Discovering that teaching teaches

Students taught directly by the teacher understand scientific concepts signfiicantly better than students who are asked to discover the science on their own, concludes a study by David Klahr, a psychology professor at Carnegie Mellon, and Milena Nigam, a research associate at the University of Pittsburgh’s Center for Biomedical Informatics. Monitor on Psychology reports:

For decades, early science education has emphasized “discovery learning,” in which children, given experimental materials such as springs and pulleys, marbles and ramps, are expected to “discover” scientific principles on their own . . .

Klahr saw three main reasons to challenge discovery learning. First, most of what students, teachers and scientists know about science was taught, not discovered, he says. Second, teacher-centered methods (in which teachers actively teach, as opposed to observe or facilitate) for direct instruction have been very effective for procedures that are typically harder for students to discover on their own, such as algebra and computer programming. Third, he adds, only vague theory backed the predicted superiority of discovery methods — and what there is clashes with data on learning and memory. For example, discovery learning can include mixed or missing feedback, encoding errors, causal misattributions and more, which could actually cause frustration and set a learner back, says Klahr.

Seventy-seven percent of “direct instruction” third and fourth graders — but only 23 percent of “discovery learning” students — “were able to design at least three out of four experiments” without confounding factors.

About a week later, a different experimenter asked the children to evaluate two science-fair posters by suggesting how to make them “good enough to enter in a state-level science fair.” Both posters described deeply flawed experiments. Again, significantly more children exposed to direct instruction were able to critically evaluate experiments.

Critics of the research say that discovery learning used in today’s classrooms is not as unguided as the model studied. It sounds like teachers are sneaking in more direct teaching.

Via Education Gadfly.

About Joanne


  1. Discovery learning, when used correctly, can be a valuable teaching tool, but if used incorrectly or too often, it can cause devastating results, as this article shows.

    Math and science are closely paralleled here, because both require the outcome to be knowledge and understanding of distinct principles and laws, not opinions. If discovery learning is used and not monitored, the kids will just as easily come to the wrong conclusions. On the flip side, direct instruction gets rather monotonous day after day. An occasional discovery learning activity tends to break the monotony and gets the kids to refocus. If the kids are asked to “discover” what is to be taught and then the class discusses it, it is just as good as the teacher teaching the concept, sometimes even better. But discovery learning takes longer than direct instruction, so a teacher has to be careful. Too many “new curriculums” have been developed saying that discovery is the be all to end all, and that’s part of the reason why our kids are lacking in basic math and science skills today. We need to use what works, and stop trying to reinvent the wheel.

  2. I remember my lab days in high school generally consisted of me and my partner calculating the expected results, introducing a ‘random’ error, then recording the faked data while we hooked up the bunsen burner and conducted our own experiments on the melting points of various objects.

  3. Er, I suppose that should be, “my partner and I.” I’m still not sure where to put punctuations relative to quotation marks, though.

  4. Ack, scratch again – I was closer the first time. Damned edublogs – I’ve turned into a neurotic mess ever since the time Joanne pointed out I had mis-spelled “Independent”.

  5. Duh, to quote Isaac Newton “If I have seen further than others, it is by standing on the shoulders of giants.”

    Teaching is supposed to be the shoulders we stand on, giving us the knowledge to go see further. If every generation had to re-invent the wheel I don’t think we would have any cars today. Experiments in science class are important, but they are to generate interest anbd re-inforce what has already been taught.

  6. Second paragraph should read:

    Teaching is supposed to be the shoulders we stand on, giving us the knowledge to see further. If every generation had to re-invent the wheel I don’t think we would have any cars today. Experiments in science class are important, but they are to generate interest anbd re-inforce what has already been taught.

    The go was inadvertently left in after I realized I had mixed a metaphor.

  7. Doug Sundseth says:

    What is to prevent “discovery learners” from discovering phlogiston? Regardless of its current status in science, it was widely considered to be a good model by very bright people for many years.

    My point is that basic science was hard for people without a the grounding in physical principles that we now have. Why should we expect even quite bright children to be able to replicate the paths that required genius the first time?

    And for our next hymn, choir, ….

  8. Bruce Lagasse says:

    “Why should we expect even quite bright children to be able to replicate the paths that required genius the first time?”

    Exactly. Even Srinivasa Ramunajun (sp?), probably the closest thing to a true auto-didact the world will ever see, benefitted from some guidance by G.H. Hardy.

  9. Here we go again with the strawmen…

    If by “discovery learning,” you mean, no teaching, just throw the kids in a room of “stuff” and have them go at it… well, duuuhh, of course that will fail… might as well throw a million monkeys in a room with typewriters and see if they produce Shakespeare…

    As Jill said above, when discovery learning is done right, and in CONJUNCTION with traditional classroom teaching,
    it can be a highly effective tool.

    An example from my own education: I took a laboratory electronics course in college… we had lectures and exams on the theory behind basic circuit design… I could ace the homework/exams, but I really didn’t have any real understanding… my understanding was very mechanical, in that I could do it on paper, but have no intuitive feel…
    We also had a 6 hour/wk hands-on lab component… for some of it, we had guided lab exercises, teaching us how to build certain curcuits… but we also had “play” time… when we were encouraged to do whatever, try whatever, to experiment… to play with the electronics… I swear I learned WAY more during this “discovery learning” time then in any other part of the course… I would try one thing, see it fail, try something different, see it fail, see a patern develop in my head, try something new… eureka… It was a wonderful learning experience…

    Of course, this was at the college level… and obviously dealing with highly motivated students at an ivy league school…

    But I’ve taught high school and middle school kids before, and I have tried to replicate this scenario, and gotten some decent success at it… The key is, of course, that “discovery learning” can NEVER be a stand-alone method of teaching… you NEED traditional classroom teaching both before and after to both prepare and debrief the students…

    To sum up: Discovery learning can be a phenomenal tool that SUPPLEMENTS traditional classroom approaches…
    I think most proponents of discovery learning would entirely agree… It frustrates me when people make up ridiculous, extreme scenarios (as if they wanted it to fail)…
    and then say, see “discovery learning” is a failure… yeah, only in the ridiculous way it was implemented in this case.

  10. Anonymous says:

    From the article:
    “Psychologist Rich Shavelson, PhD, professor of education and (by courtesy) psychology at Stanford University, notes that totally unguided discovery of the type used in the study is rarely used in the classroom. Still, he says, “This study uses a strong research design. I’d like to see a replication with [the more typical] guided discovery.”

    The kind of “discovery learning” they tested was of the “totally unguided” variety… which is almost never used anyway. This was a strawman. As the psychologist above stated, it would be more relevant if they had used a more “honest” guided discovery approach.

  11. Sorry, that was me above.

  12. jab wrote:
    … “and then say, see “discovery learning” is a failure… ”

    Who said this? It sounds like you are defending some turf. The problem is not a “strawman”. I see this fuzzy discovery process all of the time.

    I don’t like vague discussion topics either, but you have to know that in many schools, “Discovery Learning” means a child-centered, mixed ability, mostly unguided, approach to problem solving, usually before the students have the basic knowledge and skills to solve the problem.

    Classes have used “labs” and hands-on work for ages (with very mixed results) without calling it “Discovery Learning”. “Discovery Learning” is an Ed. School term has way too much baggage associated with it to now claim that it has a very narrow definition. Do you think that people are opposed to hands-on learning or getting kids to discover knowledge? Many people (myself included) have seen that term abused so badly that they can’t help but rush to put on their hip boots. Instead of talking about discovery learning, show me the curriculum for the course, the text books used, the homework assignments, the lab assignments and example tests and then I will have a better feel for the course. After examining the breadth and depth of material covered, then we can talk about “discovery” as one of many tools the teacher can use.

  13. I’ve read articles about progressive educational theories that were written in the 1970’s, and they did push a lot of untaught discovery methods. Some of them expressed hostility at the idea that a teacher should intrude AT ALL in a child’s discovery process by actual teaching.

    I remember making a Wheatstone bridge, which measures resistance, in my college physics class. We did the theory first, then made the thing, measured several resistors, and did the calculations. It was such a blast to see that the electrons really did what they were supposed to do.

  14. Having recently finished some ed-classes on the whole “discovery vs. direct instruction” debate, I can definitely tell you the ed-schools (at least the ones using my textbook) are really pushing discovery learning. Direct instruction was mentioned, but it was lumped in with “drill and kill” and spoken of with marked condescension. Discovery learning, on the other hand, was held up as this shining method of making learning more meaningful to the students. Of course, they didn’t actually instruct us how to work as “facilitators” in this new method. I guess I was supposed to discover that for myself as well. **g**

  15. In other words, direct instruction works best — despite what we are told by so many high-priced consultants and presenters brought in by our districts to make us more effective teachers.

  16. Mike in Texas says:

    Amen TexasTeacher!!

  17. Amen to Sandra’s comment! I’m currently on my last two semesters of classes to complete an elementary/special ed. degree, and I agree wholeheartedly with Sandra as to the indoctrination of progressive techniques in the teacher education programs. While I have nothing against my professors presenting the progressive constructivist point of view, it would be nice to also see a more traditional, direct instruction point of view presented. The fact that only the party line is given in the classroom leaves me feeling woefully underprepared for when I actually will have to stand in front of a classroom full of kids and actually teach.

  18. In regards to science teaching and labwork, there is one huge factor not mentioned here: whether you have a teacher that not only understands the theory, but understands its application in the lab and has the physical skills to make it work. Fall down on any one of those three points and science lab becomes a joke.

    In addition, the teacher has to keep track of what each student is doing and give them a nudge in the right direction when needed. This requires quite small classes, say no more than 12 students per teacher. Colleges handle this by splitting a huge lecture section up into a number of lab sections, and by hiring graduate students or talented upperclassmen as assistants for the labs. (Depending on the school, the professors may leave the lab sections entirely to their assistants – and for some professors who aren’t very coordinated, this was definitely a good thing.) K-12, usually the lab and lecture sections are identical, and assistants are pretty rare.

    My high school science teachers were excellent in the lab and out of it – but that was a very good high school, as good as public education ever got in the 60’s. Before high-school though, there was just one teacher who was competent in the lab. Worse, a large part of the science teaching was done by ed. majors who had read the science books but didn’t really understand them, and even the ones that understood the theory were a disaster with the physical aspects. So through 8th grade, I learned science from reading on my own, from my Dad (who taught Biology and Geology at the community college), and sometimes by getting together with a couple of other mini-geeks and cooking up something smelly with our chemistry sets.


  1. Hube's Cube says:


    Joanne Jacobs notes a study that shows direct instruction by teachers is more effective than previously thought: Students taught directly by the teacher understand scientific concepts signfiicantly better than students who are asked to discover the sci…