WIXLIB

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5of the course that worked well in Biol 111 did not achievecomparable results in Biol 211, reinforcing the importance ofempirical evaluation, even when implementing “best practices” ina course.METHODSCellular and Organismal Biology (Biol 211), is the second of twointroductory courses for biology majors at our institution, but theonly introductory biology course taken by biochemistry and prenursing students. The lecture course is a 3-credit course, and isseparate from the 1-credit Biol 211L laboratory course.Concurrent registration in the lecture and lab is not required,although the vast majority of students enrolled in the lecture doconcurrently enroll in the lab. Total enrollments range betweenapproximately 225 and 310 students per semester, and eitherone or two sections may be offered in each semester. Thus,section sizes range from approximately 125 to 310 students.Course topics include the scientific method, atoms, bonds andmolecules, cell structure, enzyme activity, cellular respirationand photosynthesis, molecular genetics (DNA replication,transcription and translation) and some physiology. The courseis taught by different instructors, who have the flexibility tospend different amounts of time on individual topics and toadjust the grading scheme for their sections. However, allinstructors followed the general course models as describedbelow during this extended course reform process, and oneinstructor taught 13 of the 25 sections included in this analysis(baseline through three distinct course models).Control (Baseline)The control, or baseline, condition was in place from Fall 2003through Fall 2005 (we are only considering academic yearsemesters and are excluding summer sessions). During thistime, 9 sections were offered by 6 instructors. The 3-creditlecture course met for three 50-minute lectures per week. Thesewere largely traditional lectures, with some activities such asthink-pair-share or small group discussion. Beginning in the Fall2005 semester, clickers were introduced into some sections,adding an element of active learning that was intended tohttps://doi.org/10.20429/ijsotl.2014.0802055

Introductory biology course reformengage all students, through a small percent of the final gradebeing earned by scored clicker responses (see Preszler, Dawe,Shuster & Shuster, 2007).Lecture Cases (LC)In Spring 2006, in-class case studies were introduced into thelecture. Eight class meetings (one meeting approximately everytwo weeks) were devoted to working through a case study. Inorder to ensure that students were accountable for the casestudies, case study work product (e.g. an in-class assignment orworksheet) accounted for between 30 and 35% of students’grades in this model, and exams included specific questionsrelated to the case studies. The case studies were intended toreinforce the lecture content as well allow students to applylecture content to novel, interesting and relevant scenarios.Some of the case studies were adapted from published casestudies at the National Center for Case Study Teaching inScience (http://sciencecases.lib.buffalo.edu/cs/), and some werewritten by the instructor. Students were required to completesome form of preparation for the case studies. Typically this wasa reading (from the textbook or a website) accompanied byreading questions to be completed before the in-class casestudy. Some of the questions were more specific to the case,and were essential to work through the case (e.g. cancerstatistics or nutrition information from specific foods).The instructor acted as a facilitator during each case study,ensuring that students kept on task and on time. A graduatestudent teaching assistant also helped facilitate the case studysessions, by circulating through the lecture room with theinstructor and helping student groups that had questions. Boththe instructor and teaching assistant were careful not to providedirect answers during these sessions, but did provide scaffolds tohelp students break their questions down into more manageable(and answerable) questions.In addition to the case studies, clickers continued to be acomponent of the LC model and grading scheme, accounting for15% of the final grade.https://doi.org/10.20429/ijsotl.2014.0802056

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5Workshops (WRK)We decided to capitalize on the success of the case studyapproach by having students work in small-group, peerfacilitated workshops (see Preszler, 2009 for a completedescription of the our workshop and peer instructor model). Inthis model, students met in the large lecture for two 50-minutemeetings per week. Instead of a third lecture meeting, eachstudent registered for and attended a mandatory 65-minuteworkshop. Each workshop enrolled up to 24 students and wasfacilitated by a BioCat (undergraduate peer instructor), who alsoattended every lecture in the course. The workshop activitieswere very similar to the in-class case studies in design andintent. However, the workshop format and length allowed formore student interactions. Each student group had a largewhiteboard, allowing them to present their work to other groups.The instructors prepared all the workshop materials, and spentapproximately one hour per week training the BioCats on the upcoming workshop. The BioCats suggested modifications duringeach training session and provided feedback on the previousweek’s workshop. The BioCats also graded the students’workshop assignments.The workshops contributed to between approximately 20%and 30% of the course grade, depending on the semester. Theworkshop model also included interactive lectures with clickersand a variety of forms of in-class student talk (Tanner, 2009).The workshop model was implemented for four semesters.Lecture Cases with BioCats (LCBC)Due to disappointing outcomes with the WRK model (see resultsbelow), we decided to revise the course. It was clear that the LCmodel had been more successful than the WRK model in Biol211. We thus decided to build on the prior experience,enhancing it with the addition of BioCats.In the current LCBC model, there are two 75-minutelectures each week. The lectures are interactive, with clickerquestions (students are encouraged to discuss the questionswith their neighbors), think-pair-shares and student-generatedquestions. Approximately once every two weeks, one of thehttps://doi.org/10.20429/ijsotl.2014.0802057

Introductory biology course reformlecture sessions is devoted to an in-class case study, facilitatedby the Instructor and the BioCats (eight in a single large section,or four BioCats in each of two smaller sections). As in the WRKmodel, the BioCats attend every lecture, and grade student casestudy assignments. As in the LC model, students complete apreparatory assignment before each in-class case study. A seriesof clicker questions based on the prep assignment is used in aneffort to ensure student accountability for the prep assignment.One feature of the LCBC model that extends the impact ofBioCats is a Bio Help Desk. Each BioCat schedules three hours ofBio Help desk each week, resulting in approximately 24 hours ofBiol 211 Help Desk each week. BioCats at the Help Desk areavailable to help students with questions about the coursematerial. The BioCats help the students by breaking theirquestions down into smaller steps, asking students to draw aprocess on the whiteboard available at the Help Desk, and/orasking the students to explain their answers. In a recent(typical) semester, 21% of course students signed in at HelpDesk at least once, and 6% of students signed in more than twotimes during the semester.Table 1 summarizes each of the course models describedhere. We are reporting on 16 academic year semesters from Fall2003 to Spring 2011.Table 1: Summary of the Different Course ModelsCourse ModelSemesters/Instructors* DescriptionInteractiveFall 2003-Fall 2005Traditionallectures with(5 semesters/9(CNTRL)sections, 6 instructors; clickers(1,180 students)A (2 sec), B, C (3 sec),D, E & F)Lecture Cases (LC) Spring and Fall 2006Interactive lecturewith clickers; In(432 students)(2 semesters/3sections, 2 instructors; class case studies( 8 per semester)C (2 sec.) & E)facilitated bysingle graduateteaching assistantand 2058

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5Workshops (WRK)(905 students)Spring 2007- Fall 2008(4 semesters/4sections; 2 instructors;C (3 sec) & E)Lecture Cases withBioCats (LCBC)(1,444 students)Spring 2009- Spring2011(5 semesters/9sections; 3 instructors;C (5 sec), G (3 sec) &H)Two interactive 50min lectures withclickers, one 65minute workshopfacilitated byBioCatsTwo 75-minutelectures per week;interactive plusfull-period in-classstudies every 2weeks, led byInstructor &facilitated byBioCats. BioCatsfacilitate HelpDesk.*Individual instructors are designated with a letter (A-H), and listed in eachmodel. For instructors that taught more than one section in each model, thenumber of sections is indicated.Assessment OverviewIn order to determine whether the course models were meetingour goals of increasing student success, we examined overallcourse grades in each course model across all course instructors.We additionally examined grades based on gender and ethnicity,to see if any course model was differentially impacting specificgroups of students in the course. We recognize that trackingcourse grades can be confounded by grading schemes,particularly the proportion of points associated with exams. Asdiscussed below, all three revised models added pointsassociated with the main intervention (relative to control/nointervention). Among all the revised models, the percent ofpoints associated with each feature (e.g. in-class case studiesand workshops) ranged from a low of 20.5% (one WRKsemester) to a high of 35% (the first LC semester), typicallyhovering around 30%.We used specific questions from student evaluations of thecourse in order to determine student opinions of each coursehttps://doi.org/10.20429/ijsotl.2014.0802059

Introductory biology course reformmodel. As many factors influence student opinions andresponses on student evaluations, we are only reporting thestudent evaluations for one instructor (“C”) who taught a largenumber of the sections in each of the revised models (two of theLC sections, three of the WRK sections, and five of the LCBCsections).We also monitored student performance on multiple-choiceexam questions on a traditionally challenging topic (cellularrespiration), to see if overall improvements in student gradeswere paralleled by improvements in performance on a specificcourse topic. Again, to reduce sources of variability in this morefine-scale analysis, we are only reporting exam performancefrom the same single instructor.Course GradesRelationships between course grades and course model, genderand ethnicity were evaluated using two-way and three-waycontingency table analyses. In all cases, if the probabilityassociated with the Pearson χ2 was 0.01, we concluded that thevariable(s) in question had a significant impact on studentgrades.We used two-way contingency tables to look at the impactof course model on course grades- specifically, whether thedistribution of grades differed in the different course models. Wealso used two-way contingency tables to determine whethergrade distributions differed between females and males, andwhether grade distributions differed between URMs and nonURMs. URMs are students who self-identified as being AfricanAmerican, Latino or Native American, and non-URMs arestudents who have self-identified as Asian American orCaucasian. Students who chose not to identify a race orethnicity, or who selected “other” during the institutionalapplication process were not included in the ethnicity analysis.The three-way contingency table analyses were used todetermine the impact of the different course models on therelationship between gender and grades (did females and malesrespond similarly to the different course models?) and 0510

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5and grades (did URM and non-URM students respond similarly tothe different course models?).In addition to evaluating grade distributions, we alsoexamined percent changes in grades, relative to the control. Inthese cases, we first calculated the percent of students earningeach letter grade in each case (e.g. the percent A’s, B’s, C’s, D’s,F’s and W’s earned by URMs in each course model). We thensubtracted the % of each grade in the control from the % ofeach grade in a given model (e.g. the % of A’s earned by URMsin the control was subtracted from the % of A’s earned by URMsin the LC model). This difference was then expressed as a % ofthe value in the control. As a hypothetical example, if the % ofA’s earned in the control was 14%, and the % of A’s earned in aparticular model was 20%, the difference is 6%, which is a42.9% improvement in A’s relative to the 14% in the control.Scores on cellular respiration exam questionsIn order to determine if observed improvement in course gradeswas accompanied by an improvement in understanding of aspecific topic, rather than an artifact of changing course gradingschemes, midterm and final exam questions pertaining to thistopic were analyzed from certain sections taught by a singleinstructor (“C”). Each question was evaluated for the percentageof students that answered it correctly within a class section. Forsome of the older semesters, either no data was available, oronly partial data was available (e.g. questions from only oneversion of the midterm, representing only a subset of students).Table 2 shows the data available for this analysis. In addition toplotting the averages for each course model, we used an ANOVAto investigate whether there were significant differences (p value 0.05) in cellular respiration exam question scores in thedifferent course models.Table 2: Cellular Respiration ExamSemester Course# tudents335411

Introductory biology course udent EvaluationsAs one instructor taught a substantial number of the sections ineach revised format, we examined course evaluation data forthat instructor. Focusing on a single instructor who taught in allfour versions of the course allowed us to compare studentevaluations of the four course models without confounding thecomparisons with instructor effects. Specifically, we focused onhow students responded to three questions about the courseformat: whether the specific format (LC, WRK, LCBC) madethem more interested in the course content, whether the specificformat (LC, WRK, LCBC) helped them understand the content,and whether the specific format (LC, WRK) was a positiveaddition to the course. Students were also asked about theircurrent interest in biology. These questions were embedded onthe anonymous end-of-semester student evaluations of thecourse. Students responded on a 5-point Likert scale (stronglyagree, agree, neutral/no opinion, disagree and stronglydisagree). The percent of students selecting each response wascalculated for each section/semester. These percentages wereaveraged for each course model, to obtain an overall studentevaluation of each course model.This research was reviewed and approved by theinstitutional IRB (protocol # 354).RESULTSGrades and Course ModelThe distributions of grades differed significantly between the fourdifferent course models (Pearson chi-squared p 0.001). Thedistributions of the percentages of each letter grade are shownin Table 3. The percent change of each letter grade relative tohttps://doi.org/10.20429/ijsotl.2014.08020512

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5the control is shown in Figure 1. The significance of thedifferences appear to be driven by large increases in B’s relativeto the control, and a decrease in F’s and W’s relative to thecontrol (Figure 1). These trends are generally consistent for eachof the three revised course models.Table 3: Distribution of the Percentages of Letter GradesCourse 11.08in Each“W”11.956.948.298.10Figure 1: Changes in Grades with Each Course ModelGrades and GenderThere were no significant differences in the grade distributions ofmales and females, when pooled from Fall 2003 through Spring2011 (i.e. across all the course models) (Pearson Chi-squaredp 0.75; Table 4).https://doi.org/10.20429/ijsotl.2014.08020513

Introductory biology course reformTable 4: Overall Course Grades (Fall 2003- Spring 2011)Males and 9124.3524.0312.3413.88Females 14.8424.3723.2713.4915.02for“W”9.509.01While males and females did not have differentdistributions of grades when averaged across all four coursemodels, males and females did respond differently to changes incourse models (3 way contingency table, Pearson chi-squaredp 0.001). The percent changes (relative to control) are shown inFigure 2. In comparison to the control semester, female’spercent increase in A’s and B’s was highest in the LC model, andthe LC model resulted in the largest reduction of W’s (coursewithdrawals) for females. In contrast, males showed no increasein A’s with the LC model, large increases in A’s and B’s with theLCBC model and concurrent reductions in D, F, and W’s with theLCBC model. In general, females performed best with the LCmodel, while males’ performance was highest with the LCBCmodel.Figure 2: Changes in Grades for Females and Males in DifferentCourse Models a. Female grades b. Male gradesGrades and URM StatusWhen looking at the overall distribution of grades pooled fromFall 2003 through Spring 2011 (i.e. across all the 4

IJ-SoTL, Vol. 8 [2014], No. 2, Art. 5models), URMs and non-URMS performed significantly differentlyfrom one another (Pearson Chi-squared p 0.001) (Table 5). Inthi

Michele I. Shuster PhD New Mexico State University - Main Campus, mshuster@nmsu.edu Ralph Preszler New Mexico State University, Main Campus, rpreszle@ad.nmsu.edu Recommended Citation Shuster, Michele I. PhD and Preszler, Ralph (2014) "Introductory biology course reform: A tale of two courses,"International Journal