Increasing student engagement in large classes: a departmental case study.
S. Pollock and K. Perkins, University of Colorado, Boulder.
As a large state school, we are faced with the challenge of educating about 1800 students each semester in our introductory courses. The result is large lecture courses of up to 600 students held in theatre-style rooms. In 1997, several faculty members began promoting the use of interactive methods in the classroom. They cited the improved learning gains reported by others using Eric Mazur's Peer Instruction method3 and began using such techniques. Each year since then additional faculty members have adopted this style of lecture. In fall 2002, we wired our large lecture halls for the H-iTT student response system4. Each student now votes anonymously using a personal IR transmitter, locally known as a "clicker". We have developed a local database of "clicker questions" with roughly 10 categories of questions used in the department. These include questions that 1) quiz on the reading, 2) elicit/reveal misconceptions, 3) test conceptual understanding, 4) require prediction of experimental outcomes or simulation response, 5) recall a lecture point, 6) require reasoning to apply concepts in different contexts, 7) relate different representations, 8) do a calculation, 9) draw on intuition from everyday life, and 10) survey students. The mix and types of questions used varies with course and instructor.
The manner in which active engagement methods are used varies with instructor. However, all implementations have included an increased emphasis on collaborative learning by encouraging peer discussions related to clicker questions. In some courses, this simply means that students discuss their ideas with neighbors. In other courses, groups of 3-4 students are formed and are required to come to a consensus before voting. Some instructors follow up clicker questions with teacher-led, full-class discussions. In addition to clicker questions, some faculty use interactive lecture demonstrations5 which require each student to graph or otherwise predict the measured outcomes of a experiment.
This increased emphasis on collaborative learning exists outside the lecture hall as well. In 2000, the department created a public Physics Help Room for all introductory classes. Open 9 am to 5 pm and staffed by instructors and TAs, the "Help Room" is extremely popular. Typically thirty to ninty students are present in the room. Often they are working together in small groups. In addition, most of our introductory courses make use of CAPA, a computerized homework system6 which personalizes problems. This further encouraging a culture of student collaboration since student can work together on problems without feeling like someone is copying the work of another. In fall 2003, traditional recitations in the calculus-based mechanics course were replaced with tutorials run by TAs and undergraduate learning assistants. These tutorials replicate the University of Washington Tutorials7 as faithfully as possible, given local constraints.
Over the last 7 semesters, 28 out of 35 introductory physics courses included some degree of interactive engagement. Of the roughly 45 regular department faculty, 12 have now taught large lectures using these methods. Some faculty members are quite tentative, while others have designed their entire course around the interactive engagement methods. A smaller subset of faculty members have taken steps to assess the effectiveness of these methods. We have given the Force Concept Inventory (and other validated assessments) in a handful of classes. For comparison, traditional lectures generally result in normalized gains of about 25% on these tests8. In the mid-90's, one of our award winning lecturers measured a normalized gain of about this size in his popular but traditional course. Since shifting to peer-instruction style classes, we have measured gains of 33% in a more tentatively reformed class, 45% in an algebra-based course using clickers, 43% in a "pure concept test" style course, and 62% in our most recent calculus-based mechanics class using both the University of Washington style Tutorials and clickers.
Incorporating these new teaching methods has not had any adverse effects on course or instructor ratings. In fact, on average our interactive engagement-based courses rate higher than the traditional lecture-based courses on student evaluations. When asked to evaluate how the use of clickers in the classroom contributed to their learning, 96% of students in the most recent calculus-based course and 81% of students in the non-scientist course rated clickers as beneficial to their learning. Another value of using clickers has been higher attendance. When "clicker points" contribute to the grade, average attendance has exceeded 85%. When they are extra credit, average attendance has been above 75%.
It is interesting to reflect on the history which has led to a department where the climate and practices encourage the use of innovative, research-based teaching methods. Colorado University at Boulder has a legacy of dedicated and innovative teachers, including George Gamow, Frank Oppenheimer who developed freshman labs in a pre-Exploratorium style and Al Bartlett with his enormously popular classes that started in the 1960's. Nonetheless, our department was fairly conventional, with large teacher-centered lectures and graduate student instructors in recitation and laboratories. The shift towards an increasing awareness of physics education research and the use of active engagement methods started with the efforts of a modest number of energetic faculty. The growing interest and dedication of a highly respected senior research physicist to these efforts - evident through his local education initiatives and his vocal promotion of their value in improving education - drew the attention of the faculty. The Department Chair recognized and supported these fledgling efforts and contributed significantly himself. Among other things, he initiated a Preparing Future Faculty program (funded through AAPT) and modified the yearly teaching evaluation to include criteria that acknowledged the scholarship of teaching and learning.
The department now holds bi-weekly brown bag lunch meetings where interested faculty members can discuss education issues. Local physics education researchers to are often invited to speak. Typical attendance at the "brown bag" meetings is up to 30% of the faculty. This is an informal but powerful forum for sharing interest and ideas, spreading pedagogical theories and practical approaches, and encouraging reflection on individual and departmental practices. The department also began inviting high-profile colloquia speakers, Lillian McDermott, Eric Mazur and Lorrie Shepard among others. These (and other) colloquia had a noticeable impact and were well attended by the faculty. Providing forums for dissemination of ideas - hallway discussions, colloquia, and lunch meetings - has certainly contributed to the spread of interest in and awareness of new teaching methods.
In addition to these bottom-up efforts, financial support from the administration, the department, and other university programs including the graduate school has been invaluable. The Colorado University (CU) administration has a unique mandate to improve undergraduate education which provides both monetary support through student lab fees and top-down pressure on the department. The CU Faculty Teaching Excellence Program and Graduate Teacher Program are involved with our efforts, supporting faculty members individually and providing training and support for our graduate teaching assistants (TAs), including developing a lead TA program. An NSF Teacher Preparation grant9 allowed us to hire undergraduate learning assistants to team up with graduate students as learning coaches in the Tutorial sessions. Increased collaboration with the School of Education provided learning assistant training, joint meetings on science education, and collaborative research activities in our classes.
More recently, our efforts gained momentum, with a hire this year of a junior faculty member whose research is in the field of physics education research (PER) , the development of research programs in PER by two tenured physics faculty members, and the ongoing interest of a senior instructor. We now have a rapidly growing PER group with internal and external funding, graduate students, and a post-doc. For more information on our group, please visit our web page at www.colorado.edu/physics/EducationIssues. We welcome feedback on efforts at other institutions to implement sustainable and effective change.
(1) McDermott, L., Redish, E., American Journal of Physics, 67(9), 755-767, 1999
(2) www.colorado.edu/physics/phet, supported by Kavli Foundation, NSF, and CU.
(3) Mazur, E., Peer Instruction: A User's Manual. Prentice-Hall, NJ 1997
(4) H-ITT: See www.h-itt.com
(5) See e.g. Sokoloff, D., Thornton R., The Physics Teacher, vol. 35, pgs 340-346 (1997).
(6) CAPA: See www.lon-capa.org
(7) McDermott, L., Shaffer, P., and the PEG, Tutorials in Introductory Physics, Prentice Hall, NJ 2002
(8) Hake, R., Conservation Ecology 5(2):28, '02, www.consecol.org/vol5/iss2/art28
(9) Supported by a STEM-TP grant from the National Science Foundation