The Room Itself is Not Enough: Student Engagement in Active Learning Classrooms

College Teaching ISSN: 8756-7555 (Print) 1930-8299 (Online) Journal homepage: https://www.tandfonline.com/loi/vcol20 The Room Itself is Not Enough: Student Engagement in Active Learning Classrooms Kelsey J. Metzger & David Langley To cite this article: Kelsey J. Metzger & David Langley (2020) The Room Itself is Not Enough: Student Engagement in Active Learning Classrooms, College Teaching, 68:3, 150-160, DOI: 10.1080/87567555.2020.1768357 To link to this article: https://doi.org/10.1080/87567555.2020.1768357 Published online: 01 Jun 2020. Submit your article to this journal Article views: 31 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=vcol20 COLLEGE TEACHING 2020, VOL. 68, NO. 3, 150–160 https://doi.org/10.1080/87567555.2020.1768357 The Room Itself is Not Enough: Student Engagement in Active Learning Classrooms Kelsey J. Metzgera and David Langleyb a University of Minnesota Rochester; bUniversity of Minnesota Twin Cities ABSTRACT KEYWORDS The primary purpose of this paper is to describe the variety of active engagements that characterize student behaviors in active learning classrooms (ALCs) across an undergraduate degree program. The number of different engagement types observed during a single class meeting varied between two and eight across 23 different courses. Three forms of engagement accounted for nearly 75% of all observed time, regardless of the subgroups (e.g., consistent in STEM vs Non-STEM courses, lower division vs upper division courses). In addition, unique patterns of student engagement characterized the pedagogical “signature” of a given course. We conclude that intensive class observation focused on student engagement not only has value for comprehensive undergraduate program review but also serves as a lever that invites faculty reflection on pedagogical practice toward course improvement. Curriculum; educational environments; pedagogy; student behavior; teaching methods Introduction Student engagement has occupied the attention of researchers in higher education for decades. Chickering and Gamson (1987) well-known “seven principles for good practice in undergraduate education” contained classroom behaviors that hinged on student engagement. Pascarella and Terenzini’s (2005) exhaustive review of the literature concluded that the level of student involvement in academic work directly influenced knowledge acquisition and general cognitive development. Two groundbreaking surveys–the National Survey of Student Engagement (NSSE) and the Student Experience in the Research University (SERU)–reaffirmed student engagement as a key proxy for student learning. Based on the NSSE results, George Kuh and colleagues (Kuh 2008; Kuh, O’Donnell, and Schneider 2017) identified High Impact Practices (HIP’s) as key levers for student success. These practices – e.g., research with faculty, internships, and culminating senior experiences – induce high levels of student engagement and are known to have especially broad and significant positive impacts on student learning (Kilgo, Ezell Sheets, and Pascarella 2015). Both research and the classroom experience of teachers leave little doubt that student involvement in the subject matter is the “grand meta principle” for student learning (Cross 1993). Fink (2003) views active engagement as an CONTACT Kelsey J. Metzger Rochester, MN 55904, USA. ß 2020 Taylor & Francis Group, LLC kmetzger@umn.edu umbrella term that has three components: a) gathering information and ideas, b) experiences that include direct involvement (doing) or indirect involvement (observing, listening), and c) reflecting on what is being learned and how one learns. Fink’s definition implies that student engagement can be directed by intentional instructional strategies. In the classroom, “active learning” is a phrase used to indicate instructional strategies that introduce student activity into traditional lecture courses and promote student engagement (Prince 2004, Bonwell and Eisen 1991). Extensive literature has demonstrated the significant positive impact on student learning and retention that results from the implementation of active learning pedagogies in the undergraduate classroom (c.f. Freeman et al. 2014). Active learning strategies have enabled students to outperform their predicted performance extrapolated from previous performance measures (Cotner et al. 2013), and results in a decrease in achievement gaps between URM students and majority students (Eddy and Hogan 2014). Active learning strategies are associated with changes in student behaviors that correspond with increased academic success in the course (e.g., distributed studying), and attitudinal shifts in which students in active learning classes perceive their class as a community (Eddy and Hogan 2014). Moreover, active learning strategies University of Minnesota Rochester, Center for Learning Innovation, 111 South Broadway, COLLEGE TEACHING enable students to interact closely and in meaningful ways with faculty members in the classroom, increasing the perception of instructor accessibility and promoting relationship-building (Metzger 2015). The enactment of these components by instructors is not confined to a particular learning environment; indeed, as Kuh, O’Donnell, and Schneider (2017) have emphasized, HIPs are found both inside and outside the classroom. It seems reasonable to assume, however, that certain environments can be structured to leverage the advantages of student engagement. We turn to one of those environments below – the active learning classroom – and examine how their affordances have been exploited to support active engagement by students in a multidisciplinary undergraduate degree program. Active learning classrooms as sites for leveraging student engagement Active learning classrooms (ALCs) are learning spaces intentionally designed to promote student-centered, collaborative, and technology-enhanced instruction and learning (Beichner 2014). These spaces have a few typical features: … round or curved tables with moveable seating that allow students to face each other and thus support small-group work. The tables are often paired with their own whiteboards for brainstorming and diagramming. Many tables are linked to large LCD displays so students can project their computer screens to the group, and the instructor can choose a table’s work to share with the entire class. Wireless Internet plays an important role in retrieving resources and linking to content management systems, and depending upon the size of the room, table microphones can be critical so that every student’s voice can be broadcast across the room. (Baepler et al. 2016, 10) Numerous studies, with many focusing on students in STEM disciplines, have demonstrated significant, positive student outcomes in ALCs operationalized as meaningful student interactions, decreased failure rates, and increased student satisfaction (Gaffney et al. 2008; Whiteside, Brooks, and Walker 2010). In particular, ALCs can affect the ways in which faculty and students relate to one another; i.e., the social context of the classroom can undergo beneficial change compared to traditional learning spaces (Walker and Baepler 2017). Significant relative learning gains have also been demonstrated when using active learning strategies in low-technology SCALE-UP mock up classrooms (i.e., low-tech ALC design with moveable rectangular tables and portable whiteboards) (Soneral and Wyse 2017) and when active learning strategies are 151 implemented in traditional, non-ALC settings (Stoltzfus and Libarkin 2016; Soneral and Wyse 2017). The use of interactive learning strategies has also been found to decrease or eliminate the achievement gap between male and female students in undergraduate physics (Lorenzo, Crouch, and Mazur 2006). Although the current level of deployment of ALCs in higher education is considered “experimental,” ALCs are predicted to rise in pervasiveness to the level of “mainstream” deployment by 2022 and were determined to be the top strategic technology in 2017 by the EDUCAUSE Center for Analysis and Research (2017). Despite demonstrated significant, positive student outcomes in ALCs, it does not directly follow that increasing the number of ALC spaces in higher education will lead to the seamless adoption of instructional strategies that promote more or deeper student engagement. As Daniel Wilson (2015) has observed, “space is simply a context for human behavior.” Affordances such as round tables and whiteboards merely present opportunities for instructional design; learner-centered teaching and subsequent student engagement are intentional acts. The physical architecture of ALCs is not a causative mechanism that directly and invariantly results in improved student learning (Brooks 2012). It may be better to view the architecture and affordances of ALCs as a potential “nudge” (Thaler and Sunstein 2009) for decision-making, i.e., encouraging faculty action but ultimately leaving the choice to individual faculty. A recent analysis of observations from over 2000 classes taught by more than 500 STEM faculty across 25 institutions concluded that “ … simply providing infrastructure or small class size does not necessarily change instructional practices, as about half of the classes with flexible seating and about half of the small- and medium-size courses were classified as didactic.” (Stains et al. 2018, 1468). While architecture and affordances provide a nudge, the value of studentcentered classrooms such as ALCs hinges on the alignment between a teacher’s epistemology and the enactment of student-centered pedagogies (Lasry, Charles, and Whittaker 2014). As Stoltzfus and Libarkin (2016) conclude, “ … while SCALE-UP-type classrooms may facilitate implementation of active learning, it is the active learning and not the SCALE-UP infrastructure that enhances student performance” (1). Student engagement and the University of Minnesota Rochester The vision for the [study location] since its inception in 2006 is to provide a personalized, integrated, 152 K. J. METZGER AND D. LANGLEY technology-enhanced, and evidence-based curriculum in health sciences. The University of Minnesota Rochester (UMR) offers two undergraduate degrees: a Bachelor of Science in Health Sciences (BSHS), and a junior-admitting Bachelor of Science in Health Professions (BSHP), which is delivered in partnership with the Mayo Clinic School of Health Sciences. The design and implementation of the undergraduate health sciences curriculum was vested in applying innovative and emerging best practices in support of student learning and development. The academic home of the BSHS degree is a multidisciplinary unit, the Center for Learning Innovation (CLI), which includes faculty from Biochemistry, Biology, Chemistry, History, Literature, Mathematics, Philosophy, Physics, Psychology, Public Health, Sociology, Spanish, and Writing. At the time of writing, this unit comprises 12 tenured/tenure-track faculty and 21 lecturers/teaching specialists whose primary academic experiences, training, and research was in non-education disciplines (e.g., biology, sociology, literature, philosophy). All faculty participate in teaching and support a “high contact” faculty model to assist students both in and outside of the classroom. Tenured/tenure-track faculty are primarily responsible for continuous and longitudinal assessment of student learning, which is often collaborative and focuses on the realization of course, curriculum, and/or institutional level learning and developmental outcomes. In position recruitment materials, student-centered teaching is an explicitly stated ideal and expectation for all faculty in the unit. For many faculty, however, their UMR appointment is their first full-time teaching opportunity in higher education post graduate school/professional training. In the Spring of 2017, UMR engaged in an extensive academic program review with a specific emphasis on pedagogy and curriculum. One part of the review involved a systematic program of observation of student engagement in courses across the BSHS curriculum. An important consideration of these observations was that all courses in the BSHS degree program are taught in ALCs (Supplementary document 1: Example ALC layout at UMR). This physical and curricular context provided an extraordinary opportunity to examine a programwide curriculum built on the assumption of instructional strategies that promote active learning engagement to enable significant learning and development of students. Data collection Data acquisition on student performance and other attributes is not only common at UMR, but an assumption of enrollment for students. From its initiation as a campus, UMR has been committed to continuous investigation of student learning across all courses and activities in the BSHS degree program. While the current classroom observation project was aligned with this institutional culture and programwide approach to interrogate variables that contribute to student learning, the goals of the classroom observation project were not primarily research-driven. Instead, the initial and primary goals were to a) acknowledge and honor both creative and routine work that signified student engagement in undergraduate coursework, b) provide feedback to faculty on the type and range of student engagement to improve pedagogy, and c) gain an initial perspective on the varied and common forms of engagement across the full degree program. To fulfill this goal, our focus was to identify behaviors on a macro/categorical level (e.g., discussion, problem solving) rather than on specific classroom activities (e.g., the use of jigsaws, brainstorming on a whiteboard). A number of structured instruments have been developed to facilitate classroom observations of student engagement and to determine the extent to which faculty use student-centered approaches (e.g., Classroom Observation Protocol for Undergraduate STEM (COPUS) (Smith et al. 2013) or the Teaching Dimensions Observation Protocol (TDOP) (Hora, Oleson, and Ferrare 2012). Both offered a finer grain of analysis than needed for meeting the essential goals of the current project, which were formative and descriptive rather than quantitative and evaluative. A broad literature base was consulted to derive different forms of classroom engagement that might be present in university courses. For example, the protocol for COPUS lists student behaviors such as listening, problem solving, discussing, and others as codes for student engagement. Svinicki and McKeachie (2014) list reading as another form of active engagement. Fink (2003) includes activities such as observing and reflecting under the banner of active learning, but the identification of “doing” activities in authentic settings opens the door to many additional categories such as performing (e.g., music, dance, theater) creating/constructing artifacts (e.g., studio arts), and designing or planning (e.g., architecture, clothing design). A full list of the ten possible student engagement formats used in our observational protocol is shown in Table 1. Not all of these formats were expected to be present in observations at UMR, given its focus on the health sciences and the limited number of degree programs. Nevertheless, most forms are well-known to instructors COLLEGE TEACHING 153 Table 1. Categories of engagement formats included in observation protocol. Engagement Type Creating/Constructing Definition building an artifact or product related to professional practice Examples 1. 2. 3. Designing/Planning preparing a scheme or approach to address a problem 1. 2. 3. Discussing verbal or digital dialoguing with one or more individuals 1. 2. 3. Investigating/Problem Solving conducting a detailed inquiry or exploration through a systematic process 1. 2. 3. 4. Listening/Processing attentional focus on auditory information to gain knowledge or understanding 1. 2. 3. Observing attentional focus on visual information directed at an object or process 1. 2. 3. Performing/Demonstrating/ Presenting goal-oriented action aimed at achieving a task within an applied setting 1. 2. 3. 4. 5. Reading/Studying information acquisition by directing attention to text-based content 1. 2. Reflecting active, persistent, and careful consideration of any belief or form of knowledge in light of the grounds that support it (Dewey 1933) 1. 2. 3. Writing (diagramming, drawing, keyboarding) use of written or digital methods to analyze, transcribe, or transform information 1. 2. 3. 4. constructing a physical model of a virus in microbiology lab stitching together a new garment in an apparel design course completing a project in the studio arts (e.g., painting, sculpture, drawing, ceramics) designing a mini-experiment to test a theory in abnormal psychology using AutoCad to design a new structure in an architecture course designing a prosthetic limb (digitally) prior to its fabrication in a biomedical engineering course talking with peers in a think/pair/share exercise in a literature course small group discussion in a philosophy course on the meaning of a text practicing language skills with peers in a Spanish course combining two reactants to determine the reaction rate in a general chemistry lab locating information via a web search in a public health course using interviewing to conduct a qualitative study in a sociology class diagnosing the severity of a contusion in a clinical nursing course listening to a math instructor explain the sequence for solving an equation listening to a lecture in a history course listening to a procedure being outlined by the instructor in an immunology class watching a physics instructor solve an equation on the whiteboard watching a video on children’s play in a child development course observing an instructor demonstrate a new routine in a modern dance class physical movement in a dance course or other kinesiology activity class (e.g., tennis, basketball) instrumental performance in a music course (e.g., piano, violin, trumpet) acting in a theater class carrying out a public speaking assignment in a speech class practicing the delivery of anesthesia on a simulator in dentistry reading a text (book, article) to acquire information in a gender and sexuality course reading or studying procedures for carrying out a biology lab assignment considering your beliefs about an ethical dilemma prior to discussion in an ethics of medicine course thinking about how your MBTI score has been influencing group dynamics in your psychology lab section contemplating how your choice of college major has been affected by your family of origin drawing a figure or table on a whiteboard writing out the steps to solve a mathematical problem prose writing in completing a report or as an exercise in a writing studio course taking notes on a lecture in a communications course 154 K. J. METZGER AND D. LANGLEY and include easily observable behaviors, as well as behaviors that require more inference about their connection to engagement. Clear definitions were generated as well as broad examples that signified likely behavior in each domain of engagement (Table 1). All classroom observations were conducted by an educational program specialist employed by the University of Minnesota Center for Educational Innovation (CEI). Classroom observations were scheduled with individual instructors of courses. An early observation took place within the first 7–8 weeks of the semester and was followed by a second observation during the latter half of the semester. Two observations were completed for each class involved in the study with a singular exception in which only oneearly observation took place. In some instances, the early and late observation course sessions were taughtby the same faculty member. In other instances in which a course was taught by a team of instructors, a different faculty member taught (or acted as lead instructor for the session) during the early and late observation session. An observational protocol was designed to help systematize the data collection. The observer was seated in the classroom in a position to view the majority of students in the class. At 15 second intervals, the observer noted the primary engagement behavior (the central focus of student behavior in following the intent of the instructor) of half or more students at that time and any secondary behavior simultaneously occurring (the “means” to the “end” - the primary engagement). For example, if an instructor prompted students to spend time on a practice problem, “problem solving” is the primary engagement behavior; to accomplish that, students might discuss with one another, write on a board, observe others writing on a board, etc., which would be the secondary engagement behaviors. At the conclusion of a given 15 second interval, a “1” was entered into an Excel spreadsheet in the appropriate column that designated the primary form of engagement. This procedure continued until the end of a 50 minute or 75 minute class. When the class session concluded, the amount of time in a primary engagement behavior was summed, and the overall percent of course time spent in these formats was compiled into a pie chart. Finally, the observer engaged in debriefing dialog with the instructor(s) and occasionally elicited the instructor’s estimate of time spent in engagement formats prior to revealing the observational data and pie chart summary. Debrief dialogues with instructors varied in length and content, but were never used to assess the quality or efficacy of the teaching and learning activities in the class session, or to question the decision of the instructor to design the session as they had. Rather, debrief dialogues were used only to prompt reflection from the instructor, e.g., “is this kind of engagement what you want?”. Notes on the responses of instructors were taken and used to help construct the qualitative portion of the current paper. In the current project, we focus on the primary behaviors for quantitative analysis, which secondary behaviors providing additional context for debriefing discussions with faculty. We stress again that the formative and descriptive goals of the project remained foremost. Nevertheless, we adopted particular systematic practices for this project, consistent with the intent of research instruments. Clear definitions and examples of engagement, a systematic recording interval, and the ability to summarize the observations in a meaningful way across all courses remained essential features of the observation endeavor. Results In total, 45 observations across 23 different courses were conducted, including 28 observations of 14 lower division courses (1000–2000 course number designators) and 17 observations of nine upper division courses (3000–4000 course number designators). Examined from the standpoint of distinguishing STEM or non-STEM courses, the observations can be described as 21 observations in 11 STEM courses and 24 observations in 12 non-STEM courses. An observation was conducted for each course in the first half of the semester, with a second observation of the same course in the second half of the semester. Of the ten possible engagement behaviors included in the observation protocol, eight were observed, while two – Creating/Constructing and Performing/Presenting – were never observed. The number of different engagements observed during any one class meeting ranged from two to eight, with an average number of engagements across all courses of 4.9. This finding was generally consistent across upper division and lower division courses (average 4.6 in lower division, 5.3 in upper division courses) and across STEM and nonSTEM courses (4.6 in STEM courses, 4.3 in nonSTEM courses). The engagement profiles resulting from the early and late session observations were occasionally consistent (Figure 1(a)). However, variability in student engagement profiles was generally COLLEGE TEACHING 155 Figure 1. Engagement behavior from selected courses. a) Pie charts comparing very similar profiles from same instructor from early and late observation period. b) Pie charts contrasting very different profiles from same instructor from early and late observation period. c) Pie charts demonstrating very different profiles for course Instructor A and course Instructor B from early and late observation period of the same course. 156 K. J. METZGER AND D. LANGLEY Figure 2. Summary of top engagement formats observed across all courses and observation sessions. Figure 2 displays the mean percent of time spent in the top five student engagements observed across lower division courses (n ¼ 14, 28 observations), upper division courses (n ¼ 9, 17 observations), STEM courses (n ¼ 11, observations ¼ 21), Non-STEM courses (n ¼ 12, 24 observations), and Overall (n ¼ 23, observations ¼ 45). observed between the early semester observation vs. late semester observation(Figure 1(b)), between early and late observations of two different instructors in the same class (Figure 1(c)), and certainly between different courses. Despite the variability in engagements observed, three engagement behaviors - Listening/Processing, Discussing, and Problem Solving - accounted for approximately 74% of all student time observed. Across all courses, the most frequently and extensively observed student engagement behavior was Listening/Processing, which comprised, on average, 36% of all student time, with a range during single class meetings from 0% to 79%. Listening/Processing represented seven of the top ten highest percentages of engagement across all courses, and was the highest percentage of engagement time in 24 of 45 class meetings observed. A summary of the top five student engagement formats observed across the courses and class meetings observed is displayed in Figure 2. Table 2. Most frequently observed student engagements. Listening Discussion Reading Writing/Drawing Observing Lower Division Upper Division STEM Non-STEM Overall 27/28 27/28 20/28 17/28 18/28 17/17 17/17 16/17 14/17 12/17 20/21 20/21 14/21 16/21 16/21 24/24 24/24 22/24 15/24 14/24 44/45 44/45 36/45 31/45 30/45 The number of class observations in which the engagement formats were observed out of the total number of classroom observations is displayed above. These general trends in amount and types of student behaviors were consistent across both STEM and nonSTEM courses, as well as across both upper division and lower division courses (Table 2). The finding that engagement profiles are similar across upper and lower division courses is consistent with the results of a recent large-scale investigation characterizing student engagement in undergraduate STEM that found no significant differencesin instructional styles usedacross different course levels (i.e., 1000 level courses vs. 4000 level courses), concluding that similar COLLEGE TEACHING engagements are used across the curriculum (Stains et al. 2018). Stains et al. describe three groups of classroom instruction/engagement profiles based on COPUS data using analyses of instructor and student behavior, categorized into three different profile types: didactic instruction, in which lecture comprises 80% or more of class time; interactive lecture, in which instructors supplement lecture with more student-centered activities such as group activities and clicker questions with group work; and student-centered instruction, in which “instructors incorporated student-centered strategies into large portions of their classes” (Stains et al. 2018, 1469). When these three profile types were examined for differences across upper and lower division courses, no significant different was found by Stains et al (v2 (8, N ¼ 1927) ¼ 11.0, p ¼ 0.20). Taken together, the findings of the current project coupled with the results of Stains et al. indicate that course level is not predictive of the type of student engagement observed. This conclusion is in contrast with the perception of many faculty, who predicted that the student engagement profiles of upper division courses would be markedly different from lower division courses: at the outset of our study, multiple faculty members predicted that lower division courses would be characterized by more frequent listening/processing, while upper division courses would be characterized by more student-centered engagements such as discussion and problem solving. Although Listening/Processing was the most consistently and frequently observed student engagement across the courses in our study, the majority of student engagement observed (>60% of the total time) was in other engagement types. Using these benchmarks of demarcation, the overall nature of the curriculum as a whole could be described as student-centered, with individual class meetings varying along the spectrum from didactic to highly student-centered. Our finding that Listening and Discussion are central and predominant behaviors in ALCs and other classroom settings employing student-centered approaches, although seemingly paradoxical, is consistent with observational data reported elsewhere (Cotner et al. 2013; Smith et al. 2013; Lane and Harris 2015; Stains et al. 2018). Given the prevalence and importance of communication between faculty and students, the classroom setting–a highly verbal environment–places demands on instructors and students to listen well and speak informatively. Classroom observation is a tool to present a mirror to faculty regarding student behavior and invite faculty reflection: “Are these forms of engagement helping students master the content and develop as individuals?” 157 Feedback and discussion between professional developer and faculty Providing feedback to faculty regarding the nature of student engagement in their class sessions was a key goal of the project. Despite our focus on descriptively characterizing student engagement across the curriculum, the responses from faculty regarding the observation data from their class meetings provides insight into how such data can be a powerful tool to elicit reflection on pedagogical choices and inform such decisions in the future. Such reflection can be elicited on an individual instructor/course basis or on a larger scale to encourage reflection about a course sequence or the totality of a degree program. In the context of our project, feedback sessions between the professional developer conducting the classroom observation and the individual faculty member or team of faculty members for the course illuminated consistent themes. In particular, faculty responses to the observation data could be examined relative to 1) intentions for eliciting student engagement for the session and 2) faculty perceptions of the student engagement actually elicited during the session. An additional but less prevalent theme was a misalignment between faculty’s desired student engagement and the type of engagement they felt they were able to plan to implement. When presented with the summary of student engagement data following a class meeting (e.g., breakdown of the percent of time spent in categories represented in a pie chart figure), instructors had varying initial responses. In some cases, faculty perceptions of how students were engaged in their class – and the relative time spent in those forms of engagement – closely matched the observational data. Typically, these faculty had highly structured plans for the class session and connected a clear pedagogical rationale with the engagement strategies chosen. Faculty cited a high value on the time spent face to face in class with students and a desire to make the most of the limited time together with intentionally planned engagement formats. In other cases, faculty perceptions of the engagement profile of the class session did not align with the observational data. A focus on the “what” – content – rather than the “how” – engagement – influenced instructor choices. In addition, faculty cited limitations for being able to facilitate particular student classroom engagement strategies, such as lack of student motivation to complete pre-class preparation (i.e., a ‘flipped’ classroom approach) that would facilitate the use of more student-centered engagement strategies in class. 158 K. J. METZGER AND D. LANGLEY At times the faculty member’s intentions for which types of student engagement behaviors would be elicited during a particular class session did not align with what occurred in reality. In discussion of why a disconnect between intentions and the observational data may occur, faculty described perceiving a particular classroom activity as highly engaging (e.g., oral presentations), but realizing in retrospect (prompted by the observation data) that only a fraction of the students were actually demonstrating that particular highly engaging behavior at any given moment while the majority of their classmates were listening. In these cases, having the benefit of viewing the observational data of actual engagement facilitated the faculty’s ability to revisit intentions and reconsider if such classroom practices supported their intended learning goals. Another reason cited for a mismatch between instructor intentions and the observational data was simply mismanaged time - faculty described an intention to limit the amount of lecture to prioritize time spent in group problem solving, but ended up spending more time lecturing than anticipated, which shortened the amount of time for other activities. In cases where observational data did not match faculty intentions in spite of having a highly structured plan of engagement, faculty attributed the discrepancy to the way(s) in which students responded during the session. For example, student misconceptions or questions that arise during a class meeting may require more time than anticipated during which the instructor is providing more didactic instruction to clarify the student question. Since a hallmark of student-centered instruction is responding to students in real time, adapting to the needs of the students informed by evidence gathered on the fly can be an attribute of effective, just-in-time instruction rather than a shortcoming of not adhering to an a priori plan. For circumstances in which faculty were involved in co-teaching, the observational data provided evidence that the teaching team could use to discuss their intentions for student engagement, the reality of what was being observed in particular sessions, and how they could use the information to more effectively engage in co-planning. This was particularly important for teaching teams in which the individuals comprising the team tended to very different engagement profiles. While a diversity of engagement strategies was universally seen as positive, faculty also expressed the desire to provide a consistent message to students about the student-centered nature of the course(es) that would be reflected in their student engagement strategies. Discussion This classroom observation data project, and others, are well situated within a larger framework of change in the paradigm of higher education from a place that delivers instruction to a place that produces learning (Barr and Tagg 1995). Contrary to widely held perceptions that university faculty engage in rigid patterns of largely instructor-centered teaching behaviors that do not change over time, recent work has demonstrated that changes in pedagogy are pervasive (Beyer, Taylor, and Gillmore 2013, 9). Moreover, the pedagogical changes that faculty make are most often the result of what faculty members observe from their students’ behaviors or performance in their courses, rather than from outside directives or mandates (Beyer, Taylor, and Gillmore 2013, 10). In other words, many faculty in higher education are in the process of shifting from teaching using instructor-centered approaches to teaching in ways that are responsive to their students (i.e., student-centered approaches) as a result of insight gleaned from what students do. Thus, classroom observation data with a focus on student engagement can be a useful, non-threatening lever to promote faculty reflection on course design and how teaching practices invite, promote, facilitate, and sustain student behavior. Using systematic observations of student classroom engagement can provide units or programs with evidence to inform and shape a teaching culture focused on student behaviors, and in which instructors can reflect thoughtfully on what is occurring and specific changes to pedagogy to support student learning and development goals of the unit. Given that the observations in this study took place exclusively in ALC contexts, we can also infer some general conclusions about the impact of classroom environment on teaching and learning practices. The variability of student engagement behaviors observed within and between courses across the BSHS curriculum demonstrates that ALC environments do not result in consistent behavioral outcomes, but rather than such spaces can invite and facilitate a variety of teaching and learning behaviors. From our results, it is apparent that different instructors can use the same physical spaces and affordances in very different ways; different instructional approaches chosen can in turn impact the effectiveness of the instruction in such spaces (Lasry, Charles, and Whittaker 2014). It has been convincingly and repeatedly demonstrated, however, that actively engaging students in well-designed learning experiences can lead to significant positive learning and development (c.f. Freeman et al. 2014). Thus, the task facing educators in higher COLLEGE TEACHING education is not whether or not to pursue studentcentered pedagogies – regardless of specific classroom environment affordances – but rather to determine which student-centered approaches are the most effective for a given context (e.g., for a particular disciplinary learning objective, distinct population of learners, etc.). Faculty teaching in ALCs, or striving to incorporate more student-centered pedagogies in traditional learning spaces, need iterative support in the form of initial training, practice, feedback, and reflection to effectively facilitate effective student-centered learning experiences. While ALCs invite many kinds of student-centered engagement simply through their physical design and technological amenities, the room itself is not enough (Baepler et al. 2016, 71) Institutions that wish to promote student-centered classroom engagement can communicate institutional support for such high impact practices by investing financially in ALCs or by retrofitting traditional classroom spaces as mock-ALCs (Soneral and Wyse 2017), and by incorporating an explicit focus on effective student-centered teaching in evaluation procedures. To best facilitate effective use of ALCs and student-centered approaches, faculty development and support for reformed teaching should be provided systemically. Institutions or units in the dreaming or planning stages for acquiring active learning classrooms would do well to consider in what ways they can support faculty to promote effective student engagement and learning gains. Disclosure statement No potential the authors. conflict of interest was reported by Funding This work was Minnesota Rochester. supported by University of ORCID Kelsey J. Metzger http://orcid.org/0000-0001-6927-7786 References Baepler, P., J. D. Walker, D. C. Brooks, K. Saichaie, 2016. and, and C. Petersen. A Guide to Teaching in the Active Learning Classroom: History, Research, and Practice. Sterling, VA: Stylus. Barr, R. B., and J. T. Tagg. 1995. “From Teaching to Learning: A New Paradigm for Undergraduate 159 Education.” Change: The Magazine of Higher Learning 27 (6):12–26. doi: 10.1080/00091383.1995.10544672. Beichner, R. 2014. “History and Evolution of Active Learning Spaces.” New Directions for Teaching and Learning 2014 (137):9–16. doi: 10.1002/tl.20081. Beyer, C., E. Taylor, and G. M. Gillmore. 2013. Inside the Undergraduate Teaching Experience: The University of Washington’s Growth in Faculty Teaching Study. Albany, NY: State University of Albany Press. Bonwell, C. C., and J. A. Eison. 1991. “Active Learning: Creating Excitement in the Classroom.” ASHE-ERIC Higher Education Report. Washington DC: School of Education and Human Development, George Washington University. Chickering, A. W., and Z. F. Gamson. 1987. “Seven Principles for Good Practice in Undergraduate Education.” AAHE Bulletin 39 (7):3–7. Cotner, S., J. Loper, J. D. Walker, and D. C. Brooks. 2013. “It’s Not You, It’s the Room”: Are the High-Tech, Active Learning Classrooms Worth It?” Journal of College Science Teaching 42 (6):82–8. doi: 10.2505/4/jcst13_042_ 06_82. Cross, K. P. 1993. “Reaction to ‘Enhancing the Productivity of Learning’ by D. B. Johnstone.” AAHE Bulletin 46 (4):7. Dewey, J. 1933. How We Think: A Restatement of Reflective Thinking to the Educative Process. Boston: D. C. Heath. (Original work published in 1910). Eddy, S. L., and K. A. Hogan. 2014. “Getting under the Hood: How and for Whom Does Increasing Course Structure Work?” CBE Life Sciences Education 13 (3): 453–68. doi: 10.1187/cbe.14-03-0050. EDUCAUSE Center for Analysis and Research. 2017. “Higher Education’s Top 10 Strategic Technologies for 2017”. https://library.educause.edu/resources/2017/1/ higher-educations-top-10-strategic-technologies-for-2017. Freeman, S., S. L. Eddy, M. McDonough, M. K. Smith, N. Okoroafor, H. Jordt, and M. P. Wenderoth. 2014. “Active Learning Increases Student Performance in Science, Engineering, and Mathematics.” Proceedings of the National Academy of Sciences of the United States of America 111 (23):8410–5. doi: 10.1073/pnas.1319030111. Gaffney, J. D. H., E. Richards, M. B. Kustusch, L. Ding, and R. Beichner. 2008. “Scaling up Educational Reform.” Journal of College Science Teaching 37:48–53. Hora, M., A. Oleson, and J. Ferrare. 2012. Teaching Dimensions Observation Protocol User’s Manual. Madison, WI: University of Wisconsin Madison; Wisconsin Center for Education Research. Kilgo, C. A., J. K. Ezell Sheets, and E. T. Pascarella. 2015. “The Link between High-Impact Practices and Student Learning: Some Longitudinal Evidence.” Higher Education 69 (4):509–25. doi: 10.1007/s10734-014-9788-z. Kuh, G., K. O’Donnell, and C. G. Schneider. 2017. “HIPs at Ten.” Change: The Magazine of Higher Learning 49 (5): 8–16. doi: 10.1080/00091383.2017.1366805. Kuh, G. D. 2008. “High-Impact Educational Practices: What They Are, Who Has Access to Them, and Why They Matter.” Change: The Magazine of Higher Learning. https:// provost.tufts.edu/celt/files/High-Impact-Ed-Practices1.pdf. Lane, E. S., and S. E. Harris. 2015. “A New Tool for Measuring Student Behavioral Engagement in Large 160 K. J. METZGER AND D. LANGLEY University Classes.” Journal of College Science Teaching 44 (6):83–91. doi: 10.2505/4/jcst15_044_06_83. Lasry, N., E. Charles, and C. Whittaker. 2014. “When Teacher-Centered Instructors Are Assigned to StudentCentered Classrooms.” Physics Education 10:010116. Lorenzo, M., C. H. Crouch, and E. Mazur. 2006. “Reducing the Gender Gap in the Physics Classroom (Author Abstract).” American Journal of Physics 74 (2):118–22. doi: 10.1119/1.2162549. Metzger, Kelsey J. 2015. “Collaborative Teaching Practices in Undergraduate Active Learning Classrooms: A Report of Faculty Team Teaching Models and Student Reflections from Two Biology Courses.” Bioscene 41 (1): 3–9. Pascarella, E., and P. Terenzini. 2005. How College Affects Students. Volume 2, A Third Decade of Research. 1st ed. San Francisco, Calif.: Jossey-Bass. Smith, M. K., F. H. Jones, S. L. Gilbert, and C. E. Wieman. 2013. “The Classroom Observation Protocol for Undergraduate STEM (COPUS): A New Instrument to Characterize University STEM Classroom Practices.” CBE Life Sciences Education 12 (4):618–27. doi: 10.1187/cbe. 13-08-0154. Soneral, P., and S. Wyse. 2017. “A SCALE-UP Mock-Up: Comparison of Student Learning Gains in High- and Low-Tech Active-Learning Environments.” CBE—Life Sciences Education 16 (1):ar12. doi: 10.1187/cbe.16-070228. Stains, M., J. Harshman, M. K. Barker, S. V. Chasteen, R. Cole, S. E. DeChenne-Peters, M. K. Eagan, J. M. Esson, J. K. Knight, F. A. Laski, et al. 2018. “Anatomy of STEM Teaching in North American Universities.” Science (New York, NY) 359 (6383):1468–70. doi: 10.1126/science. aap8892. Stoltzfus, J. R., and J. Libarkin. 2016. “Does the Room Matter? Active Learning in Traditional and Enhanced Lecture Spaces.” CBE—Life Sciences Education 15 (4): ar68. doi: 10.1187/cbe.16-03-0126. Svinicki, M., and W. McKeachie. 2014. McKeachie’s Teaching Tips: Strategies, Research, and Theory for College and University Teachers. 14th ed. Belmont, CA: Wadsworth. Thaler, R., and C. Sunstein. 2009. Nudge: Improving Decisions about Health, Wealth, and Happiness. New York: Penguin Books. Walker, J. D., and P. Baepler. 2017. “Measuring Social Relations in New Classroom Spaces: Development and Validation of the Social Context and Learning Environments (SCALE) Survey.” Journal of Learning Spaces 6 (3):34–41. Whiteside, A., D. C. Brooks, and J. D. Walker. 2010. “Making the Case for Space: Three Years of Empirical Research on Learning Environments.” Educause Quarterly, 33. https://er.educause.edu/articles/2010/9/making-the-case-for-space-three-years-of-empirical-research-onlearning-environments. Wilson, D. 2015. “Learning Spaces Week at Harvard.” Harvard Initiative for Learning & Teaching. https://hilt. harvard.edu/teaching-learning-resources/learning-spacesweek-at-harvard/