No, this is not about angst in taking exams. It is about teaching intro science classes.
The New York Times just ran a piece about how teaching science in intro lectures is becoming more dynamic in an attempt to be more successful. This is old news at the University of Colorado, where we had the Science Education Initiative start systematically adjusting science education for more than 7 years. Much of the emphasis is on small group activities within lecture classes. And yet the news story made the Grumpy Geophysicist scramble over the Times webpage looking for a place to comment. Well, no such luck, so the complaints have to be here.
First one point GG agrees with was this quote: ‘“Higher education has this assumption that if you know your subject, you can teach it, and it’s not true,” Dr. Uvarov said. “I see so much that I was missing before, and that was missing in my own education.”’ In general, faculty are the ones who succeeded at getting material from books and classic lectures and so may be the very least prepared to help those for whom classic lectures and texts really don’t work. The recognition that faculty need help before teaching a class is a helpful one.
Before going further, a disclaimer: GG is not a great teacher. Probably not even good and maybe seriously subpar. So caveat emptor.
The sentence that set GG off: “There are many explanations [for why more dynamic modes of teaching are not employed], educators say, including the low value placed on teaching, tradition, pride and the belief that science should be the province of a select few.” [One should wonder who the “educators” are that inform this quote, as they are clearly not the educators teaching the classes in question].
“Belief that science should be the province of a select few”!?! Really? That is way off base. Does anybody really hold that opinion? While it is undoubtably true that some are better at doing science than others, that doesn’t mean that only a few should understand science.
First, some general comments. Yes, there are faculty who will not change how they teach an intro class (or a grad class, for that matter). It usually takes 2-3 times teaching an intro class to develop material fully enough for the class to be taught without working nearly full-time on it; asking a faculty member to go through that process again (when they are senior enough to have all the committee work and research responsibilities filling their days) will meet resistance. Is this “tradition” or “pride”? Not so much: it is prioritizing work (so the “low value placed on teaching” can indeed play a role here). And some classes, once developed, need little tuning (does basic physics change year to year? No) while others do (does paleoclimate change? Ur, more than we might like, so each iteration of a historical geology or paleoclimate course is likely to require updating).
But there are many more obstacles present, and there is also a bit of a one-size-fits-all issue. One obstacle is simply the rooms available for intro courses: it is hard to run small group activities in a room with 150 chairs all bolted to the ground and facing forward (let alone 800 in some really large lecture halls). Another is the continuing squeeze on faculty to teach more students, leading to larger classes (see GG’s screed on MOOCs for more). In general, moving to the “guide on the side” mode of teaching means covering less material, which introduces a tension between how much material is covered versus how much time is spent on small group activities. In classes where a very specific set of skills needs to be mastered (as, say, in intro physics), it is clear that mastering 50% of the material is better than mastering 25%. But in some more fluid classes, mastering 50% of 10 goals may or may not be superior to mastering 25% of 20 goals.
Frankly, a lot of the motivation is arising from what has been lost from years gone by. Students are generally not so well connected to one another, so the concept of studying in a group is generally absent. Recitations, which were wonderful places to be able to interact with a TA, are often missing now. As a result, the small group work in lectures is often absorbing the work that used to be done in these other arenas.
It certainly seems that these methods have produced great gains in physics, for instance, where in some instances the same faculty teaching the same course with and without the kind of group activities discussed here yield very different results on post-course evaluations. It is hard to argue with those results, and in general physics is now taught using these techniques at CU. But there is a cost: there are a large number of TAs who help to conduct the in-class exercises and some associated activities; the faculty who teach the monster intro courses are usually overwhelmed in that term with little time for much else.
In Geology, we have no TAs for these courses (we have used our allocation of TAs for our intro labs, which are small enrollment). Some faculty have pioneered recruiting upperclassmen as learning assistants (LAs) for the big lecture courses; these students are paid hourly wages by the department.
Anyways, unlike Physics, where there is a consensus on what skills students need to master, there is no equivalent consensus in Geology. As a result, there is no single evaluation tool to see how well or poorly different teaching philosophies work, though at least for Physical Geology (our 1010) there is well-vetted set of questions that can be used to make a learning evaluation exam [these are different from a typical class exam in that the questions are built to consider specific learning goals and are generally developed from interviews with students to identify the most common misconceptions as well as to avoid mis-stating the question, which leads to the student understanding the question as something different than what is intended to be asked]. In Historical Geology, which GG teaches, there is only a subset of questions available to use in making a pre-/post-course evaluation.
Additionally, there is a real risk that some perceived gains are somewhat like teaching to the question. Not in a deliberate way, but for instance a colleague once reported that his students were now acing a concept he had previously had trouble getting them to understand. But the exam question was phrased in the same way the in-class exercise had examined the topic. GG wondered if a better question was one that approached the same concept somewhat differently; it turned out that the answers to that question were much farther from ideal. So while there had been some gain–the students could follow the same process as in class–the concept itself was not as successfully taught as it appeared.
Historical Geology in particular poses challenges to the interactive, student-led learning paradigm. Consider, say, the concept of a proxy. A geologic proxy is a measurement of something in a rock that tells you about some other condition. So, for instance, the ratio of heavy to light oxygen in seawater tells you how much ice is in ice sheets (this is because light oxygen will evaporate more easily than heavy oxygen and so ice tends to be enriched in light oxygen). This oxygen ratio is captured by foraminifera, which incorporate oxygen as part of their shells. So measuring the ratio of oxygen in ancient seashells can tell you how much ice was in ice caps at the time that creature was alive. Nearly all paleoclimate work depends on proxies like this, but how to best teach the concept? You can hammer on the ice proxy with multiple examples, breaking it down into little pieces, etc. Or you can show it and other proxies (e.g., carbon isotopes can inform the level of oxygen in the atmosphere; ratios of certain elements in basalts can inform the amount of crust that has been made) and try to tie them together to abstract the common elements of a proxy. The first has the advantage that as the material is relatively limited, group projects and student-led learning fit well into the course timeline. However, it also carries risks: the particulars of that proxy don’t apply to all proxies, and so students can mistake certain particulars of the oxygen isotope proxy as general characteristics of all proxies. Additionally, there is a bit of catering to the weakest students in “hammering” on anything: if you spend an extra week on getting from 50% to 80% mastery of the topic, there is a fair chance that some of the 50% who already got it are getting bored and unhappy with the course. The second approach though requires a much higher ability of the student to abstract concepts from specific examples; while a student succeeding at that will have a far firmer grasp of the concept, failure is apt to be more common. Also, as there is simply more material needed to be presented to make this connection, there is less time for group work and interactive elements to class.
A great advantage of the SEI program at CU has been that faculty now discuss how to best teach intro classes. When GG started, he got a few syllabi and class notes from other faculty and, well, that was it. The level of support now for new faculty is much greater and the recognition that faculty need instruction to be good teachers has increased. These are tremendously positive developments independent of the tools being used in any given class, and arguably this more than the “guide on the side” approach has been the greatest contribution to improving science teaching at CU. But there is a real temptation to mistake change for advancement; care is needed to be sure that the changes suggested as in the New York Times piece are positive.