Beyond Facts and Formulas

Wednesday, August 1, at 1 p.m., U.S. Eastern time

Robert J. Beichner is a professor of physics and director of the STEM Education Initiative at North Carolina State University. He has developed a teaching method, called Scale-Up, influenced by educational psychology and cognitive science.

A transcript of the chat follows.

Jeffrey Brainard (Moderator):
    Hello and welcome to this week's Colloquy Live Discussion. I'm Jeffrey Brainard, a reporter for The Chronicle. I will moderate our discussion today with Robert J. Beichner, a professor of physics and director of the STEM Education Initiative at North Carolina State University. He has developed a teaching method, called Scale-Up, influenced by findings in educational psychology and cognitive science, about how students best learn. Thanks for taking the time to join us today. Please send in your questions now, and I will add a few of my own. Let's get started.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Bob, let me start things by asking you to describe the method of teaching introductory physics that you developed, called Scale-Up. What are some of the major ways in which it differs from a traditional introductory course?

Robert J. Beichner:
    The name really says it all. SCALE-UP stands for “Student-Centered Activities for Large Enrollment Undergraduate Programs.” (The name originally ended as “Undergraduate Physics” but since then many different institutions are teaching a variety of courses this way.) The basic idea is that you give students something interesting to investigate. While they work in teams the instructor is free to roam around the classroom, asking questions, sending one team to help another, or asking why someone else got a different answer. There is no separate lab class and most of the "lectures" are actually class-wide discussions. We carefully structure the groups and give them many opportunities to interact. Three teams (named a, b, and c) sit at a round table and have white boards nearby. Each team has a laptop in case they need web access. At NC State, our classes usually have 11 tables of nine students, but some schools have smaller classes and at least one I know of has bigger classes. Physics, chemistry, math, engineering, and even literature courses use this approach. I’d love to see a political science class taught this way. The teacher would pick some current event, say the Attorney General’s Congressional testimony. The “a group” at each table would see how CNN covered the event. The “b groups” would read the Washington Post coverage, while the “c groups” could find the Fox News website. Then they would compare and see what aspects were covered by all three and which things are missing is some. They might then be sent on a search to find the least biased presentation (perhaps by the BBC?). Whether the topic is current events or chemistry, the basic idea is the same. Students work while teachers coach. Note that this requires extensive preparation by the students. We make sure the chapter has been read and simple homework finished before students come to class to ask questions and work on interesting activities.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    The Scale-Up method has some pretty specific suggestions for classroom practice, such as that students sit in discussion groups at round tables seven feet in diameter – no more, no less. Tell us how you arrived at this precise number, and why it is important. Will the method suffer with eight-foot-wide tables?

Robert J. Beichner:
    The SCALE-UP classroom, which is still changing, is based on a decade of research. We knew we wanted to facilitate student group work, so we tried a variety of different table geometries before realizing that, of course, restaurant-style round tables worked best. Once we decided that, we tried four different diameter tables. We videotaped classes, made disciplined observations (i.e. looked at very specific things at regular intervals), and held focus groups with students. After students worked at one table for a few weeks, we moved them to a different size table and repeated the data collection. We kept this up until we were satisfied that we had the best table size--seven feet in diameter! Students feel cramped at smaller tables (when there are 9 students per table) and bigger tables make for noisier rooms. Once one student calls across a 10 foot table to a peer, everyone in the room must increase their volume.

Question from Cheryl, comprehensive land grant university:
    Can you address the resource issue? Small classes with lots of interaction and critical thinking exercises require well-trained faculty that are increasingly expensive. At comprehensive land-grants, teaching in the STEM disciplines is not rewarded on the same scale as research.

Robert J. Beichner:
    There is no doubt that teaching is not rewarded like research. The Carnegie Foundation’s idea of the Scholarship of Teaching and Learning addresses this in detail. (One big problem is assessing teaching. Charlie Glassick, former President of Carnegie, wrote a book on just how to do that.) More and more universities are modifying their tenure and promotion guidelines to raise the profile of teaching...but we still have a long ways to go. And your comment about small classes is well taken. That’s precisely why we developed the SCALE-UP approach, which is nothing more than using a carefully designed classroom and instructional materials so we can make a large class feel small and intimate. Teacher training does take time and money, but most things that are valuable do. I think as more people realize that a healthy US economy becomes increasingly dependent on an educated citizenry, the easier it will be to find the resources to make that happen. The recent “Rising Above the Gathering Storm” report illustrates the fact that American business has come to this realization. It is also a matter of priorities. I heard Alan Kay speak this morning and he noted that the money we spend every 3 months in Iraq would buy a laptop computer for every US high school student.

Question from Emil Chuck, George Mason University:
    One of the challenges that face instruction of introductory coursework is that many teaching faculty are adjunct or non-tenure-track, so instructors are not available to the dozens to hundreds of students who take our courses. What suggestions do you have for implementing many of these pedagogical innovations if neither the instructors nor the vested faculty have the time to nurture interactive and engaging relationships?

Robert J. Beichner:
    I’m reminded of the saying, “You can lead a horse to water, but you can’t make him drink.” That’s true, but if he’s thirsty, he’ll drink by himself! I think there will be external economic pressures to improve teaching, both from global challenges to our national, financial well-being as well as direct alternatives to traditional university coursework. Accredited schools now offer classes and degrees online. If a faculty member doesn’t want to make their courses attractive, then they might find themselves out of a job.

Question from Beth Simon, UBC Wieman Institute and UCSD:
    When I have spoken to those I know who are interested in the student classroom experience -- but still run their classes "lecture style" I often get the following feedback when I describe my active learning-oriented lectures: But there are just things students have to know in biology/economics/chemistry etc. My interpretation of the book's materials are my contribution to the classroom. Our materials for this class aren't set up to encourage teaching and evaluation of problem solving like yours are (in computer science). What do you say to that sort of challenge? Does it really require a complete redesign of how such subjects are taught from the curricular level?

Robert J. Beichner:
    I tell my students that if all I do is repeat what's in the book, they could either save time by skipping class to read, or same money by attending class but not going to the bookstore. True, students need help learning what's in the book, but we should be able to supplement what's there, not just repeat it. And there are some things that books can't do. For example, students can write code to solve particular problems, then compare what they've done with their peers. It does take a careful rethinking of what you want your students to get out of a course.

Question from Nicola Mellor, U. of Reading:
    How do you propose that we deal with the issue of individuals' academic identity? Lecturers have succeeded in the traditional system so it can be uncomfortable to admit that their own 'success' as undergraduates has been in memorization rather than skills acquisition. After all, if you compete and win, it is less appealing to question the rules.

Robert J. Beichner:
    It is probably true that nearly all higher education faculty members were very good students in traditional classroom settings. This leads to inertia--”It was good enough for me, so it’s good enough for them.” Besides, we don’t want “just anyone” to succeed in our difficult fields. We want only the best. So many introductory courses become filters instead of pumps. What ends up happening is that people who think as we do will succeed and the rest won’t. But we can no longer afford to only concern ourselves with 2 or 3% of our students. I’m not promoting lower standards. I’m saying we have to find ways to help more students learn (and then raise our standards) because we can no longer afford to be elitist while our nation falls behind in the growing global economy. It turns out physics has enjoyed an advantage over other STEM areas in promoting educational reform. For a few years now we have had some really good conceptual tests that clearly point out what students aren’t learning. The tests look so simple that it sometimes is hard to convince faculty members that they won’t insult their students’ intelligence by giving it. But if they can be persuaded, they are usually shocked at how little their students are gaining from the erudite lectures. Even students who excel on traditional tests very often reveal embarrassingly low conceptual understanding of fundamentals. Most instructors, when faced with these facts, want to do a better job of teaching.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    You’ve talked about how it can be challenging to persuade colleagues in STEM (science, technology, engineering, and mathematics) disciplines to try the new teaching approaches, but that newcomers react favorably when you show them quantitative evidence of better student outcomes. What is some of this evidence?

Robert J. Beichner:
    At NC State, we have collected data over a 5-year period on more than 16,000 students. Even though we raised our standards, we see that failure rates have dropped to 1/3 their previous values. For minorities the failure rate is only 1/4 what it was and for women it drops to 1/5 that seen in traditional sections. Other schools have seen similar trends. We take problems from exams written by traditional teachers for their students and give them to our students under identical conditions. The SCALE-UP students do a letter-grade better. One time we had a traditional student accidently sit in on a SCALE-UP test. After looking at his paper for 45 minutes, he raised his hand and asked, “Are we really supposed to know how to do these?” We’ve also tracked students in upper-level classes and found that, for example, the overall failure rate for engineering statics (the course closest in content to introductory physics) hasn’t changed. This is in spite of the fact that we fail only a third as many students in the earlier physics class. (If we didn’t fail students early but then they couldn’t succeed later, we would have done them a real disservice and wasted their time.) Students with weak math skills who took the SCALE-UP sections of their introductory physics classes actually fail the statics course at less than half the rate of students who took the more traditional sections in the introductory course. Of course, we also have lots of data from many different schools on a variety of conceptual tests. The results are clear. Students taking interactive classes learn more.

Question from ELizabeth, information officer, design firm:
    Hasn't anyone noticed that when "small groups" of students get together to work on an assignment, at least one student will totally slough off and at least one will have to do all the work? Collaboration may be a great way to solve problems in a work situation but it does not work at schools, especially commuter schools.

Robert J. Beichner:
    Actually, there are two types of students who don’t want to work in groups. The top students don’t like groupwork because their teammates slow them down. The lazy students don’t like to work in groups simply because they don’t like to work. So we try to motivate each extreme. The best students are often motivated by grades. So, we encourage teams to study together and if the team average is above 80%, each person on the team gets 5 points added to their score for “teamsmanship.” So the best student teaches the weaker students. And who better to teach than a peer who speaks the same language and just recently figured out the material. It’s been 35 years since I managed to first understand Newton’s Third Law...so I’ve forgotten what was so hard about it. We have data clearly showing that the best students learn more than anyone else when they teach their peers. To motivate the lazy students, we play to their desire to avoid work. The group writes their own contract or agreement on what each person will do to help the team succeed. If someone doesn’t pull their weight, they can be “fired” from the team. That sounds silly, but what it means is that an entire group assignment would have to be done by the fired individual...and that’s a lot of work! In a decade of doing this, we’ve only had one student fired. The team members also get to evaluate each other’s performance (and their own) in terms of the role (manager, recorder, skeptic) assigned to each person. These are done in a way that doesn’t let students see what their teammates say about them. But the teacher can quickly see if there is a problem and step in to help them solve it.

Question from Leah Macfadyen, University of British Columbia:
    Greetings, everyone, from BC. Robert, you note in your response above that you "carefully structure groups". Can you elaborate on that a bit? This is a question that comes up often, when we talk with faculty about designing group work into their courses. How can you 'carefully structure groups' if you have 400 students in your class? Leah

Robert J. Beichner:
    First we gather data to rank the students on some scale. GPA, grade in a previous course, or a concept test is sufficient. Then we make sure each group has a student from the top, middle, and bottom third of the class. (We also don't tell the students how they are grouped!) You definitely do not want students to form their own long-term teams. Ad-hoc is fine for a quick class task, but combinations like boyfriend/girlfriend or roommates in the same team cause difficulties. We actually WebAssign, a web homework system we developed, to do the sorting for us.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    In my Chronicle article, people pointed out that academics have a well-ingrained aversion to being seen as simply picking up and using another academic’s model for teaching, no matter how meritorious. They want to innovate in their own right, or at least apply their personal stamp. Given those predispositions, how do you sell Scale-Up and its rather precise requirements?

Robert J. Beichner:
    I’m not really selling SCALE-UP. I’m promoting the idea that learning is more important than teaching. We should pay attention to what we want students to learn (by writing down exactly what we want them to get out of our course--i.e. instructional objectives), carefully consider the best way to teach that material (perhaps by reading research on student understanding of the topics), do our best to design ways to help our students learn, and then check to see if they really did learn. SCALE-UP happens to be a way that has been proven to work, but it isn’t the only way. We tell people what has worked for us and, if they are interested, even provide details of the social psychology and cognitive science behind our suggestions. But we don’t “require” that they adopt every last detail of our approach. I don’t think anyone would want to do that. But we have seen that people often adjust their adaptation as they gain experience. For example, one school tried keyhole shaped tables...their next three SCALE-UP rooms are going to have round tables. Another institution put students in ad hoc groups and found little collaboration. Now they follow our guidelines for building heterogeneous groups of students and assign specific roles to each team member. Each campus should evaluate what they are doing and try to make it better. Simply lecturing because that's what has always been done is no longer appropriate. As noted by NSF in the Chronicle article, there are lots of great ways to teach that have been developed. Find one you like, adapt it to your setting, and give it a try!

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Bob, could you describe the amount of time necessary for instructors to use Scale-Up? Obviously, there is a large learning curve at the beginning because this approach is very different from the lectures with which many professors are familiar. But does this curve flatten out, and how quickly? For an experienced teacher, how does the total time required compare with that for a lecture course? And how does the time commitment for Scale-Up compare to that for the other, student-centered teaching methods?

Robert J. Beichner:
    It takes a tremendous amount of time and effort when you completely change the way you teach. Luckily, one doesn’t need to start from scratch. Many educational reforms, including SCALE-UP, offer ready-to-use materials for use in student-centered, active classrooms. It takes a lot of time to create these activities, but once they are done, it is very easy to re-use them. I find it takes me less time to prepare for a SCALE-Up class than it does a lecture. If I am lecturing, I review every single logical step before going into the classroom. For a SCALE-UP class, I just pull out the activities related to the topic for the day and walk in. My thinking is done during class instead of before I meet the students. SCALE-UP uses many of the approaches of other student-centered teaching methods, but applies them to classes of up to 100 students. The time requirements are probably quite similar.

Robert J. Beichner:
    I didn't mention in the question on forming groups that we've found it best to reshuffle groups every few weeks. If you don't, they end up becoming such good friends that group discussions are no longer talk about physics, but instead about the movie they just saw or the party they are planning! We use their scores in class to do the ranking, so it gets more accurate each time we do it.

Jeffrey Brainard (Moderator):
    We are now about halfway through our discussion today. If readers have a question, now would be a good time to submit it.

Question from amanda camp, u.s. army:
    how can systemic change in the way we teach science be brought about? can curriculum improvements be mandated?

Robert J. Beichner:
    That's really the central problem, isn't it. It has to be in people's best interest to make a change, because change is hard. Some faculty are self-motivated, others need a little more prodding. It really isn't "rocket science." Decide what you want students to learn, carefully craft instruction to help them learn it, then test to see if they really did. I think this will require pushing from the grass roots, like having reformers visit and give talks and workshops, as well as administrators pushing from above. Government funding needs to not only support innovation, but also dissemination. Progress is being made along all these fronts, but it is slow. Academic freedom rules out mandating too much, but most faculty can "read the handwriting on the wall" (or perhaps the text on the screen) and will eventually be persuaded. Everyone needs to support everyone else in the endeavor, because it is difficult.

Question from Beth Simon, UBC Wieman Institute and UCSD:
    What are your "best practices" for engaging and supporting faculty in embracing pedagogical change in their classrooms?

Robert J. Beichner:
    We do workshops where teachers practice the skills they need in an interactive classroom. What seems to work very well is having faculty sit in on a class a few times, gradually increasing their role from observer to full participant. It is also important to provide ready-made teaching materials and rationale for doing things a particular way.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Bob, some of these teaching methods have existed, and shown good results, for more than 10 years, yet they have not been widely adopted. What is being done to get out the word, and what is working on this front?

Robert J. Beichner:
    The most effective thing I’ve seen for advancing reform is being done by a collaboration of the American Association of Physics Teachers, the American Physical Society, and the American Astronomical Society through funding provided by the NSF. New faculty in physics and astronomy attend a 2 day workshop where the results of education research and the implications for teaching are described. Ways to use the latest technology are demonstrated along with discussions of today’s diverse student body. But there are also sessions on how to balance the demands of academia with real life, as well as talks by program officers from government research funding agencies on how to get grants. Over the dozen years it’s been offered, nearly 1000 young faculty members (representing about 2/3 of all PhD-granting institutions and 1/4 of all baccalaureate institutions) have seen examples of state-of-the-art, research-based college teaching. Ken Krane of Oregon State initiated this massive effort. It needs to spread to other disciplines. Young faculty seem particularly open to hearing and applying new ideas in their classrooms. As far as answering your question about the best explanation for the slow rate of adoption...I don’t think it can be pinned down so readily. There are many factors that impede progress: time and monetary constraints, lack of interest, lack of rewards, even direct, hostile opposition. There are few positive motivators to counter these disincentives, other than altruism. Changing priorities by university administrations will shift resources and rewards somewhat. Expanded use of conceptual assessments to pinpoint problems will help. And ready availability of research-based, classroom-tested alternatives to the lecture will lower the barriers to adoption. Complex problems require multi-faceted solutions. Besides, there is always PBR...Progress By Retirement!

Question from Juliet, Delaware State U.:
    How large is your typical class and what percentage of the work is group work and what percentage is lecture or do you use only group work?

Robert J. Beichner:
    My typical intro physics classes are 99 students. I'd say 3/4 of the time is group work and the rest is "lecture." Keep in mind that most of the "lecture" is actually a directed classwide discussion. " Group 4a, tell us about your graph. You've got cusps on the bottom, but group 6b does not. What does that graphical feature indicate?" Different faculty have different time divisions. I tend to lecture less than most and restrict lecturing to "big picture" organizing and motivation.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    To follow up on an earlier question, what are some other ways that you use to try to persuade skeptical faculty members about your method, especially those who defend relying on lectures? These skeptics argue, for example, that in introductory courses, students new to a discipline don’t know what questions to ask in an inquiry-driven exercise.

Robert J. Beichner:
    I absolutely agree that students need to have a common, fundamental base of knowledge. That’s why faculty shouldn’t only be relying on lecturing! Most students don’t learn well from lectures. (A few do. They go on to become professors.) I am merely suggesting that faculty should pay more attention to learning than they do to teaching, because that's what really matters. I’m not saying that students have to recreate hundreds of years of scientific advances during a semester. That obviously doesn’t work. What I am saying is that science is not just sitting and listening to someone talk. It is investigating things, asking questions, changing your mind, convincing others of your theories, etc. Why not let students practice that WHILE they learn the content of science. It isn’t a zero-sum game. In fact, students learn more content while also getting a better appreciation for the nature of the scientific enterprise.

Question from Pamela J. Hines, SCIENCE:
    How can promotion and tenure analyses of faculty be restructured to reflect professional progress in both teaching and research? What would the right balance be, and what would be the measures of ‘production’ in the teaching component? How would you analyze contributions to innovation in teaching, or even improvements based on taking on someone else’s innovation? I would also like to remind your readers that Science is now publishing articles on science education and the science of education, giving faculty an opportunity to write up their teaching innovations (and analyses of their effectiveness) for publication. The section is called the Education Forum, and can be viewed here: http://www.sciencemag.org/sciext/educationforum/ Pamela J. Hines, Senior Editor, SCIENCE

Robert J. Beichner:
    The traditional academic roles are teaching, research, and service. I think it is unreasonable to expect someone to excel at all three, although sometimes it happens. Why not let faculty, in negotiation with their department, decide on which two of the three they will focus their energies. At the same time, they should agree on appropriate metrics of success. (Teaching quality can't just be based on student evaluations, for example.) The idea of treating teaching as a scholarly activity is called, not surprisingly, "The Scholarship of Teaching and Learning." The book I mentioned earlier by Charlie Glassick specifically discusses assessing this sort of thing. In addition to the coverage in Science, there are now quite a few journals covering research on education in the disciplines. I happen to be the editor of the American Physical Society's journal called Physical Review Special Topics: Physics Education Research." Our standards are very high and should help tenure or promotion-seeking education researchers.

Question from James Meyer, St. Gregory's University:
    It seems to me that your collaborative groups merely are merely in-class homework assignments where collaboration is encouraged, possibly replacing them; if that is so it seems to be a step backwards. The method also seems to favor those who on the Myers-Briggs personality inventory are the "extraverts" and works against the others. How do you determine and make homework assignments within these collaborative methods?

Robert J. Beichner:
    We use group roles (manager, skeptic, and recorder) which rotate with assignments. The Johnson brothers at the University of Minnesota have LOTS of research on the general approach. Students evaluate team member performance in terms of those roles. So, for example, an extravert might have been assigned to record data. If they spent most of their time talking instead of recording, then their evaluations would reflect that. We ask students to work in and out of class in their groups, by the way.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Many in higher education seem to worry that the current push for better results of teaching and more “accountability” among colleges will translate into unreflective conformity and standardization. To what degree does Scale-Up, or newer teaching methods like it, represent a standardized approach? What are the pros and cons if people modify and adapt it?

Robert J. Beichner:
    The only thing that is standardized in SCALE-UP is the “big picture.” As long as students are actively working together on interesting activities while the instructor circulates and asks questions, I’d say they are following the SCALE-UP approach. I personally think the round tables are the most important technology in the classroom, but not all adopters agree with me. It really is a philosophical change. Instead of the instructor being the source of information, nature (or the source documents, or the original speeches, or whatever is being studied) is the authority. In this age of information abundance, I think this is the way we have to go. Just as changing from hunter/gatherers to fast food fanatics caused huge changes in society, we are now seeing information become instantly available. A sixth-grader’s laptop is better than the entire Library of Congress because of the incredible indexing and search capabilities available to them. So unlike the past, simply transmitting information is no longer the role of education. We have to train our students to efficiently “farm the information fields.”

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Many students say that a leading reason why they become frustrated in introductory STEM courses is that a large number of teaching assistants are foreign-born graduate students whose English can be heavily accented and/or poor, and so difficult or impossible to understand. How big an issue is this? Will efforts to spread the new teaching methods founder unless this problem is fixed first?

Robert J. Beichner:
    This can be a real problem in classes where interactions between teachers (or teaching assistants) and students are important. There are also cultural differences that can make it difficult for some foreign graduate students to aggressively question their students. We have found that providing an initial semester where the foreign graduate student is an observer in class improves their language skills and gives them opportunities to see how faculty and experienced graduate students can ask leading questions to guide student thinking.

Question from Dave, Big State U:
    Be honest, which has advanced your career more, your research or developing this teaching strategy? And how did the time and effort involved affect your research? Should tenure-track faculty members undertake education reform?

Robert J. Beichner:
    Actually, I'm extremely lucky because developing and evaluating this teaching strategy IS my research. It turns out that more and more departments are opening up faculty lines to people wanting to do education research. It's very fast growing. In fact, my group's biggest problem is that our grad students often get job offers before they've finished their dissertations. They get a PhD in Physics from NC State, have completed the standard qualifying exam (and are often among the top scorers) and successfully defended a rigorous research project. They can get positions in industry (any high-tech company that needs to explain its products to the public likes to hire people trained in helping people understand complex ideas). Academic choices include teaching at small colleges, doing some research at middle-sized institutions, or joining one of the growing number of R-1 university departments conducting discipline-based education research.

Question from Emil Chuck, George Mason University:
    I'm impressed how SCALE-UP seems to work better than straightforward lecture-only courses. That said, there is also an argument that lab sections should be the opportunity where these small group discussions should take place (as is the case perhaps with the engineering physics course that you have at NCSU). Could you comment further on how this is different or similar?

Robert J. Beichner:
    We have data showing that small classes doing SCALE-UP activities learn more than large classes doing the activities. Nonetheless, students really appreciate the opportunity to interact closely with the professor, which they wouldn't get in a lab taught by a grad student.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    To follow up on Bob's reference to faculty positions for people wanting to do education research: a growing number of institutions are creating Ph.D. programs in science education. But do these positions tend to let research-oriented faculty members off the hook, because they can say, “Those people can worry about teaching, and I can focus on my research.” How can departments and faculty members deal with that perception?

Robert J. Beichner:
    These programs are getting more and more common. Because they are carrying out rigorous research on learning and then creating and testing new pedagogies and curricula based on those studies, faculty at other institutions are listening. Not only is there solid data showing improved learning, but there are also well-developed techniques faculty can adapt to their own classrooms. These PhD programs are producing future faculty members who bear the same responsibilities as anyone else at research institutions: teaching, research, and service. And because of the training they receive, they are generally very good at all three! I recall when I was being interviewed for my current position at NC State, I asked, “Are you looking for someone to get a good research group going, or do you want someone to fix your introductory courses?” (I didn’t want to be considered a second-class citizen of the department.) The blunt response was exactly what I wanted to hear: “If you don’t get a good research group going, you won’t be here long enough to do anything in the introductory courses.” In other words, I play by the same rules as all the other faculty. So discipline-based education researchers make good faculty colleagues for the same reasons as anyone else.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Some evidence suggests that undergraduates choose not to major in STEM because average grades in these courses tend to be lower than in non-STEM courses. Students concerned about attaining a good GPA know this, and choose accordingly. In explaining why more students don’t major in STEM, how big a factor is this cause compared with inferior teaching in introductory courses? Can or should anything be done about grading?

Robert J. Beichner:
    I think poor teaching and low grades are related, and the combination does drive some students away from STEM courses. My ideal path is to continually strive to improve my teaching. As this happens, students learn more and grades naturally go up because they are based on attainment of clearly-specified objectives. (I have never understood the rationale behind “grading on a curve.” Why should one student’s grade go down because someone else’s is high? Why would a student want to help anyone else learn in that type of setting?) When I see grades creeping up, I consciously adjust my course objectives to increase the content and/or depth of what is covered. This brings the grades back down, avoiding long-term grade inflation. The process then repeats. Thus, this year’s version of my class should be a better “product” than what students had 10 years ago, just like a 2008 car is clearly superior to those produced in 1998.

Question from Ken, Faculty Developer:
    Hi Robert. I am sure you would agree that this is an issue that needs to be addressed from all angles. My question to you is, what can we do to assist new faculty or graduate students aspiring to be faculty? What can we say to have them agree that the personal investment of their time (and often a lot of it) will be worth the improved learning outcomes?

Robert J. Beichner:
    The advice I give my grad students is to "Get it in writing!" Negotiating a statement of mutual expectations about time division between teaching/research/service and how to assess performance is critical. Things like the NSF's CAREER awards help everyone recognize the importance of both teaching and research. The best way to reach physicists is to show them data on learning and give them examples of the physics done by students taught via interactive methods.

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Will efforts to get more students to study science and engineering struggle as long as many Americans stereotype scientists as geeky drones? It would seem like college faculty members have more direct control over the quality of their teaching than they do over this widespread cultural perception.

Robert J. Beichner:
    Kids start out as natural scientists and we somehow manage to drive that out of them by the 8th grade or so. It isn’t totally hopeless, however. Look at the popularity of people like Bill Nye and even shows like Mythbusters. Biology departments report a huge increase in enrollments in forensic science due to the interest in the various CSI shows. Many of today’s scientists and engineers cite Star Trek as an important early influence. I’ve even seen T-shirts that list the “Top 10 Reasons for Being an Engineer” and one of them is “Scotty was an engineer!” Maybe what we should be doing is talking to the people in Hollywood!

Jeffrey Brainard (Moderator):
    A National Science Board commission has mulled the idea of a national advertising campaign to improve scientists’ image -- maybe our readers could propose some slogans. Perhaps along the lines of, “Got Quarks?” (I didn't major in marketing.)

Comment from Leah Macfadyen, UBC:
    This isn't so much a question as a follow up comment on the exchange, above, about cultural and linguistic differences in teaching and how this can make for unhappy TA experiences. At UBC, our Centre for Intercultural Communication has been running an 'International TA Training' program for ten years or more - I've co-taught it once or twice, so I've seen the ways it makes a difference. It's not rocket science either. It combines some very elementary training in 'instructional skills' and 'communication skills' (something that most homegrown students could use, as well, by the way) with reading, activities and discussion on 'the Canadian educational context'. We also do classroom observations and feedback for students who ask. Evidence suggests that this rather light touch really helps those TAs understand their new working culture, and also gives them more confidence in their ability to communicate in English to a class of students. Leah

Question from keith Barker at UConn:
    You mention technology and the abundance of information. I wonder if you allow students access to the WWW during your round-table classes?

Robert J. Beichner:
    We do let students access the web, but sometimes we make them "close the lids" on their computers, if they are being distracted. (Today's students are able to multitask much better than I can.) We are able to view their screens remotely also. Related to this is the importance of evaluating material found on the web. For example, students needing the mass of a racquetball for an activity would rather do a Google search than go to the closet and get out a scale. So they have to justify why they can trust the value they find. The usually end up at the US Racquetball Association website for the number (41 grams, if I remember correctly). Then they point out the container the ball came in, which says the contents are approved by the USRA. That's good enough for me!

Question from Jeffrey Brainard (moderator), Chronicle of Higher Education:
    Is it realistic to expect faculty members, especially at research universities, to excel at both research and teaching? Are we expecting faculty members to be super-human? Won’t most STEM departments at research universities view calls for improved teaching as requiring a unilateral disarmament in the race for a good national ranking, which is based on research productivity, traditionally defined?

Robert J. Beichner:
    If teaching and research were entirely separate then it probably would be unrealistic to ask someone to excel at both. But most faculty teach in areas related to their research. Granted, an astrophysicist might not talk about Kerr-Newman black holes in her introductory classes (although it would be nice if she did), but certainly the related topic of angular momentum will come up. Also remember that working on research with graduate students is also teaching, probably some of the most rewarding that is done by faculty members. On the other hand, consider how we speak of “teaching loads” vs “research opportunities.” I’m afraid that really points out how many of us view our academic tasks. As Randy Bass (Georgetown Univ.) noted, a “problem” in research is exciting and something you discuss with colleagues. A “teaching problem” is embarrassing and we can’t wait for the semester to end. To return to the last part of your question, why should continuing to teach via obsolete methods be required for our advancement in the national rankings?

Question from Emil Chuck, George Mason University:
    One more question from me: what results have you had from non-science-majors taking introductory classes using your method?

Robert J. Beichner:
    At NC State, we've focused on science and engineering majors. (Departments other than physics and chemistry are now in the process of collecting data.) But other schools have lots of data from a variety of courses and student populations. There is a chapter describing some of this data available from our website. (Click the SCALE-UP link in the Chronicle article.) If you want even more detail, go to the page http://www.ncsu.edu/per/SCALEUP/Classrooms.html that shows pictures of various classrooms and contact the people listed there.

Question from Jane, mid-sized institution:
    Would you please summarize the data that shows students learn more with this method.

Robert J. Beichner:
    We have data from half a dozen conceptual tests in different areas that show clear differences (i.e. statistically significant). We've also randomly sampled exams written by lecturers for their students and gave them to ours under identical conditions. In 90% of the questions, SCALE-UP students did significantly better. We've tracked performance in later classes and find that there is no change in failure rate, even though fewer students failed the earlier SCALE-UP prerequisites. See the chapter on our website for more details.

Jeffrey Brainard (Moderator):
    Just to repeat it, the address for Bob's web site on Scale-Up is http://www.ncsu.edu/PER/scaleup.html

Jeffrey Brainard (Moderator):
    That concludes today's discussion. Thank you to our guest, Robert Beichner, and to everyone who participated. A reminder: Congress is considering this week a bill to authorize increased federal spending to improve science teaching at all educational levels -- and The Chronicle will be following the bill's progress. Good day.