interviews with 110 college students. They discovered that the positive experiences in secondary STEM classes and with teachers were the key influences on their career choices in STEM. Sirakaya et al. (2020) also investigated the dynamic relationship between STEM attitudes and students’ thinking skills. They found that secondary students with positive STEM attitudes tended to demonstrate advanced 21 st-century skills, including creativity, critical thinking, and problem-solving when compared to their peers in the comparison group.

When assessing STEM attitudes, the underlying subcomponents are typically rooted in two pivotal psychological concepts: self-efficacy and expectancy-value beliefs (Eccles & Wigfield, 2002; Unfried et al., 2015; Han et al., 2021). Self-efficacy, as defined by Bandura (1993), refers to an individual’s confidence in their ability to organize and execute a particular course of action to solve a problem or achieve a task. Expectancies pertain to individuals’ expectations regarding their performance on various tasks or activities, while value relates to the motives or goals driving their actions. Unfried et al. (2015) developed the Student Attitudes Toward STEM (S-STEM) instruments, which encompass four key factors: attitudes toward mathematics, science, engineering/technology, and attitudes toward 21 st-century skills. These instruments propose that STEM attitudes are composed of two psychometric elements: self-efficacy and expectancy-value beliefs.

STEM attitude has been investigated from the integrated perspective as well as through the single-subject approach. In technology and design education, Pupils’ Attitudes Toward Technology (PATT) has been used to measure students’ interest and attitude towards technology. Raat and DeVries (1985) developed the PATT instrument to evaluate students’ perceived concepts related to technology, thereby measuring their attitudes toward STEM fields. The PATT instrument encompasses three key components: conception, career aspirations, and interest in technology (Bame et al., 1993). Over the course of more than three decades, the PATT instrument has been widely utilized in numerous international studies and has significantly contributed to identifying trends and patterns in students’ attitudes toward technology (Svenningsson et al., 2018).

In science education, various measurements and assessing approaches were introduced (Osborne et al., 2003). Wang and Berlin (2010) introduced a tool to assess scientific attitudes to determine an opinion or position on a psychological object in the field of science. Benek and Akcay (2018) presented an attitude scale for secondary school students toward STEM. They confirmed that the psychometrics of students’ STEM attitudes have six factors: science, mathematics, engineering, technology, STEM, and career. Furthermore, Kind et al. (2007) developed a survey tool to gauge pupils’ attitudes toward science. This tool delves into various dimensions, including students’ feelings, opinions, beliefs, and preferences concerning science. It covers a range of aspects, such as learning science in school, engaging in practical work in science, interactions with science beyond the classroom, the perceived importance of science, self-concept in the field of science, and future intentions for involvement in scientific pursuits.

Regional and Gender Gaps in STEM Education

Showalter et al. (2017) reported that students residing in rural areas exhibited lower performance on national-level reading and mathematics tests in grades 4 to 8, which are the critical period of career choices in relation to STEM (Maltese & Tai, 2011). Additionally, the report noted that lower academic achievements had significantly lower high school graduation rates, exemplified by Alaska’s rate of 60.9 % compared to Connecticut’s 94.3 %. Mau and Li (2018) conducted similar research in which they investigated factors influencing high school students’ career choices. Their study discovered that race, gender, and socioeconomic status wielded significant influence on STEM career aspirations, alongside factors such as interest in science and mathematics. These results indicate that students from underrepresented groups tend to opt for non-STEM majors compared to the other groups.

Previous research has unearthed several significant factors that contribute to the gender gap in STEM (Ciftci et al., 2020). Tyler-Wood et al. (2012) contended that the disparity in STEM academic achievements between genders is more closely associated with students’ perceptions of STEM than their inherent abilities. Their study demonstrated that positive learning experiences in STEM had a notable impact on increasing STEM interest, altering perceptions, and enhancing career intentions among female students. Shin et al. (2015) investigated the factors that influence gender differences, examining both factors within the home and motivational factors. Their findings emphasized the significance of gender similarities, rather than differences, in shaping career awareness in science. For instance, when parents hold positive perceptions of STEM, their female children are more likely to place higher value on and exhibit greater motivation in STEM.

As women are underrepresented in STEM fields, it is common for female students to possess a lower perception of STEM careers and identities (Cowgill et al., 2021; Swafford & Anderson, 2020). This can subsequently lead to decreased motivation and interest in STEM (van den Hurk et al., 2018; Talafian et al., 2019). Wegemer and Eccles (2019) proposed a model based on the Expectancy-Value Theory to elucidate the gender gap in STEM. Their model revealed that values, gender-related beliefs, and one’s sense of identity in relation to STEM are intertwined with behaviors related to achievement. These behaviors, in turn, contribute to the gender disparities observed in STEM. Their study provided that STEM career paths may be influenced by the alignment of values, identity, and gender-related beliefs regarding STEM fields. Additionally, O’Dea et al. (2018) investigated gender differences in school performance to pinpoint the origins of gender disparities in STEM careers. Interestingly, for the top 10% of students, their study found that an almost equal number of male and female students were enrolled in STEM subjects. Conversely, non-STEM subjects saw a higher proportion of female students. Specifically, they found that fewer female students with strong academic achievement tended to choose STEM careers, while middle to low-performing male students are inclined to pursue STEM fields.

In addition to students’ perceptions and career choices, researchers have uncovered the significant influence of teachers’ perceptions and school culture on students’ STEM career decisions. Hand et al. (2017) reported that high school teachers often hold gender stereotypes in STEM, associating masculine traits with the typical scientist and mathematician. Furthermore, Wahono and Chang (2019) found that male teachers tend to exhibit more positive attitudes toward STEM compared to their female counterparts, even though their STEM knowledge levels did not show significant differences.

Engineering Design-based Learning

Design-based learning is an instructional approach featuring design as a primary tool in project-based settings (Gómez Puente et al., 2014). Engineering design specifically, is a systematic, intelligence process for solving problems through generating, evaluating, and specifying concepts while satisfying a set of constraints (Dym et al., 2005). Engineering design-based learning includes characteristics of project-based learning, such as being student-centered, hands-on, collaborative, critical thinking, and creative, but additionally emphasizes an open-ended design process. In contrast, Fortus et al. (2004) proposed design-based science as an approach to teaching science using design activities for real-world problem-solving. They suggested that design-based science restructures school science curricula by featuring scientific inquiry as an integral component of design to address real-world challenges. While engineering design-based instruction shares similarities with other student-centered approaches like inquiry-based, problem-based, and project-based learning, it possesses unique characteristics (ITEEA, 2020). Specifically, engineering design-based activities addressing real-world problems result in the creation of artifacts that represents the mental concept, often mirroring the original idea or actual objects (Goldschmidt, 1991). Katehi et al. (2009) defined the nature of engineering design as highly iterative, open to the idea that a problem may have multiple possible solutions, serving as a meaningful context for comprehending scientific, mathematical, and technological concepts and stimulating systems thinking, modeling, and analysis.

Numerous research studies have reported that engineering design-based learning has yielded positive outcomes, including enhanced knowledge acquisition (Wells, 2021; English & King, 2015), increased creativity (Altan & Tan, 2021), improved problem-solving skills (Lin et al., 2015), heightened interest in STEM (Hall et al., 2011), greater motivation (Haron et al., 2019), and heightened career awareness (Maltese & Tai, 2011; Sadler et al., 2012). Similarly, Thibaut et al. (2018) claimed engaging in design activities within STEM education empowers students to apply scientific knowledge, aesthetic design principles, and mathematical skills collaboratively to arrive at solutions (Riskowski et al., 2009). Shanta and Wells (2020) also identified that students who participate in engineering design-based learning significantly improve their problem-solving and computational thinking skills. Additionally, the National Research Council (2009) reported that engineering design-based learning fosters engineering habits of mind among students, including systems thinking, creativity, optimism, collaboration, communication, and ethical considerations.

Context of Study

The Teachers and Researchers Advancing Integrated Lessons in STEM (TRAILS) project (NSF AWARD #1513248) was initiated to better understand and promote pedagogical practices that increase students’ motivations and capacities to pursue careers in STEM through innovative technology experiences for students residing in rural areas in the U.S. The project was conducted during the academic years 2016–2019. This initiative involved the active participation of high school biology and engineering and technology education (ETE) teachers. As part of the TRAILS project, the team developed an exemplary lesson known as Designing Bugs and Innovative Technology (DBAIT), which seamlessly integrated concepts from the fields of entomology and biomimicry into engineering design-based instruction. To prepare the STEM teachers for the implementation of these lessons, an extensive two-week summer professional development (PD) program was conducted, encompassing over 70 hours of training. During this PD, educators were not only introduced to the DBAIT lesson but were also provided with the opportunity to experience it from a student’s perspective. This pedagogical approach aimed to equip teachers with the skills and knowledge required to effectively deliver these lessons in their own classrooms.

During the summer PD, biology and ETE teachers worked together in pairs collaborating on crafting custom lessons that seamlessly integrated scientific principles and engineering practices. In the subsequent school year, these educators implemented two integrated STEM lessons – DBAIT and their custom-designed lesson – into their respective classrooms. Throughout this process, the TRAILS researchers and graduate students supported the teachers to develop a community of practice within the classrooms. This community of practice extended beyond the classroom, as teachers reached out to local experts to enhance their understanding of the engineering design activities, which, in this case, involved delving into topics such as aquatic entomology, fish food chains, and general ecology.

Participants

The total number of participants included in this study was 597 high school students taking biology or engineering and technology education (ETE) classes in the Midwest. The demographics displayed in Table 1 show that the proportions of males and females were 62.8% and 37.2%, respectively.

Data Collection

To measure students’ attitudes toward STEM, we used the science, mathematics, and engineering attitudes from the S-STEM Survey (Unfried et al., 2015). Additionally, we used the Importance Subscale from the DET section of the Teacher’s Perceptions of Engineering and their Familiarity with Teaching Design, Engineering, and Technology (DET) survey developed by Yaşar et al. (2006). These survey data were collected after the participating teachers implemented an engineering design-based STEM unit set in a rural context. The researchers obtained approval from the Institutional Review Board (IRB) before collecting the data.

S-STEM Survey

The S-STEM survey was used to measure secondary students’ STEM attitudes and career interests (Unfried et al., 2015). The survey consists of six subscales. The present study used the first three subscales, Math Attitudes, Science Attitudes, and Engineering and Technology Attitudes, to measure student STEM attitudes (Table 2). The instrument developers confirmed the reliability and validity of the survey instrument. For the middle and high school S-STEM survey, the Cronbach’s Alpha reliability level of each Math Attitudes, Science Attitudes, and Engineering Technology Attitudes, was 0.90, 0.89, and 0.89, respectively (Unfried et al., 2015). Construct validity was confirmed through exploratory factor analysis (EFA) that the four factors of mathematics, science, engineering, and 21 st-century skills explained approximately 70% of students’ STEM attitudes. Content validity was also confirmed by subject matter experts (SMEs): “three SMEs rated each item as essential, useful but not essential, or not necessary for measuring the given construct,” and item-level content validity ratios were calculated from the responses (Unfried et al., 2015, p. 625). Table 2 displays the survey items for STEM attitudes. A 5-point Likert scale (strongly disagree (1) to strongly agree (5)) was used for the STEM attitudes subscales. The items denoted (-) in Table 2 represent negatively worded item.

Design, Engineering, and Technology (DET) Survey

To investigate the relationships between teacher’s perceptions of STEM and students’ attitudes toward STEM, we used the Teachers’ Perceptions of Engineers and Familiarity with Teaching Design, Engineering, and Technology (DET) survey (Yaşar et al., 2006). The original survey consists of 41 items, but our study used the first subscale Importance of DET (Table 3), to measure teachers’ perception of the importance of STEM. The alpha coefficient for this subscale, Importance of DET, was 0.91.

Data Analysis

Given students were nested within each class, multilevel modeling (MLM) analysis was conducted using the HLM 8 software. For the descriptive statistics, the SPSS 26 software was used. Student STEM attitude was set as a dependent variable. Student gender (Female = 1, Male = 0) was a level 1 (student level) predictor variable. The level 2 (classroom level) predictors included teachers’ perception of the importance of design, engineering, and technology (DET) as measured by the DET survey (Yaşar et al., 2006), and class locale. The class locale type (Rural, Town, Suburb, and City) was dummy-coded, and Rural was set as the reference group.

The unconditional model (null model) did not include any predictors. The random coefficient model included student gender as a level 1 predictor (student level) to examine if the students’ STEM attitude following engineering design-based integrated STEM instruction, differs by gender (RQ 1). The final wholly conditional model (intercept-and slope-as-outcomes model) contained three independent variables including gender, DET, and class locale, to examine if the average scores of STEM attitude of the classes in a rural area differed from those of other classes (RQ 2) and if teachers’ perception of the importance of design, engineering, and technology (DET) relates to the class average scores of STEM attitude (RQ 3). The Level 1 predictor, gender, was uncentered, and the Level 2 predictors were group-centered. The Level 1 and Level 2 final, fully conditional models (intercept- and slopes-as-outcomes model), are as follows:

Level 1 Model:

Level 2 Model:

Mixed Model:

In the Mixed Model formula, STEM attitude_ij represents STEM attitude score for student i in class j, β_0j and β_1j represent the random coefficients of each group, γ₀₀ – γ₁₄ represents the fixed effects coefficients, u_0j represents the variation across classes, and r_ij represents the variation within class j (Suárez, & Wright, 2019).

Descriptive statistics

The analysis of Multilevel Modeling included 597 students nested within 35 classes. Tables 4 and 5 display descriptive statistics by gender and locale type where classes were taught. In the Level 1 Model, the average score of the dependent variable, STEM attitude, was 88.88 (SD = 16.33, skewness = 0.03, kurtosis = 0.293), and the averages of female and male were 83.12 and 92.30, respectively (see Table 4). For the Level 2 Model, we calculated the students’ STEM attitudes and DET at class level to identify the effect of locale that the class belongs to instead of individual student scores. The averages of STEM attitudes by locale had a moderate variation range from 86.90 in town to 89.90 in suburban. We collected the DET scores from the classroom teachers and the average of four locales was 75.57 (SD = 9.998, skewness = 0.041, kurtosis = 0.528)

Gender Differences in STEM Attitudes

To answer Research Question 1 (Do rural students exhibit distinct STEM attitudes compared to students from other locales after engineering design-based learning within a rural context?) we conducted an MLM analysis using the Level 1 model. Table 6 shows the results of the MLM analysis. The intraclass correlation coefficient for the unconditional is 0.221, which indicates that about 22.1% of the total variation in the student STEM attitude scores is associated with the classroom difference. The analysis of the random coefficient model revealed that gender (Female) is a significant predictor of student STEM attitude (γ₁₀ = –5.037, t (34) = –3.141, p = 0.003), which indicated that a female student’s STEM attitude score was 5.037 points lower than that of a male student. The gender differences were consistent across classrooms (τ₁² = 15.857, χ² (28) = 36.975, p = 0.119), which means that the size of the gender difference is consistent as 5.037 points across classes. This result found that female students’ STEM attitudes are lower than male students across the locales.

Locational Difference in STEM Attitudes

We conducted a mixed model analysis to test Research Question 2 (Do female students exhibit district STEM attitudes compared to their male counterparts after engineering design-based learning within a rural context?), examining the locale effect on students’ STEM attitudes. The results shown in Table 6 indicate that there were no significant regional differences in the students’ STEM attitudes. Reviewing mean STEM attitude scores of students shows slight differences in Town (γ₀₂) (B = –0.117, t (30) = –0.013, p = 0.990), Suburb (γ₀₃) (B = 0.671, t (30) = 0.145, p = 0.886), and City areas (γ₀₄) (B = 3.183, t (30) = 0.968, p = 0.341), but our model did not find statistical evidence that the difference in STEM attitudes is related to the locale of students. This result shows no difference in the effect by region when the same engineering design-based learning is provided.

Gender (γ₁₀) was found to be a significant predictor of student STEM attitude (B = –5.299, t (30) = –2.861, p = 0.008), which indicates that the average STEM attitude score of female students in rural classrooms was 5.299 points lower than that of male students in Rural classrooms. However, when gender was assigned as a fixed effect, the intercept-and slope-as outcome model showed no significant interaction effects between gender and other predictors.

Relationship between Teachers’ Perceptions of STEM and STEM Attitudes

To examine Research Question 3 (What relationships exist between teacher’s perceptions of the importance of STEM and students’ attitudes toward STEM following engagement in an engineering design-based learning within a rural context?), we conducted a Mixed Model analysis. Table 6 indicates that the variations of DET and Locale did not statistically contribute to the variation of gender effect. The random effect in this model suggests that the conditional mean of student STEM attitude varies across schools (τ₀² = 26.003, χ² (24) = 60.660, p < 0.001). Additionally, the association between STEM attitude and gender varies across schools (τ₁² = 29.217, χ² (24) = 36.505, p < 0.049). Compared to the unconditional model, 57.6% of the variance is reduced by adding level 2 predictors, DET and Locale, to the fully conditional model (τ₀² = 26.003) (Raudenbush & Bryk, 2002). Therefore, the final fully conditional model, intercept-and slope-as-outcomes model, revealed that teacher perception of the importance of DET (γ₀₁) was positively associated with student STEM attitude (B = 0.329, t (30) = 2.514, p < 0.018).