How important are labs for learning science?

Steve Gimbel has a provocative post that suggests the costs of undergraduate lab classes may outweigh the benefits. Quoth Steve:

[E]verything I know about physics, I learned from my theory classes. You see, science classes come in two flavors. There are theory classes where a prof stands in front of the room and lectures and then there are lab classes where for many hours, students walk in ill-prepared and tried to figure out which one of these things we've never seen before is a potentiometer, fumble their way through procedures that yield results that are not even close to what they were led to expect, and then plug and chug their way through scientific and error calculations that frankly mean little to them. I will freely admit that all my experiences in lab classes were a waste of intellectual time and curricular space that could have much better utilized.

Now, I'm supposed to be writing a serious academic paper right now*, but Steve, as a fellow philosopher who is well aware of my misspent scientific, actually emailed me to see if I'd weigh in on the (as did another blogger coming at the issue from the perspective of a working scientist). And, coincidentally, just the night before Steve published his post, my better half and I were reminiscing about our undergraduate experiences with laboratory classes. So really, what choice do I have but to respond?

As I read it, Steve 's post is advancing roughly the following argument:

  1. Science pedagogy assumes that the theoretical content of a science course (what scientists in field X know about the phenomena they are concerned to study) is intimately linked to the methodology by which scientists in that field probe the phenomena. In Steve's words, "You learn the theories, but how do you know they don't really work, if you don't test them? Science has two parts -- observing phenomena and accounting for them, science education ought to cover both".
  2. Actual lab courses, however, often amount to following a ready-formulated procedure, trying to get the equipment to work, arriving at results that don't clearly demonstrate the concept or relationship they are intended to demonstrate, and following more recipes for computations that don't connect to anything much like "understanding" in the student's head.
  3. Since the undergraduate's understanding of science seems more likely to come from a lecture course than a laboratory course (at least if the laboratory course is as characterized in #2), students might get a better understanding of science if they elected to take more science lecture classes.
  4. However, when science lecture courses require that students also take the corresponding lab course, the disincentive to take more than the minimum number of science classes is huge, because the lab course will chew up a whole afternoon (or more) of the academic week (when students might well be taking some other courses they want or need to take).**
  5. From the point of view of allocating faculty, the demand for a relatively small student-to-faculty ratio in lab sections means that a science course with a largish lecture and multiple lab sections can account for one science faculty member's entire teaching load in a given term.
  6. This means, in turn, that the faculty in science departments are likely to have less opportunity to teach additional science offerings -- electives, courses team-taught with faculty from other disciplines, etc. -- that could broaden and deepen students' understanding of and appreciation for science. Even if students could get over the tendency toward the minimum number of science classes described in #4, these potential additional science offerings aren't made actual because of the staffing demands imposed by lab classes
  7. Therefore, requiring a lab class for the lion's share of science lecture classes undermines the goal of getting more students an understanding and appreciation for science.

Now Steve is careful here to distinguish between what is needed for science students who are planning to become scientists and science students who are not:

Those who are being trained to go on in science do need training, they need to learn how to work in labs. But there is a difference between training and teaching and the overabundance of labs put the training aspect before the teaching aspect. Most students don't need training, they simply want to be taught about science -- and for most of them they are right, that is what they need... Students want to know about science and need to know about science, but are not learning about science and part of the reason is the idea that learning science means you have to be trained to do science.

On the one hand, I'm in complete agreement with Steve that the world would be a better place if more non-science majors*** took science courses for fun. Having at least a basic understanding for how science works strikes me as a crucial competency for an educated person in the 21st century, and curricular changes that increase the odds of developing this competency ought to be investigated. And it's true (as I've noted before) that the handskills one develops in a lab are distinct from conceptual understanding -- and, they're distinct from being a competent observer and from being able to draw good judgments from data.

However, I think there's something to the impulse to include laboratory experiences in science classes. The important thing science teachers hope to teach, after all, is not some particular pile of facts or some canonical set of models of their phenomena. Rather, they want to convey something about the methodology of science -- here's what we know about these systems and here's how scientists came to know it. Given the kinds of information we can access in these systems, here are the kinds of questions we can answer, the kinds of accounts that seem reasonable (and those that we've ruled out), the sorts of problem-solving strategies that give us traction. Without that part of the story about science, students don't really understand what scientists take to be distinctive about how they're grappling with the world.

The big question, of course, is whether one must mess around with the actual systems to grasp these lessons about the methodology of science. A much easier question is whether undergraduate lab classes are a sure route to grasping such lessons: on the whole, many are not.

A good part of this can probably be laid at the feet of lab exercises that have more to do with following cookbook protocols than thinking. And, as I've speculated before, the cookbook protocols may seem like a necessary evil when you only get four or five hours a week in the lab, for however many weeks there are in the academic term, to transmit all the canonical experiments for a particular level of study of a particular scientific discipline. The kind of trial-and-error exploration involved in thinking up your own approach to a scientific problem could well take more time, and the sheer time investment is already one of the things that may be driving students away from taking more science courses.

Before I go on, let me share some details of my chat with my better half about our undergraduate laboratory experiences. At the science and engineering oriented institution where my better half was an undergraduate, lecture courses and lab courses were wholly distinct, which is to say you could take all manner of science lecture courses without ever setting foot in the lab. (This came up when my better half copped to never having taken a physics lab, despite having taken a bunch of physics lecture courses.) I, on the other hand, attended a small liberal arts college where one not only had to take the corresponding lab course for nearly all the science lecture courses offered**** but one also received precisely the same number of credits for taking the lecture and lab portions of the course as one would have received for taking a much less time-intensive course such as a philosophy seminar.

Perhaps paradoxically, this means liberal artsy me may have logged more hours in lab courses of various sorts than my techy better half. This wasn't because I thought from the start that I was going to be a scientist.***** And, though some of the red ink from my old lab notebooks indicates that my lab technique struck my professors as "better than most", I can tell you that at least at the introductory and intermediate levels of study, there was next to no correlation between my grasp of the conceptual subject matter of a science course and my ability to get the stupid lab experiments to work, let alone to grasp any additional understanding of the phenomena beyond "This sucker is a pain to control precisely!" I might well have been a physics major on the sheer elegance of theory, but the damned labs (which started at 7:00 PM, by the way) were a perfectly optimized blend of impossible to conduct and unlikely to convey new insights. To be honest, my first couple chemistry labs weren't much better (and I still bear a grudge against the inventor of the bomb calorimeter that makes me thankful that I'm neither a violent person nor armed, lest I encounter that inventor or his progeny).

If my recollections are to be trusted, it is fair to say that I became a science major despite these early college laboratory experiences rather than because of them.

However, a funny thing happened around my junior year: my labs (in chemistry and biology both) started to be engaging.

How they got to be engaging was by presenting us with laboratory problems for which we needed to work out our own strategies, protocols, follow-ups and interpretations -- in other words, we had moved away from cookbook land. Also, the experiments were no longer cover-the-whole-thing-in-one-lab-session procedures, but developed into ongoing experiments that we had time to mull over, adjust (based on how things had gone the last time), and digest. In short, what we were doing in our lab classes was suddenly a lot more like "real" science.

Removing the insane time pressure and shifting the terms of our success to how well we thought as we planned our experiments (rather than on whether we were skilled enough technicians driving someone else's protocol) surely made the labs more fun to do, and I definitely learned more about the process from these labs than I ever did from my intro labs. I couldn't have gone right to these relatively advanced labs, however, without a bunch of the techniques I learned in the detestable intro labs -- such techniques were like a necessary vocabulary for framing the problems we were asked to solve in the "fun" labs and conceiving of reasonable plans of attack. If I had been at a school where the lab component of science courses is optional, the nature of the entry-level labs might well have discouraged me from electing to take the lab components of these courses where the labs turned out to be really engaging. (No one had warned me that they would be really engaging, so it was a marvelous surprise.)

In some ways, then, I'm a beneficiary of the forced march through the science lab.

Still, if the point of a lab course is to convey more than specific techniques -- if the point is to cultivate something like an understanding of how experiments can be used to build scientific knowledge, and a feel for what it's like to engage in scientific problem-solving -- rethinking the design of intro lab classes might be a very good thing. It's good that this is starting to happen. But perhaps the real measure of success will be when the majority of students who don't have to take lab courses enroll because they want to take them.
*The paper must be ready for submission by April 1. I've made significant progress on it, but there's lots more to do. However, at the moment my brain is fried, so you get this post.

**Additional fees for lab courses can be another disincentive, especially for those of us with a tendency to break a lot of glassware.

***And science majors too, for that matter.

****The first time I actually encountered such a course was when I took a labless seminar on organometallic chemistry in my junior year. It was kind of weird.

*****I was pre-med for a spell. It passed.


More like this

Semester-long independent projects. You need killer TAs to do it right, but that is the best way to do a lab.

This came out of an independent project. The bottom figure of this came out of an independent project. Sophomore core course for biology majors. They forgot all their dissection and cookbook experiments, but years after, they all remember their projects.

I've already come down in favor of doing away with lab courses. Because the ones I'm familiar with have all been horrible.

It would be great to give undergraduates a feeling for how experiments are done. But that's much more effectively accomplished by having them do something project-oriented (whether it be real research or something more purely pedagogical) than to step through a sequence of pre-prepared labs. Which, I may have mentioned, are always horrible.

I think that too much of a premium has put on brute force effort in our undergrad and graduate teaching experiences. There is actually a lot that a merely warm body can accomplish in a lab, however rigorous theoretical experience tends to come mostly from course work and I would say this is lacking from many graduate schools and certainly from undergraduate training. This is kind of a sore spot for me because I feel like I appear to be somewhat successful but I feel woefully untrained in actual science even though I quite well know my way around a lab.

I'm an undergraduate engineering student. I've found that, in general, labs are time-consuming and don't help me learn theory. For mastery of a subject, labs are largely useless--at least, for me. Well-done labs can be useful for gaining practical information about how to use the equipment. I wouldn't know how to use an oscilloscope or build a circuit without a lab. However, I think my physics labs were largely useless, though some were cool.

I think that if people running the labs made sure that the equipment actually works, labs would be far more useful and pleasant. Much of my time in lab is spent trying to figure out what's wrong with the equipment, finally calling over a TA who can't fix it either, and trying to scrounge up another piece of equipment to use.

By far the most engaging intro labs I had were physics. This was because we had to design all of the protocols ourselves, based on the equipment available and the theories we had learned in lecture up to that point. I can no longer recall the formula for moment of inertia, but I vividly remember spinning my lab partner, with barbells in her outstretched arms, in order to determine hers.

Labs are also good for showing why a particular theory is actually pertinent to the actual physical world in which we exist. And they address core competencies that practicioners are expected to have. As a geologist, people are constantly asking me "what kind of rock is this?" But that question is extremely hard to answer using theory alone.

I never ended up taking a single science class after high school other than math (unless you count the history of physics class I sat in on for most of a semester a few years ago in grad school), but in thinking much more about philosophy of science, and talking to my boyfriend (who is a chemist) I've come to have a much greater appreciation for so much of what was going on in those science classes I did have as an undergrad. And now I wish I could do some of those labs again, now that I've got more of a perspective to see why those experiments were interesting the first time they were done. It's too bad that they're forced on unwilling high school and early college students, rather than people who have more of a perspective to have an interest in them.

I think this may be a problem of emphasis; if you're made to feel a failure for not getting the expected outcome from 'classic' experiments or standard techniques, then it will be frustrating and counterproductive. But with the right attitude you can often learn a lot from these 'failures'. Back when I flirted with Cell Biology, every lab I did uniformly went horribly wrong - but this also meant that I had to do a lot of thinking about why they went horribly wrong, which often meant that I came out understanding the experiment and processes involved much better (and I did surprisingly well in the evil end-of-year practical exam). Likewise, in my Physics practicals there was a huge concentration on identifying the sources of, and quantifying, random and systematic error - something which I've been highly appreciative of ever since.

So, if the emphasis is on more than getting the 'correct' result, I don't think that it's a bad thing to learn that these things are difficult to get right; otherwise people don't really learn what doing science is actually like.

Of course, some subjects face more challenges than others in this area; Geology labs are not so much experimental as observational and interpretative - the only way you can really do experimental geology is with an extended project.

Most of the complaints here seem to be by scientists about their intro level courses, which are not necessarily the kinds of courses the nonscience people are taking to fulfill their science credits. I know at the small liberal arts school that I teach at, we offer those 100 level courses for nonscience majors that very rarely see a science person in them.

If you're trying to get nonscientists interested in taking science courses, standing up in front of them and talking on and on means a significant portion of their brain just turns off. Now, put them in the lab and make them actually do something we've talked about in class, and they are involved and talking about it and it becomes much more real to them. That being said, I often take a different approach in these courses than "apply the scientific method to this topic," and I'm teaching in Biology rather than physics, so maybe it's a little easier to do there.

Whenever we brought undergraduates into any research lab (i.e., independent study or work study), we always assumed that they needed to learn every technique. The labs don't teach technique, and they certainly don't teach what doing 'actual' research is like.

I think that the subject makes a difference. I can see how one could learn a lot about physics without ever stepping in a lab, I never learned much in any physics lab I took. In general chemistry the only labs that helped me were the ionic crystal lab and rate reaction lab; however, I don't think I would have made it through organic without the labs. I learned more in the labs then I did in lecture, that is where it all came together. I don't think one can really teach biology without labs in at least some of the classes. Labs in biology (and geology) often involve investigating the subject you are talking about in lecture, dissections, studying structures under a microscope, going into the field and identifying plants, finding and identifying animals, etc.. Things one could never really cover in a lecture, and you had better pay attention because the lab practical will be coming up soon and they are usually harder then the lecture exams.

Undergraduate labs should not be about training undergrads to be lab techs, every lab is different and need techs with different skills. Undergraduate labs for science majors should be about teaching the students science and getting them comfortable working in a lab. Undergraduate labs for non-majors should be for keeping their attention and getting basic concepts through like osmosis, behavior of salts in solutions and other simple concepts that a majors student should be able to grasp without a lab but where a lab may really help a non-science major.

Labs do take a lot of resources, both time and budget, but doing science takes time and budget so I don't think one can train scientists without lab classes. Sure we could hire a professor or two at the CC I am at if we cut out all the labs but would we actually teach more or just offer more options with less content.

Good labs are essential motivation for me. After spending too long following theory and extrapolating, I start to wonder how real any of this is. There's only so far I can go out on a limb from a familiar and secure trunk before I have nagging doubts about what I'm doing.

But then comes an experiment where I personally test some of the predictions and... holy shit, it actually works! Then, I feel much more confident about what I've learned and can take the next step forward. Things like the dependence of Johnson noise on temperature and bandwidth, and 90 degree phase shift at resonance, and how Bode plots can predict stability and oscillation. They were all a little mysterious at first, but after seeing them happen in front of me, and having the chance to play, I can reason about them without being distracted my doubts.

Now, as I've done this more, there's also an increase in how far from experimental evidence I'm comfortable going. So the theory learned is superlinear in the number of experiments conducted. But it's still nice to see experimental validation occasionally.

Thinking about it, the doubts aren't quite that what I'm being taught is wrong, but that what I'm not being taught is important. Pretty much all of science depends on simplified models. Some effects are important, and worth calculating, and others are insignificant and can be ignored. Real systems are affected by all sorts of things, and you can spend eternity calculating minor factors which turn out to e imperceptible.

The thing I appreciate the chance to see is "the real world actually behaves a lot like the simplified model we've been discussing in class". The simplifications are actually justified, and we haven't forgotten anything important.

By Heath Robinson (not verified) on 29 Mar 2007 #permalink

How they got to be engaging was by presenting us with laboratory problems for which we needed to work out our own strategies, protocols, follow-ups and interpretations -- in other words, we had moved away from cookbook land.

I'm totally in accord with this view. Too much of the thinking about teaching seems to have to do with shoving as much information into students minds as possible. Learning information has nothing to do with developing skills. I think how science is taught needs a big overhaul, but in terms of attitude and approach and expectations of what students learn.

It's too bad there is such a disconnect between the subject experts who are expected to teach subjects and the education experts who study how to teach effectively.

I am a chemistry major right now (going into biology for grad school), and I must say that most of the lab classes I took were pretty much useless. I don't know if it's because they were structured badly, taught badly, or whether it's just endemic to lab classes, but I didn't learn much in the way of experimental techniques (except for very, very basic things like how to use a sep funnel and how to pipette), they didn't integrate well with the lecture courses, and I had to relearn everything anyway, down to the basics, when I started doing research in an organic chemistry lab. I think part of the problem was that the way techniques were done in the undergraduate labs was very different from the way the techniques were done in research labs (probably due to cost, space, and safety constraints), but consequently it taught us very little useful knowledge about the actual science that goes on in the world.

There only good lab class I ever took was graduate lab course, which was essentially like a bunch of mini lab rotations. There were six different projects in six different research labs that was supposed to teach not only how to use different techniques in biophysics (optical microscopy and micromanipulation, magnetic tweezers, FRET, microrheometry, and so on), but also challenged us to get some preliminary results, as these projects were on things that had never been done before. I learned a lot and really enjoyed it. I think the main thing that made it succeed was that I was learning in a very small group in research labs setting, working with the grad students, post-docs, and professors who were experts in these techniques. I started to grasp both the power and the limitations of certain experimental techniques, which would inform my ability to interpret experiments from the literature. It succeeded in a way that none of my other lab classes did.

As a someone who was a laboratory TA for first-year chemistry for eight years (three years while working as a research assistant and five years during my Ph.D.) in a mid-sized Canadian university I've a couple comments about the labs at least in first year.

At our university the laboratory component was not directly linked to the lecture component of the course. You had to pass both to complete the course and they served different purposes. Much of first year lab time is spent introducing students to the tools and skills that they need to understand and use later in their academic careers. Before you can do those complex and interesting problem-solving labs in second, third and fourth year you have to learn the ropes in first year. This includes learning to handle glassware safely, to work relatively independently, to struggle through problems and to come to the laboratory prepared. Due to restraints in high-school few students have had the opportunity to get these skills before they get to University so we have to drill them in during first year. Think of first-year labs as the time you spend with your learner's license before you are allowed to drive alone. It is amazing to see the difference in confidence, skill and preparation between the first and last week in the lab course. So to all the commentors saying that labs don't teach you theory, I would argue that, in first year at least, that is not their role.

The other, equally important role of laboratories in larger schools is to provide students with mentors and help for their studies. In first year chem the prof may have 150+ students in his/her class. The lab is the one opportunity where students can get individual attention and help. I spent more of my office hours helping students with their class-work then I did helping them with their lab-work. Maybe in a small school this wouldn't be the case but in the larger schools this is one of the few places where one-on-one teaching can reasonably happen.

I once went to a talk by Carl Wieman, who claimed that there was empirical evidence showing that most lab courses are educationally ineffective. In my personal experience, labs are the single biggest reason why many students in my university decide not to continue with a physics major (often switching to math instead and restricting themselves to pure theory physics electives). Maybe those in favour of labs think it's a good thing to 'weed' these people out.

Oh, and I have the same experience as Janet: labs come in addition to traditional lectures in a single course, so I effectively have five times the workload for a typical lab-based physics course compared to a philosophy course. And while I never emerge from a philosophy course not having learnt anything, I estimate that I learnt nothing from about half of my physics courses. Perhaps it's just bad implementation on the part of my university, but I suspect bad implementation is more the norm than satisfactory or even good implementation. Which suggests that perhaps there is something inherent in the traditional system which makes it difficult to implement it well.

I've posted about this at least twice today at Uncertain Principles. It's so outrageous that I find effort is needed to stay calm. I've decided not to answer with mere anecdotes about Physics, Astronomy, Chemistry, or Biology labs I've taken and/or taught.

Steve Gimbel: "[E]verything I know about physics, I learned from my theory classes."

I have to say that it makes me wonder:

(1) How much DOES Steve Gimbel know about physics?

(2) How bad must his Lab teachers have been?

(3) Does he believe that every student has the same learning style?

(4) Does he believe that every teacher has the same teaching style?

(5) Does he appreciate how important Laboratory experience was to many of the greatest theorists? For instance, Albert Einstein's early lab time, before he was able to set that aside in favor of gedankenexperiments?

It's hard for me to address such an absurd claim by Steve Gimbel as if were a genuine proposition, and not some half-assed political ploy, or pig-ignorant rant. It is about as "provocative" as throwing a black widow spider into a facukty lounge.

As a physics TA back in the 70's, I aquired a distinctly negative impression of the value of the laboratory experience from the perspective of one whose back is to the chalk board.

To GET that TA, of course, I spent 4 years with my face to the chalk board in laboratories. And I loved it. But I was a science major, and a hands-on kind of learner who expected that it was not the principle but the care in technique, observation and recording that I was learning. God! if I needed to SEE gravity work to accept its operation I should have taken up philosophy!

Not at all the same however for 15 or 20 pre-med weenies who just want to be at the beach or anywhere but in a basement purgatory called "kinetics and electrostatics" with some geek who possesses the teaching finesse of a farmer but holds the key to that "easy A" they need for their science requirement.

The little shits broke things I had never even managed to scratch. I resorted to writing simulations of the experirments so they could effectively run them a hundred times to get a feel for the behavior of the underlying equations..which after all, were what they would be tested on.

In the end, it all worked out ok...I became a software engineer based on all the programming I had secretly enjoyed more than the teaching and probably have made a much better living because of it.

i'm gonna disagree with most folks on here.

i had a great introductory laboratory experience in general chemistry and then the upper labs were not so good. but again, it has to do with how "cook book" the lab is. the lab course was taught by someone who was DEDICATED to teaching us students how to be scientists, not just read instructions. teaching us to be comfortable in a lab using equipment, thinking up problems. and by the end of the lab we were. granted, most of us spent about 10hrs per week in the lab and another 10 outside learning to interpret our results, and understanding the science behind what we were doing. and all of this for zero credit (the lab course being an "extension" of the lecture from which we learned nothing and for which we earned four credits). takes a dedicated teacher and dedicated students. even tho it was a "general chem" course it was an advanced one that you had to take an exam to get into. it was the best lab course i ever took.

and i also had some good bio lab experiences.

in fact, my only bad undergrad lab experience was in physics.

the school i'm at now (doing my phd work) there is virtually no undergrad bio and no biochem labs--they take the tack of the kids should learn lab skills/science by doing undergrad research. well, guess what? that doesn't work any better than bad lab courses because the prof/postdoc/grad student basically has to spell out every step of what they are supposed to do in that case too.

I'm going to take a middle ground here. As a professor at a small liberal arts school, I agree with Steve that we have too many labs. Some labs are necessary. Some labs are absolutely essential. But we overdo it because it's the paradigm that we use. If everyone else is doing a huge lab, then so should I. I feel that at my school, the lab courses are so involved that it prevents the students from doing more independent projects, and it occupies so much faculty time that we cannot direct more independent student work.

But, to say that labs aren't necessary (which I realize that Steve isn't saying that) is going too far. A few well designed lab courses in a curriculum are very essential, but the difficulty is assessing the balance, both for the students and the faculty. As I said, at my school, pretty much every upper division course has a lab, and I feel that some of the reason is equity in faculty workload. If some of us don't have a lab, then either we have a lighter teaching load, or we have to teach a completely different prep.

It would be great if we could have a few, really well-designed labs in our curriculum, and then free up student and faculty time to direct more independent work. Unfortunately, at my school, I just don't see that happening any time soon due to politics.

Ah, Geology 101 lab, where one learned, inter alia, that it was a slam dunk to identify halite by taste and limestone by the application of a little dilute HCl. Of course there had to be one student on the lab exam who dropped HCl on anything and everything, no matter how un-carbonate-like, and then everyone downstream got to taste HCl on the most likely halite sample.

By biosparite (not verified) on 02 Apr 2007 #permalink