Since Michael believed his “way of thinking” made him feel out of place in his engineering major (Interview 6), we appeal to three areas of study to explain that marginalization. First, research on disciplinary practice explores the ways of knowing, seeing, and interacting that characterize how individuals approach problems within domains. In reviewing that research we pay particular attention to relationships between disciplinary practice and personal epistemologies, or individuals' stances toward knowledge and knowing. Second, identity research asks how students define themselves as individuals and how they come to belong (or not belong) as members of a community. Lastly, retention research asks what factors keep students in the discipline, why students choose to leave it, and what we might do to keep them.
To motivate and contextualize our data analysis, we argue from the literature that sense-making is an epistemological aspect of disciplinary practice, identity and disciplinary ways of knowing are deeply connected (i.e., a student can identify as someone who approaches learning and problem solving in a certain way), and engineering retention research has focused on students' identities but has not foregrounded the epistemological aspects of those identities.
Sense-Making: An Epistemological Aspect of Disciplinary Practice
Research suggests that successful engineers (and some successful engineering majors) take a sense-making approach to learning and using disciplinary knowledge. Again, by sense-making, we mean that practicing engineers and scientists often try to seek coherence and meaning across multiple representations and sources of knowledge. Documented instances of practitioners engaged in such sense-making include architects debating acceptable seismic loading thresholds for a building (Hall, Stevens, & Torralba, 2002), engineers designing and validating computational models of physical systems (Ball, Onarheim, & Christensen, 2010; Bucciarelli, 1994; Gainsburg, 2006; Jackson, 2010; Rooksby, 2010; Tang, Aleti, Burge, & van Vliet, 2010), engineers coordinating cartographic and schematic representations of roadways (Hall & Stevens, 1995; Stevens & Hall, 1998), and scientists reconciling mathematics across gestural and pictographic space (Hall et al., 2002).
An example helps illustrate the depth of participants' sense-making and its connection to epistemology. In Bucciarelli's (1994) ethnography of a firm designing a photovoltaic system, Beth, an engineer, is building a computer model to help reveal the source of a mysterious performance fault in a solar array. She runs her model against sensor data from the field data; the results reveal a discrepancy. The sensors report the array is overefficient: it is outputting more energy than Beth's model predicts it should (p. 60). Beth ultimately discovers her model was right; a faulty sensor was overstating the array power. Crucially, Bucciarelli argues that Beth's thinking entangles mathematical relationships, graphical representations, and the real behavior of physical hardware as she works her way through the solar array malfunction:
To Beth the object photovoltaic module is as much the symbolic, mathematical relationships describing how the current produced by the module depends upon the sun's intensity, the module's temperature, and the battery voltage as it is the artifactual, physical panel in itself. The “I-V” (current versus voltage curve) as a whole can be read as a module's signature, an image Beth can sketch out from memory in a few seconds. (p. 63)
In brief, Beth sense-makes by coordinating and reconciling among multiple kinds of knowledge, including a mathematical model and sensor output. Furthermore, her sense-making reflects sophisticated epistemological stances toward the complex but integrated nature of knowledge, in this case knowledge encoded by a mathematical model and knowledge of physical hardware, and the need to monitor the comparative tentativeness of different bits of knowledge, in this case the sensor output versus her model (Buehl & Alexander, 2001; Elby & Hammer, 2001; Hammer & Elby, 2003; Hofer & Pintrich, 1997; Muis et al., 2006). Bucciarelli's (1994) detailed analysis of her practices gives insight into the associated disciplinary ways of knowing. Moreover, this contextual, practice-centered approach is methodologically different from those of other kinds of epistemology research. Specifically, the ways Bucciarelli, Hall and Stevens, Gainsburg, and others explore disciplinary ways of knowing is distinct from epistemology research that categorizes students along relatively static, context-independent epistemological dimensions (Carey & Smith, 1993; Marra, Palmer, & Litzinger, 2000; Palmer & Marra, 2004; Schommer et al., 1992; Smith, Maclin, Houghton, & Hennessey, 2000) or group students into clusters to describe their behavior (Bernold, Spurlin, & Anson, 2007; Lumsdaine & Lumsdaine, 1995).
We discussed Bucciarelli's example in detail because it epitomizes the nuanced multimodal nature of sense-making in engineering. Furthermore, it is not unique. Hall and Stevens (Hall & Stevens, 2000; Stevens & Hall, 1998) document two civil engineers engaged in intense sense-making about a proposed roadway design. As the engineers interact through talk, gesture, and inscriptions, they “make visible” features of the roadway that were otherwise obscured or flattened in existing schematic representations (Stevens & Hall, 1998, p. 133). Gainsburg (2006) finds similar phenomena when Tim, a structural engineer in her study, struggles to create an accurate model of weight loading on floor beams. At the heart of Tim's struggle is his attempt to reconcile how his physical intuitions about the world should guide how he computationally models the loads on the floor.
Ultimately, these in-the-wild findings mirror the conclusions of expert-novice studies more generally, which emphasize how experts possess deep, interconnected knowledge (Chi, Feltovich, & Glaser, 1981; Reif, 2007) and how successful students blend intuitive conceptual reasoning with symbolic mathematics (Sherin, 2001). Our point here, though, is that sense-making is an authentic epistemic practice with respect to “making and evaluating knowledge” in science and engineering (Sandoval & Reiser, 2004, p. 368). In the next section, we argue that particular strands of identity-based research theorize strong, educationally-consequential relationships among identity, knowledge, and practice.
Identity Interrelates with Domain Knowledge and Disciplinary Practice
How students identify as engineers matters for their academic and personal development (Jocuns, Stevens, Garrison, & Amos, 2008; Stevens et al., 2005, 2008), for retention (Foor et al., 2007; Walden & Foor, 2008), and for the kinds of engineers they may become (Downey & Lucena, 2003; Tonso, 2006a, 2006b). Within engineering education, research has focused specifically on the developing professional identities of undergraduate students (Eliot & Turns, 2011; Loui, 2005), students' sense of identification with the discipline (Foor et al., 2007; Murphy et al., 2007; Stevens et al., 2005, 2008; Tonso, 2006a, 2006b), and students' identification with the culture and values of engineering as a profession (Tonso, 2006a, 2006b; Walden & Foor, 2008). Since our analysis of Michael will connect his identity to disciplinary knowledge and practices, this section focuses on work that relates students' identities to the ways they interact with domain knowledge.
Stevens and colleagues (2005, 2008) presented a model in which “becoming an engineer” is a journey through a complex landscape that couples learners' identification with engineering, their developing disciplinary knowledge, and their trajectories of experiences in engineering education. Stevens and colleagues rely on an ethnographic conceptualization of disciplinary knowledge called “accountable disciplinary knowledge,” defined in terms of what counts as a student knowing something in engineering, when, and to whom. This ethnographic view of disciplinary knowledge offers perspective on how institutional structures present students “with different images of engineering knowledge across the many contexts they inhabit and over the four years of their engineering education careers” (Stevens et al., 2008, p. 357).
Structures that hold students accountable for knowing things can also work to inhibit or restrict students' access to domain practices. Nasir and Hand (2008) argue that identity formation is tied to access to practice and to social recognition. Studying how the same high school students identify while playing varsity basketball versus doing mathematics, the authors show that students' identity with respect to each practice is in part tied to what participation structures are available to them. For example, a basketball practice drill afforded players socially valued ways to express themselves through finesse and technique in a way that the rigid call-and-response assessment of their math classroom did not (Nasir & Hand, 2008). These differences in recognized practice in turn shape how students perceived their roles (as an empowered agent on the court versus constrained in the classroom) and how they were perceived by their peers.
Ultimately these contextual, practice-centered analyses reflect “a shift from a focus only upon knowledge, to one that attends to the inter-relationships of knowledge, practice and identity” (Boaler, 2002, p. 47). In other words, learners' perceptions of which practices constitute knowing and performing in a discipline can link to their identification or what we term disidentification with the discipline. In the next section, we argue that disidentification with certain epistemic practices (Sandoval & Reiser, 2004, p. 368) embodied by an academic program may make students feel marginalized; such marginalization puts them at risk of abandoning the discipline.
Retention Research Has Underemphasized Epistemology
Engineering retention research has sought to identify factors that predict whether students stay in the major (Felder, Felder, & Dietz, 1998; Felder, Forrest, Baker-Ward, Dietz, & Mohr, 1993; Moller-Wong & Eide, 1997), assess the impact of curricular changes (Felder, 1995; Froyd & Ohland, 2005; Olds & Miller, 2004), and understand what works to retain students and increase diversity in engineering majors (Murphy et al., 2007; Walden & Foor, 2008). Large-N studies include multi-institution ethnographic research on why students decide to switch or leave (Seymour & Hewitt, 1997), multiyear mixed methods studies on student persistence (Felder et al., 1993 and subsequent work; Marra et al., 2012; Ohland et al., 2008), and the creation of quantitative models to help understand whether and when students switch or leave (Min, Zhang, Long, Anderson, & Ohland, 2011).
Over the past decade, however, researchers have also analyzed small groups or even a single student in detail to reveal underexplored facets of student retention (Foor et al., 2007; Stevens et al., 2007, 2005, 2008; Tonso, 2006a, 2006b; Walden & Foor, 2008). Many of these studies point to the role of students' developing sense of self, i.e., their identities. Bryn, a student in Stevens et al. (2005), chose to leave engineering not because of academic difficulties or lack of internship opportunities but because the engineering work she had done “behind a desk” clashed with her identification as a “people person” (p. 6). Stevens et al. (2005) also argue Bryn felt “engineering education is not a site for personal development” (p. 7).
Students in other studies also express identity-linked alienation toward their engineering programs. Surveying 113 students who left engineering, Marra, Rodgers, Shen, and Bogue (2012) found that lack of belonging was the only significant predictor of whether those students switched to a nontechnical major (p. 16). More pointedly, Foor, Walden, and Trytten (2007) offer a case study of Inez, a student who in her own words wished she “belonged more” (p. 104) in engineering. Those researchers show the cultural and institutional obstacles Inez faced in her engineering major: her strenuous (and financially necessary) work hours that made it difficult to see faculty (p. 108); a physics professor who told her she should quit the major (p. 110); and the fact that she had not yet participated in co-op or internship programs, which cost her cultural capital in her classes (p. 110). Her experiences show how students from nondominant backgrounds with respect to socioeconomic class, gender, ethnicity, or other social categories can be disenfranchised within the education system. This disenfranchisement stems in part from assumptions and structures, including the myth that we need not change pedagogy to accommodate different ways of knowing, that are embedded within the education system and reflect the experiences of those from dominant backgrounds (Foor et al., 2007, p. 112).
So far in this section, we have shown how mismatches between a student's identity and various aspects of an academic program can lead to marginalization and a student's desire to leave. We now concentrate on one particular aspect of academic programs: the epistemic practices they enact and value. In doing so, we build on our earlier argument that a student's identity can be coupled to their stance toward knowledge and knowing in a discipline. This epistemic aspect of students' identities and of academic programs, and the possible mismatches between them, has received little attention in the retention literature. Yet we see glimpses of this pathway to marginalization in some of the studies mentioned above and in studies of mathematics students.
Boaler's prominent line of mathematics education research suggests that epistemological aspects of the classroom culture can lead to learners' forming a sense of self that is alienated from the discipline (Boaler & Greeno, 2000; Boaler, 1998, 2000). Studying 76 students across six British schools, Boaler (2000) argues:
for many [students], mathematics was of another world and to fully engage in that world, students needed to suspend their knowledge of the real world, suppress their desire to interact with others, and strive to reproduce standard procedures that held little meaning for them. (p. 392)
Epistemological issues come to the fore when students report feeling that mathematics classrooms relegate their roles to those of calculating “robots,” where “being good at mathematics appeared to some students to involve being less than human” (p. 386).
In another study, Boaler and Greeno (2000) documented that high school students' reactions to their math classes – and their likelihood of taking math in the future – were strongly influenced by whether the classroom pedagogy aligned with their senses of themselves as knowers. Students who viewed themselves as creative thinkers and defined themselves in part by this characteristic tended to dislike a traditionally taught math class. They perceived the traditional math class as too authority-driven and inhibiting their agency over their own thinking, but they tended to like a reform-oriented math class in which students worked together to figure out how to solve problems. By contrast, students who identified as good rule-followers in some cases found the reform-oriented course disconcerting. Boaler and Greeno (2000) ultimately argue that how students develop their senses of identity and agency with respect to mathematics pedagogy shapes their decisions about studying mathematics in the future. In other words, students decided whether to persist in mathematics in part by judging whether their identities resonated with the classroom's “epistemic climate” (Bendixen & Rule, 2004; Feucht, 2010; Haerle & Bendixen, 2008).
Within engineering education research, Inez's story (Foor et al., 2007) highlights the need to explore epistemological aspects of identity and their role in student retention. Explaining why she found a physics course particularly challenging, Inez noted, “They say: ‘Here's the equation, plug and chug.' But it wasn't like that for me. I pretty much failed every test I took” (p. 109). By contrast, Inez enjoyed laboratory and project-based courses that allowed more opportunities for collaborative sense-making. Describing one such course, Inez said:
I am a more hands-on person than a test taker … and it really interested me. I like a lot of discussion. I don't mean when the teacher asks the class and someone answers. Like I like it when five kids know the answer and want to share the answer. And we are supposed to give back examples. I think it helps us interact better with the teacher. And then of course when we are given labs to do things … to know how to use the things we are learning to do in some classes. (p. 110)
While Foor et al. (2007) do not label this episode as being about Inez's personal epistemology or epistemological aspects of her identity, Inez is referring to exactly that. Her interest and success in a course were tied to the alignment, or lack thereof, between her personal approaches to knowing and learning and the epistemic practices valued within the course. By contrast, she perceived physics class as rewarding “plug and chug” (memorization and rote use of equations), and she gives hints of disidentifying with that epistemic practice: “It wasn't like that for me.” Such epistemological disidentification with the culture of a classroom or program does not just create “barriers to [students'] knowing” (Boaler, 2000, p. 387); it potentially augments or diminishes students' conviction to pursue STEM (Boaler & Greeno, 2000). Our case study of Michael contributes to this argument.