In the United States, the attrition rate of undergraduate science, technology, engineering, and mathematics (STEM) students is notoriously high; more than 50% of first-year STEM students either change their major or fail to earn a degree (Aulck et al., 2017; Chen, 2013), which is significantly higher than the attrition rate of non-STEM majors (Drew, 2011; King, 2015). Not only are STEM attrition rates among the highest, but STEM programs have a historically small number of students switching into STEM majors after college entry, leading to an overall low number of students earning STEM degrees (King, 2015). Compared with graduate outcomes of other industrialized nations, the United States lags in STEM education (Aulck et al., 2017), and there is predicted to be a shortage of 1 million STEM-trained professionals in the coming years (Gates & Mirkin, 2012). The U.S. government has launched several campaigns to address the projected decline, as reducing STEM attrition has become a national priority (Aulck et al., 2017; Chen, 2013; Drew, 2011).

There are few comprehensive, statistically backed studies to pinpoint the causes of STEM attrition in the United States. However, reports in the academic literature propose multiple potential contributors to this phenomenon. Although precollege preparation (Tyson, Lee, Borman, & Hanson, 2007) and achievement in mathematics (Astin & Astin, 1992) are important predictors of persistence in college-level STEM programs, several of the key factors linked to attrition are associated with colleges and universities rather than individual students. These include the traditional lecture-based teaching style and resulting social isolation, a lack of role models, hypercompetitive peer environments, and overall lack of student support during the first year (Aulck et al., 2017; Chen, 2013; Drew, 2011; Forest & Rayne, 2009; Metoyer, Miller, Mount, & Westmoreland, 2014; Rask, 2010).

At our small liberal arts college in New York City (NYC), we recently introduced a first-year experience (FYE) course, “The Urban Ecosystem,” as a requirement for all freshman life science majors. In this course, students explore the science of the urban landscape, as well as the particular environmental challenges faced by NYC, through weekly site visits and field experiments. Although focused primarily on environmental themes, the course is highly interdisciplinary, drawing from biology, chemistry, social sciences, and the arts. Unlike in most introductory STEM courses, the goals of the class extend beyond the content. The course aims to promote success and persistence of first-year science majors by fostering a welcoming academic culture and also to facilitate students’ adjustment into college and city life. To achieve these aims, we incorporate field trips, social media–based assignments, student collaboration on field-based research projects, and near-peer mentorship. These interventions address the specific factors associated with undergraduate STEM attrition by reducing social isolation, providing role models, and establishing an informal support network for first-year STEM students. Here we assess the effectiveness of each aspect of the course with survey data from two consecutive cohorts of first-year students.

Research has revealed measurable differences in learning styles between current and previous generations, with today’s students benefiting most from visual and experience-based learning (Forest & Rayne, 2009). There is ample evidence suggesting that lectures are the least effective teaching method, with one study finding that students retain an average of 5% of information conveyed through lecture, which rises to 20% through audiovisual presentations and 30% when demonstrations are incorporated (Duderstadt, Atkins, & Van Houweling, 2002). The same study found that students retain 75% of information learned through experience (Duderstadt et al., 2002). Our Urban Ecosystem course implements an immersive experiential learning approach with a curriculum based on visits and experimentation at various sites relevant to the local ecological landscape and its environmental issues. Such trips include visits to aqueducts and reservoirs to learn about the challenges of urban water supply, a local recycling facility to learn about urban waste management, a wastewater treatment plant to learn about water quality issues, a green roof to learn about green infrastructure, and repeated visits to Central Park to study biodiversity and measure water chemistry (Table 1). This interactive learning environment contextualizes science and technology in NYC while targeting the learning style of this generation of college students. It also addresses the low peer interaction reported among STEM students (Chen, 2013; Drew, 2011; Metoyer et al., 2014). Traveling as a class to all sites gives freshmen ample opportunity to interact and bond with faculty, their peer mentor, and each other, helping to establish an academic support system. The trips accelerate cohort formation while giving students real-world context for learning scientific concepts.

Rigorous assessment of student learning through field trips can be difficult to achieve. Prior to each Urban Ecosystem excursion, students are assigned background readings that connect the trip to course themes. Following each outing, students are required to submit photojournalistic assignments, termed “Mixed Media Microessays,” on Instagram (Figure 1). Instagram is the most commonly used social media platform among college students (Knight-McCord et al., 2016). Microessays, which substitute for traditional response papers, consist of photos/videos taken during our excursions with detailed captions about what students learned and how it relates to our readings and course content (specific prompts are provided for each excursion). Students may use their personal Instagram account or create a separate profile for the course. Those who do not wish to use Instagram are offered the alternative of writing a response paper in place of each microessay; only one student in 2 years has chosen this option.

The goals of the Instagram-based microessays are threefold: (a) to hold students accountable for content conveyed during class excursions and strengthen links with course readings, (b) to stimulate discussions among students about class activities and content, and (c) to encourage peer interaction through social media, thereby boosting cohort formation among our freshmen. The average college student spends eight hours a day online, and 83% of 18- to 29-year-olds interact through social networking sites (Knight-McCord et al., 2016). Current undergraduates use Instagram and similar online media as a primary mechanism for social development at the start of college (DeAndrea, Ellison, LaRose, Steinfield, & Fiore, 2012). The microessays are intended to motivate students to complete their assignments (knowing their work will be viewed not only by the instructor but by their classmates) using a familiar and entertaining platform.

Microessays are graded on (a) how effectively they address the prompt for the week (3 points); (b) connections with readings (3 points); (c) writing quality, including correct grammar and spelling, coherent sentence structure, and clear expression of ideas (2 points); and (d) originality of ideas and how they are conveyed through words and images (2 points), for a total of 10 points. The Instagram caption limit of 2,200 characters requires students to synthesize the key points of each topic concisely, a hallmark of scientific writing. We have found that the Instagram-based assignments confer the unanticipated benefits of providing our course with a unifying theme (emblematized by #urbanecogram, a unique hashtag invented for the class) and fostering substantive peer interactions.

Two multiweek projects are incorporated into our FYE curriculum to provide students with another interactive approach to learning scientific topics. These two field-based projects include a semester-long research study of water quality in Central Park and a creative climate change project entitled HighWaterLine (Table 1). Both projects follow the team-based learning (TBL) approach (Metoyer et al., 2014), requiring students to actively synthesize information and solve problems in collaboration with each other. Research has shown that implementing TBL strategies in undergraduate science courses increases student attendance, participation, and academic success, while fostering higher levels of thinking and increased interest in the course (Metoyer et al., 2014). In addition, TBL provides another opportunity for peer interaction and bonding.

For the Water Quality research project, students collect and analyze samples from several bodies of water in New York’s Central Park once a month throughout the semester. Working together, students measure a variety of environmental parameters, including pH; temperature; turbidity; dissolved oxygen; and levels of phosphate, nitrate, and iron. The goals of this project are for students to monitor water chemistry as a function of changing season (from late summer to early winter) and to link their measurements with the persistent problem of eutrophication and algae blooms in Central Park’s lakes and ponds. Over the course of the semester, students assemble a data set that forms the basis for a scientific research paper and slide presentation on the last day of class. This original research project constitutes a substantial fraction (30%) of the course grade.

For the HighWaterLine project, students work in teams to put the effects of climate change into local context by mapping out scenarios for sea-level rise in NYC. This project is modeled after the work of performance artist Eve Mosher (). In 2017, to commemorate the 5-year anniversary of Hurricane Sandy, each student team visited a different neighborhood in lower Manhattan that was inundated during the storm to symbolically trace the highwater lines. Students achieved this using flood zone data from the U.S. Army Corps of Engineers and creative tools of their choice. Some teams used sidewalk chalk in emulation of the original work of Eve Mosher, others ribbon or rope. One team represented the line through a performance in which dancers painted a vivid red line on poster paper using their hands and feet, an exceptional example of the sort of arts-based science communication that particularly resonates with the general public (Lesen, Rogan, & Blum, 2016). Part of the HighWaterLine assignment was to engage passersby in discussion of climate change and storm surge, and to remind the public of the consequences of cataclysmic weather events such as Hurricane Sandy. Students were asked to create an edited video documentary of their experience to be presented and discussed during a subsequent class period. Through this project, students are able to visualize and contextualize the ramifications of climate change in NYC.

The Water Quality and HighWaterLine projects both require that students present their work in front of the class and incorporate background reading and research into their presentations. Studies show that students retain 90% of information when they teach material to others, making presentations one of the most effective teaching strategies (Duderstadt et al., 2002). These TBL projects also provide an opportunity for sustained academic collaboration, further contributing to cohort formation during the first semester. The Water Quality project is accessible for first-year students and provides a means to explore open-ended research questions. Early research experiences (during the first 2 years of college) have been shown to be highly effective in promoting persistence of undergraduates in STEM majors (Graham, Frederick, Byars-Winston, Hunter, & Handelsman, 2013), particularly students from underrepresented groups and students with low achievement (Nagda, Gregerman, Jonides, von Hippel, & Lerner, 1998).

There is little quantitative research examining the relationship between attrition and the social climate of STEM programs. However, several studies have reported anecdotal evidence that suggests feelings of isolation and lack of student role models in STEM programs do contribute to attrition (Chen, 2013; Drew, 2011; Forest & Rayne, 2009; Metoyer et al., 2014). To address these concerns, our FYE course makes use of a peer leader—an upper class science major who simultaneously serves as a guide, counselor, chaperone, and teaching assistant for the freshmen in the course. Peer leaders from all disciplines participate in a collegewide training program to learn how to help freshmen become acclimated to college and life in NYC. For STEM students, the peer leader also facilitates the difficult transition into a college science program, promoting persistence by providing a positive role model. The peer leader uses handouts, emails, and discussion to convey information regarding academic life, including STEM-specific study tips, and extracurricular activities. The peer leader also accompanies the class on most field trips, serving a vital function as a chaperone for a large group. While in transit on the excursions, the peer leader makes an effort to personally connect with students. Students also have the opportunity to communicate and/or meet with the peer leader outside of the classroom to aid student integration into a college STEM program. This experience is mutually beneficial for freshmen and their peer leader, consistent with previous findings that near-peer mentorship in STEM programs promotes growth and development of both mentors and mentees (Tenenbaum, Anderson, Jett, & Yourick, 2014).

Outcomes were assessed through voluntary, anonymous surveys administered in spring 2018 to students who had taken the Urban Ecosystem course in fall 2016 or fall 2017 (N = 20). Survey data were collected several months after the conclusion of the course, from a population of students retained beyond the first semester. Quantitative student feedback, summarized in Figure 2a (and described in more detail in the Appendix, available at ), reveals generally high ratings of the learning effectiveness of various course components. On a scale from 1 (poor) to 4 (excellent), the average student rating of the learning effectiveness of the field trips is 3.7 (Figure 2a). A large majority (80%) of students prefer field trips over traditional classroom structure (5%), with 15% of students considering their learning effectiveness equivalent (Figure 2b). In response to open-ended queries about the field trips (Table 2), students express enthusiasm for the experiential and hands-on learning aspects of the course. They repeatedly mention how field trips enhance their understanding of course topics. Some students find the field trips fatiguing, but the overall response is quite positive. Although this field trip–based course by no means precludes students from encountering traditional lecture-based teaching in other college STEM courses, it serves to engender more positive attitudes toward science as a whole.

On a scale from 1 (poor) to 4 (excellent), the average student ratings of the learning effectiveness of the team-based projects are 3.5 and 3.6 for the HighWaterLine and Water Quality projects, respectively (Figure 2a). In their free responses, students indicate that they benefit from “real world” application of course concepts and appreciate the creative aspects of the projects (Table 2). Similarly, the average student rating of the learning effectiveness of social media assignments is 3.6 on the 4.0 scale (Figure 2a). Students report that they enjoy using a familiar app (Instagram) to complete course assignments (Table 2). Overall, students’ comments indicate that the nonlecture-based teaching methods (i.e., field trips and projects) and the use of social media are productive and enjoyable experiences for learning scientific information and facilitating the transition to college.

On a scale from 1 (poor) to 4 (excellent), the average student rating of the effectiveness of their peer leader is 3.9 (Figure 2a), indicating clear advantages of the support provided through peer mentorship. A majority (70%) of students benefit both socially and academically from the advice and resources provided by the peer leader, while 15% report academic benefits only and 10% social benefits only. Only 5% of students do not benefit in either aspect (Figure 3a). Students’ responses to open-ended questions demonstrate that the peer leader provides vital support for both the academic and social challenges faced by first-year college science students (Table 2). These results reinforce the crucial dual role of the peer leader in facilitating students’ adjustment to the academic demands of a STEM program as well as college life in general.

Students also provided feedback on the effectiveness of the course for the adjustment to urban life and navigation around NYC (Figure 3b). For this measure, we differentiated the responses of students from NYC (N = 3) and those from elsewhere (N = 17). The average rating of the effectiveness of the course in the transition to urban life by students from out of town is 3.3 on the 4.0 scale. The local students are understandably less appreciative of this aspect (Figure 3b). Despite the lower ratings for this aspect of the course, students are enthusiastic in their free responses about how the field trips help them adapt to city life and learn about the environmental problems unique to NYC (Table 2). Overall, students report that their first-year experience in this course facilitates their transition to college as STEM majors and to life in our urban environment.

The quantitative and qualitative data from students’ questionnaires support our prediction that the unique elements of our integrated course, including the field trip–based structure, team-based projects, social media use, increased peer interaction, and near-peer mentorship, help first-year students adjust to a college STEM program. As Table 2 shows, student comments are generally positive and sometimes reveal advantages of the course we had not explicitly incorporated into our surveys, including the opportunity for individual interactions with the professor, and that the course motivated several students to add an additional major or minor in Environmental Studies. Anecdotally, we have found that the Urban Ecosystem course particularly benefits students who aspire to major in STEM but arrive at college lacking the academic background to begin the standard freshman science sequence in biology and chemistry. These underprepared students take the FYE course with the other science majors in the fall alongside remedial math courses with the goal of starting college-level science courses in their freshman spring. We have noted that after taking the Urban Ecosystem course, such students, who previously tended to drop out of the major after the first semester, appear not only to be retained in the program but also to form a strong part of the freshman life science cohort.

The ultimate measure of success of any FYE program lies in retention data. Although it is too early to assess long-term persistence and graduation rates for the two cohorts that have initiated the program, retention of life science majors to the second year (third semester) has been greater than 70% for both cohorts, the highest our program has seen in years (the prior 3-year average retention to third semester was 56%).

Future studies will focus on larger and longer term data sets to solidify the link between our FYE course and STEM persistence. This course could easily be adapted to any geographical community, with a focus on local environmental and/or technological themes.