Essays
The Scientific Method
|
I became discouraged over the way science was taught in elementary and high schools. So I took the opportunity to post to an online discussion forum at the University of Houston called the
Hun blog. It is no longer online but this discussion is taken from The Internet Archive
October 20, 2006 Following is a guest post from Michael T. Martin, Research Analyst, Arizona School Boards Association. Comments are included from our e-mail discussion of his original post. Though all comments appear to be from me, they are not. Our signatures are at the end of each comment. Let us know what you think! We recently exchanged emails over the issue of lead poisoning, which you finally published on your blog in its entirety. Now I would like to comment on your "new" teaching science posting, because I believe the problem is much more fundamental. I don't have quite the research about science education that I have about lead poisoning, but I believe the way schools teach science is wrong. It is my perception that the "scientific method" taught in schools does not correspond to what scientists actually do. What I read as the the scientific method involves "observations, hypothesis, prediction, testing" but it seems to me this is inaccurate and almost automatically turns students off. I don't think scientists do this, and in particular I don't think scientists have a single ritual that they perform as scientists. However, from my readings of science and biographies of scientists, I perceive an entirely different way in which science and scientists operate, which is much more conducive to explaining science to students and encouraging them to pursue science even if they don't "do" science. The traditional method, noted above, is taught as if it were a single process that is always performed by individual scientists. You can't "do" science unless you rotely follow these steps. What I perceive as happening in science is more diffuse. I see a four step process where in many cases individual scientists often specialize in one step of the process. My "scientific method" is: "discovery, taxonomy, theory, and experimentation." I see hundreds of scientists involved in pure discovery, such as oceanographers and jungle explorers looking for new species, astronomers looking for new bodies. I don't see any of these scientists using much in the way of hypothesis and prediction or experimentation. They are primarily involved in observation and recording descriptive data in reference to existing taxonomies. They may know theory but they primarily know how to look in new ways. Galileo, for example, used the telescope to better see the heavens. Jacques Cousteau used the new aqua lungs and diving bells to explore ocean depths. But they just explored, they didn't form theories or hypotheses (although Galileo realized what he saw supported Copernicus' theory). I also see hundreds of scientists involved mostly in taxonomy, such as the recent controversy over Pluto, such as analyzing newly discovered fossils, or any number of other sciences where they attempt to place new discoveries into categories according to certain criteria. The formation of taxonomies represents a crucial step in science. Linnaeus is the most obvious example. I don't dispute that discovery scientists and taxonomic scientists often overlap, but they are distinct processes in the scientific method. Knowing the classifications of something is one of the reasons we go to college to learn science. If you study ornithology, you learn the classifications of birds, and if you discover a bird the first thing that enters your mind is "what classification of bird does this belong in?" And the same thing is true of astronomy and subatomics and any field of learning. Fundamentally, what transformed Galileo's discovery of the moons of Jupiter was the realization that the bodies seen near Jupiter did not fit the classification system of Ptolemy. If they had, it would have been of no great importance. I also see hundreds of theoretical scientists who may never do discovery nor taxonomy nor experimentation. Theoretical scientists, such as Murray Gell-Mann, and those involved in string theory, or even Albert Einstein, primarily attempt to find continuous models to explain the discrete phenomena in the taxonomies. These are the scientists who most use mathematics, but Murray Gell-Mann's quark theory was largely arithmetic, and some just built computer models using genetic algorithms. But theoretical scientists play a crucial role in science because essentially they are the ones who develop theory. Theory represents a continuous procedure for classification. It is called an 'explanation' even THOUGH Quantum Mechanics provides a theory to explain classification that nobody really understands. Theory does nothing except provide a procedure for classification into a taxonomy. In the process, it will frequently suggest new classifications that may not be known at the time. Finally, I see hundreds of scientists involved in the experimental testing of ideas developed from theory. These scientists may be mostly engineers and are involved in the more mechanical aspects of investigation. They build large and complicated machines to test things like gravity waves and Higgs Bosons and the mass of subatomic particles. These are the "rocket scientists" who may never do much in the way of discovery or taxonomy or theory building. They primarily devise clever ways of expressing the consequences of some theory in a hypothesis that can be tested. Albert Einstein may or may not have realized that his theory of relativity would bend the positions of stars, but experimental scientists devised the idea of looking during a solar eclipse to test this new classification (which then became gravitational lensing). I have argued that you cannot create a scientific hypothesis without a theory. The traditional explanation of the scientific method almost sounds like you see some phenomena and then make some uninformed guess and call it a hypothesis. But I don't see scientists doing that. What I see scientists doing is creating a theory that explains existing discrete taxonomic groups from which testable hypotheses may be derived for predicting other discrete groups. I think this is a crucial concept, particularly in the Evolution/Creationist debate. I think the main reason students can be confused by this debate is because we present a theory as some conjecture. In fact, I see a theory as accounting for facts, you cannot have a theory unless there are facts which have been placed in a taxonomy. Young people seem surprised when I point out to them that even if the theory of evolution is totally disproved, the taxonomic facts of evolution will remain for a new theory to explain. In other words, the neophyte thinks you can disprove facts by disproving theory, because we don't teach students that a theory is derived from a taxonomy of facts. Indeed, it is my contention that we almost entirely leave out the process of taxonomy when we teach science, but I don't think any science can get far without a taxonomy. And without the taxonomy we forget about discovery scientists who do all of the exploration and discovery of new things that challenge our existing taxonomies. More importantly, what we actually show kids as science is almost always the discovery scientists who explore the oceans like Jacques Cousteau, or dig dinosaur bones in the Gobi Desert. These are the scientists they see, and they don't see them sitting around creating hypotheses and making predictions and testing them. They see a disconnect between what we say is science and what they see as science. When you say the PROCESS of science, I see primarily descriptive data recording and the creation of taxonomies from that. What I see scientists doing is noticing (observing) something different and taking notice of that and trying to see how that fits into the existing taxonomies. For example, you begin a scientific paper by reviewing the literature and then stating the difference between the existing phenomena and existing explanatary taxonomies. If you have truly new phenomena that do not fit existing taxonomies, then you almost certainly begin by creating a taxonomy, such as the subatomic "zoo" that catalogued different kinds of new "particles." Or perhaps the discovery of "vitamins" categorized into A and B and then other "vitamins." Or the Messier catalog with stars, planets, and "nebula." These are the crucial processes of science that I don't see people really teaching kids about because we always rush into asking them to form some "hypothesis." More importantly I don't see us teaching them the difference between discrete phenomena that we observe, and the concept of a theory which involves a continuous conception which we can use to predict other discrete phenomena that should be there but we don't currently have in our taxonomy. I don't think we teach students why a theory can predict things. Perhaps more importantly, we tend to tell kids that they have to have algebra and calculus and beyond to be scientists. Like it is some torturous hazing to belong. But you don't need those to be scientists. You can be a discovery scientist without knowing much more than arithmetic. You can work in taxonomy without deep math. You can even work in theory and experimentation without deep math. But math is a handy tool in those fields. I think if you interest students in science, and teach them the process of discovery, taxonomy, theory formation, and experimentation, they will soon come to appreciate the advantage of knowing math. And having a need for it, they will learn it much more avidly. Take any field of science and think about it for a while in terms of how that field functions. Take drug research, for example. There are numerous discovery scientists who explore different chemicals or toxins or bacteria and viruses. They are just trying to find new things, like prions, or new elements such as proteins or epigenetic processes. They are just trying to find what is out there. When they find something different they try to classify it, to see how it relates to others of that "type" and thus fit it in a taxonomy. You can find all kinds of new things that fit into existing taxonomic classifications. That generally will not affect the theory of the field at all. Occasionally you find something that doesn't seem to fit the taxonomy and thus may not be accounted for by the standard theory of that field. Thus someone has to modify the theory, or develop a new one. But any change in a theory will generally create new permutations that suggest something new might be located. The experimental scientists then try to figure out how to test for things in that new location. I believe this makes a more coherent explanation of how science functions that students can reconcile with their own knowledge. When you define a "hypothesis" as an implication from a theory that needs to be tested, it positions it on the opposite side of traditional teaching, that implies a hypothesis precedes a theory. But it takes theory away from being a conjecture into making it an explanation of an existing taxonomy. Astronomers first observed Novas, and then realized there were differences that created a new category of supernova, and then they observed differences that could be categorized as Type I and Type II and then they developed Type Ia and Type Ib, and each one required people to reconsider the theory that formerly was just the main sequence of stars and then became more involved. But it was the taxonomy change that required the theory change. And then the theory began to imply that there were "black holes" which started people testing to find them. People did not see some stars and then hypothesize there were "black holes" because you cannot create a hypothesis from an observation as the traditional scientific method avers. I think the average student recognizes that implicitly and thus has great difficulty accepting the teaching of science. But if they understood the concept of a taxonomy and how a theory is merely an explanation of a taxonomy, it all hangs together as a rational means of viewing the universe. I realize this is wild esoterica, but I would be interested in hearing how someone who teaches kids science would react.
Michael T. Martin Comments [by blog owner] 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. Posted by: Brad Hoge | May 10, 2007 at 02:21 PM [Response by Martin] 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 [Response by blog owner] [Response by Martin] 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 been 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. IMHO
Michael T. Martin
|