« Marble Steps | Main | A Gedanken Experiment for Teachers »

October 20, 2006


Michael T. Martin

I disagree that science has "the single common factor of controlled experimentation." I see experimentation as a specific phase of science, but you may be referring to "procedures" that typically are taught as "experiments" in schools. When I think of an "experiment" it refers to a particular phase of science that involves deriving an implication from a theory that one can test. If your computer doesn't turn on it could be due to numerous problems with the computer, but someone might have a "theory" that the fuse has blown on the wall socket which you can test by the simple experiment of plugging a fan into the wall socket to see if it works. A theory is always an attempt to explain known observations (the computer doesn't light up) but each theory has implications for other phenomena which you can test with an experiment. But you can do science without experimentation.

Consider, for example, a comet hunter, or the discovery of a planet, that is done primarily with "blink" equipment that projects two identical views of space to see if something in that view changes. This is not an "experiment" it is just a procedure, done repetitively, to see if something can be found, but a scientist does it. Or take an archeologist who digs out artifacts from an ancient trash heap. There are procedures for doing this "scientifically" (meaning primarily to preserve data) but there is no experiment. You could be an astronomer and never run an experiment but still do science.

What I see as the "single common factor" of science is the recording of data, and particularly measureable data. That recording of data is
procedural: you must follow particular procedures to make measurements (the size of a horse depends entirely on whether you measure to the top of the head, the top of the haunches, the top of the withers, or from the tip of the tail to the tip of the nose). So when we teach "science" we frequently teach standard procedures for making measurements and mistakenly couch these procedural practices as "experiments" when they are really just procedures. When you actually do experiments you often have to make measurements, so you employ those procedures, but just following a procedure to obtain a measurement is not an "experiment" because we generally are pretty confident of the outcome. When you go to the doctor they put you on a scale to weigh you and find your height, but I don't think they call that an experiment. What they are doing is gathering data.

I think it is crucial that students understand that all science derives from the procedural recording of replicable data, but that the next step must always be to categorize your data. Even with something as simple as your doctor taking your weight and height, it is for the purpose of categorizing you as obese or emaciated or normal. This categorization is what I just don't see being taught in typical science classes. Instead, most science classes start with the procedures to produce certain categories of data and then record the data without even mentioning the categorization. But in real science, the discovery phase involves finding things and recording the descriptive data which can include color, size, density, weight, translucence, and other characteristics because initially you don't know what characteristics are crucial for categorization. Often determining the means of categorization is a crucial development in science. Once discrete data has be categorized it is natural to ask, why or how these categories exist. Then you enter the realm of the theorist. The theorist tries to develop a continuous function that will explain the discrete data in the categories. Thus the categories of solid, liquid and gas can be explained by a theory of kinetic energy which might include sublimation as well. Once you have a theory, then because it is continuous it will have implications for discrete consequences that you may not be aware of, and which you can devise an experiment for.

But the importance of science is that you collect replicable data through standardized procedures, categorize this data according to discrete characteristics, and then seek a logical explanation for this categorization as a theory, which then has implications for other discrete characteristics that can be tested. Which is not to imply that you cannot have data other than a measurement. Sightings are not measurements, but they must be replicable, or presumed replicable from evidence (the coelacanth was initially a single fish which implied it was replicable, and incidentally was primarily significant for taxonomic reasons). Typically data is first qualitative until quantitative procedures are derived (measurement requires a "unit of measure" to be valid, which is why I don't consider academic tests as valid measures).

You mentioned that there are unguided wasteful lessons by some teachers, and one incident that always pops into my mind was visiting a school science fair and seeing a posterboard display about dinosaurs that stated "Tyrannosaurus Rex ate meat and therefore had sharp teeth." I remarked to the person in charge of the fair that this was totally non-scientific. We don't know what T.Rex ate, we only know they had sharp teeth. Thus the scientific method is to first record facts (sharp
teeth) and then to categorize these facts (T.Rex was a sharp toothed
animal) and then to devise a theory that explains the taxonomy (dentition implies diet: animals with sharp teeth typically are carnivores). But the crucial point is that whether a T.Rex ate meat or not has no educational value: the educational value is purely from the logical thinking of facts to category to theory. From this theory about dinosaur diet it may be difficult to devise an experiment to test the theory, but one that was actually done was to consider that typically there is a predator to prey ratio and therefore the relative abundance of T.Rex skeletons in comparison to herbivore skeletons would provide a test. And in this case, the "experiment" was purely statistical.

Constructivism is especially suited to teaching science because the entire point of science is devising a logical sequence to account for discrete phenomena. Thus what we want the student to learn is not really the discrete phenomena (such as dinosaur teeth or diet) but rather the logical process of accounting for those phenomena. I have argued, for example, that the theory of evolution has little educational or scientific value, what has all the value is the logical process of selectionism. We teach evolution in order for students to see an obvious example of selection at work, but selection works in many disparate sciences, including the science of cognitive development (google Esther Thelen or Gerald Edelman). When we teach science it is the process of reasoning based on facts to categories to theory to experiment that is important for students to learn. Constructivism is essentially a method of having students "learn" by reasoning out from their existing knowledge. Science is essentially a method of having people learn by reasoning from existing data to taxonomy to theory and experimentation. It is a kind of hand in glove thing.


Michael T. Martin
Research Analyst
Arizona School Boards Association

Brad Hoge

I think you've hit on an extremely important issue that does need to be re-examined. If we say that constructivism is the best way to teach science because that is how science is done, then we'd better be clear what we mean. I've never liked teaching scientific method as a standardized procedure. Science is done in many ways, with the single common factor of controlled experimentation, whether that is in a laboratory or the field or through mathematics. Emphasizing the key features of scientific method is beneficial, but we can easily miss teaching much about how science works by not being thoughtful in our inquiry. Constructivism is effective because it models scientific problem solving, but teachers have to know what they are doing. If they don't, then hands-on science becomes the unguided wasteful lessons our critics are so fond of pointing to in order to indict the whole philosophy.

Michael T. Martin

The point I was making was not really constructivist/traditional, which is a methodological concern, but rather that what we actually teach is incorrect. It is true that if you teach something that is wrong, then students will be unable to construct an understanding of it, but my concern is that when we teach science we simply have the wrong model.

I was looking at your 10/4 posting and the responses were interesting. First, someone noted that you cannot have a theory without facts, which I don't believe we put enough emphasis on. But more importantly, the critic who talked about his fifth grade kid mentioned categorizing leaves, when my main point is that our existing explanation of the "scientific method" simply ignores the process of categorization (which I call taxonomy).

However, I am a constructivist, and perhaps in explaining it to others you may want to adopt my view of why they don't understand. In my opinion, these people condemn constructivism because they have the traditional paradigm and cannot comprehend the difference is more fundamental. The issue in constructivism is not "the guide at the side" telling the student where to go, which these people interpret it to be, as this is simply "the sage on the stage" who has stepped off the stage. The purpose of the guide is to ask, not to tell. The guide at the side asks the student what is already known and how that might apply to the present circumstance. Thus it attempts to build a logical connection between what the student knows already and what is being learned.

The issue is that students do not need so much to be taught subjects, but rather to be taught that subjects are logical. The purpose of education should primarily be to teach students how to reason. The focus must always be on the logical process of developing an answer to a problem, not the problem nor the answer. We teach science primarily to learn how to think scientifically, how to reason in the ways that scientists reason, based on logic and parsimony. The same is true in other subjects such as history, where the traditional names and dates approach is worthless.

The issue in the history of the American Revolution is not that something happened in 1770, and then something else happened in 1775, and that something further happened in 1776. But rather that something happened in 1770 that created a logical circumstance which led to a sequence of events that logically precipitated people to consider a change in circumstances that evolved through rational debate into the Declaration of Independence. Students need to know the rationale and logic of the events rather than thinking history is some surreal sequence of events.

Once students know how to think, they can derive facts and concepts. In many cases, knowing how to think will totally change how we look at facts and concepts. The perfect example of that is paleontology which was initially a dead science of categorizing fossils until the "young turks" looked at the fossils and asked "what does this tell us?" They transformed dinosaurs from cold-blooded reptiles into warm-blooded ancestors of birds. Sure they could have continued to learn the same misinformation of dinosaurs being cold-blooded reptiles, but teaching them to think caused them to ask about these fossil bones would have logically functioned in an animal and then it became clear that the fossils did not logically fit into how reptiles functioned. They, in essence, "constructed" dinosaurs in a new way using constructivist thinking.

But, as I said, I am more interested in dissecting the scientific method that we currently teach.

Michael T. Martin
Research Analyst
Arizona School Boards Association

Brad Hoge

This is hardly wild esoterica, it is in fact constructivism which is how I teach science and hope to have my teachers teach it in their classrooms. If you've read my blog for awhile, you must realize that I agree with just about everything you said. Constructivist pedagogy, done right, accomplishes what you see as needed. Unfortunately it is rarely done well, and is currently under attack from NCLB adherents. If you want to know more about constructivism and how best to teach science, check out the links here. Particularly, How People Learn, BIE and the Inquiry Page.

Thanks for you passionate correspondence. I'd love to continue the conversation.

The comments to this entry are closed.

My Photo
DAILY CARTOON click to enlarge

June 2016

Sun Mon Tue Wed Thu Fri Sat
      1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30    
Blog powered by Typepad

Powered by FeedBurner

Bookmark and Share