“It’s Not You, It’s The Room”— Are The High-Tech, Active .

RESEARCHANDTEACHING“It’s Not You, It’s the Room”—Are the High-Tech, Active LearningClassrooms Worth It?By Sehoya Cotner, Jessica Loper, J. D. Walker, and D. Christopher BrooksSeveral institutions have redesignedtraditional learning spaces to betterrealize the potential of active,experiential learning. We comparestudent performance in traditionaland active learning classroomsin a large, introductory biologycourse using the same syllabus,course goals, exams, and instructor.Using ACT scores as predictive,we found that students in the activelearning classroom outperformedexpectations, whereas those in thetraditional classroom did not. Byreplicating initial work, our resultsprovide empirical confirmationthat new, technology-enhancedlearning environments positivelyand independently affect studentlearning. Our data suggest thatcreating space for active learningcan improve student performancein science courses. However, werecognize that such a commitmentof resources is impractical formany institutions, and we offerrecommendations for applying whatwe have learned to more traditionalspaces.82Journal of College Science TeachingAmong active learning strategies, team-based learning(or cooperative learning)has perhaps the longesthistory and the richest evidentiarybasis (Michael, 2006; Prince, 2004;Springer, Stanne, & Donovan, 1999).Yet architecturally, traditional classrooms with rows of students facinga single focal point—the instructoror a central screen or board—are notnecessarily conducive to peer interaction (Milne, 2006; Oblinger, 2006).In response to this perceived barrierto the implementation of active learning strategies, a few institutions havepioneered the reconfiguration of entire classrooms (e.g., North CarolinaState University’s SCALE-UP classrooms [Beichner et al., 2007] and theTEAL project at MIT [Dori, 2007]).These rooms are designed to encourage student interaction and facilitateactive or team-based collaborativelearning by including features such asround tables, movable chairs, studentlaptop connections for sharing workon overhead projectors, and tablesidewhiteboards.Some work has been done to assess the effectiveness of these roomsin contributing to meaningful studentinteractions and in increasing studentunderstanding of course material (Dori& Belcher, 2005; Gaffney, Richards,Kustusch, Ding, & Beichner, 2008).At both North Carolina State University and MIT, students in the modifiedclassrooms had lower failure ratesand increased levels of conceptualunderstanding compared with studentstaking the course in a traditional classroom using a lecture-based approach.However, the interpretation of theseearly results is constrained because ofmethodological limitations; specifically, the previous work lacks sufficientcontrols to make a case for the physicalspace as opposed to the instructor (orpedagogical approach) in contributingto student gains.After a pilot study of two activelearning classrooms (or ALCs), theUniversity of Minnesota constructedthe new Science Teaching and StudentServices (STSS) building with 10ALCs (Whiteside, Brooks, & Walker,2010). These classrooms, modeled inpart on North Carolina State University’s SCALE-UP classrooms, consistof a centralized teaching station withtechnological controls and from 3 to14 nine-person round tables, eachof which has several laptop connections, a dedicated large overhead LCDscreen, whiteboard space, severalmicrophones, and visual access tolarge projection screens (Figure 1).Constructing these rooms requireda significant up-front and long-termcommitment of scarce resources, expenditures that are especially onerousfor a public institution facing budgetconstraints and increased tuitioncosts in a strapped economy. Was theinvestment worth it?

Learning gains inactive learningclassroomsTo determine whetherthe investment has hadthe desired effects, researchers at the University of Minnesota haveengaged in a longitudinal investigation of theALCs’ impact on howinstructors teach andhow students learn.Early work on theALCs (Whiteside et al.,2010) focused on studentand faculty attitudes,expectations, and perceptions of the rooms.Faculty members hadhigh expectations whenthey began teaching inthe ALCs and stronglypositive attitudes towardthe spaces at the end ofthe term; many notedthat their role changedin the ALCs, shiftingto the role of a learningcoach or facilitator. Students also had stronglypositive attitudes towardthe ALCs, particularlyregarding the rooms’ facilitation of teamworkand collaboration withtheir classmates and theability of ALCs to appealto a variety of learningstyles. Both faculty andstudents noted the importance of the round-tabledesign in altering theclassroom dynamic inimportant ways. However, some differencesemerged: Freshmen andsophomores rated therooms more highly thanFIGURE 1Top: Layout of traditional classroom (STSS 220). Bottom: Layout of active learningclassroom (STSS 330). More details available at http://www.classroom.umn.edu/.Vol. 42, No. 6, 201383

researchandteachingdid upperclassmen, and metropolitanstudents perceived the rooms as moreuseful than did students from ruralbackgrounds.Preliminary studyFollowing initial work on perceptions and attitudes, researchersworked with a faculty member todesign a quasi-experimental studyin an effort to isolate the impactof the room itself on student learning. The instructor taught the samesection of an introductory biologycourse (Postsecondary Teachingand Learning 1131), but in two different rooms—an ALC and a traditional classroom. The two sectionswere taught at the same time of day,yet on different days of the week.Course materials, assignments, exams, and pedagogical approacheswere controlled across sections; theonly factor that varied systematically was the room itself. Researchers were unable to assign studentsrandomly to the treatment and experimental sections; however, students were unaware of the room differences during course registration,allowing for post hoc demographicequivalency to be established. Theonly anomaly that emerged betweenthe sections was that students in thetraditional classroom had, on average, significantly higher ACT scoresTABLE 1Students in the traditional classroom had significantly higher ACTscores (yet did not perform significantly better in the course).Traditional 1020.65Sex (female )1640.78(0.04)1020.04Year (senior 4;first year .04)1070.66(0.06)740.09ACT score26.36(0.31)13925.32(0.37)811.04*Grade (% of total)77.77(0.69)16176.69(0.84)1021.28Note: Cell entries for each classroom are means, standard errors (in parentheses),and the number of cases for two-group, mean-comparison tests. ALC activelearning classroom.*p .05.84Journal of College Science Teachingthan did students in the ALC (22.54vs. 20.52; p .05). Given the predictive nature of ACT scores (e.g.,ACT, 2007; Marsh, Vandehey, &Diekhoff, 2008; Stumpf & Stanley,2002), we expected students in thetraditional classroom to earn highergrades than their peers in the ALC.However, at the end of the term,there was no significant differencein class performance, on identicalmetrics, between the two sections(Brooks, 2011). This finding suggested that the ALCs positively affected student learning.Methods and data collectionIn spring 2011, we sought to replicate these initial results using asimilar quasi-experimental designwith a different course, different instructor, and groups of students thatwere both larger and more representative of our general student population than the students involved inthe initial study. Specifically, oneinstructor worked with two groupsof students enrolled in an introductory biology course for nonsciencemajors (Biology 1003). One groupmet in a traditional classroom, theother in an ALC. In addition to controlling for instructor, every attemptwas made to keep course materialand designed activities the sameacross the two sections. Laboratoryexercises were identical, as werequizzes, homework assignments,and all three major exams.Biology 1003 is a large introductory class (N 161 and 102 for thetraditional and active sections, respectively). Survey data were collected viasurveys administered in class on thelast day of the term. The Universityof Minnesota’s Institutional ReviewBoard approved the protocol, andwe obtained informed consent forall subjects. Demographic and grade

data were supplied by the UniversityOffice of Institutional Research andthe instructor, respectively.A project PI and a trained studentresearcher collected observationaldata on 50% of randomly selectedclass periods for both sections. Anobserver recorded the levels at whichstudents appeared to be “on task,” aswell as specific characteristic behaviors of the instructor (e.g., lecturing,consulting individuals or groups, andworking problems with the documentscanner) and the students (e.g., consulting in a group, asking questions,and working on a group activity).And finally, students were surveyedabout their experiences and perceptions in their respective classroomson the last day of class. The designwas intentionally quasi-experimentalin that we used principles of experimental design, but we were unable torandomly assign subjects to controland experimental groups. However,we worked with the same instructor, same syllabus, and same testitems, and the sections were offeredback-to-back in the late morning onthe same days (Tuesdays and Thursdays); only the space was allowedto vary systematically. Although therandom assignment of students intosections that would have afforded afully experimental design was notpossible, the enrollment process (e.g.,registering for a specific lab section),coupled with post hoc equivalencytests, essentially approximates randomization. We suspect dialoguebetween the sections was minimal tononexistent: Students enrolled froma variety of majors within a verylarge university, lab sections werespecific to each class section, andthe sections—although offered backto-back—were on different floors ofthe classroom building with only 15minutes separating them.FIGURE 2Expected versus actual grades (BIOL 1003). Students in the ALC earnedsignificantly higher final grades than their ACT scores predicted (****p .0001).All instruments used in this research have been tested for scale reliability and validity and are availableonline at http://z.umn.edu/lsr.ResultsLike our earlier study, students in thetraditional classroom had, on average, significantly higher ACT scoresand were thus expected to outperformstudents in the ALC. And, like ourearlier study, final scores—on identical metrics—were not significantlydifferent across sections (Table 1).Given what we know about the predictive capacity of the ACT scoresfor grades, this finding is surprising.Using a point estimation regressionmodel, we expected students in theALC to earn approximately 6 percentage points lower on their finalgrades than their peers in the traditional classroom; instead, students inthe ALC earned half of a letter grademore than expected (p .0001; Figure 2). However, just as we observedin the first experiment (Brooks,2011), the altered environment didnot undermine the ACT’s predictivepower. In both classrooms, the ACTscore served as a reliable predictorof performance, predicting 20% and23% of variation in student grades inthe traditional and ALC spaces, respectively (Table 2, Model 1). Evenwhen we control for a host of demographic variables, ACT compositescores continued to be the only significant predictor of student gradeswith little explanatory improvementover the initial model (Table 2, Models 2 and 3). Thus, the patterns of evidence in support of our initial findings—that ALCs have a significantand independent impact on studentperformance—are identical.In the ALC, the same instructor,teaching the same material, spentVol. 42, No. 6, 201385