J. Jose Bonner's Outreach Pages
DEPARTMENT OF BIOLOGY
The biggest gap that adults have in their scientific knowledge is not that they've forgotten the details of DNA; it's rather that they don't know what science is about. Understanding the nature of science is even more important than mastering its details.
-- Alan Leshner, AAAS
Educational Leadership, Dec. 2006
Details on the role of...
Why Understanding the Nature of Science Matters
...in food choices
...in the evolution battles
Teaching the Nature of Science
A simple statement of the Nature of Science is presented in Constructing Science in Elementary Classrooms [Lederman, NG, Lederman, JS, Bell, RL. (2004); Pearson Education, Inc. Boston]. Lederman et al. offer the first four of the following five statements. We have added #5.
The Nature of Science
1. Scientific knowledge is partially the product of human creative imagination.
2. Scientific knowledge is tentative.
3. Scientific knowledge is partially a function of human subjectivity.
4. Scientific knowledge necessarily involves a combination of observation and inference.
5. Scientific knowledge is grounded in evidence.
A more extensive statement can be found at the National Science Teachers' Association.
What do these statements mean, why are they true of science, and what is their impact on the reliability of scientific knowledge? Why does Dr. Leshner consider this to be such a serious issue?
Because so much of the natural world is too small, too large, too fast, or too slow to be detected by unaided human senses, it is essential to visualize natural processes in the mind. This requires Imagination. It also requires imagination to recognize links between different pieces of information. Many of the greatest scientists, who are responsible for fundamental leaps of understanding, saw relationships that others did not.
It is important not to mistake "imagination" for "making things up." Here, we use "imagination" in its scientific sense -- using the mind to create images. This is the same skill that one uses when visualizing the action while reading a story. We do not use "imagination" in its Conversational English sense of seeing things that do not exist (as is sometimes suggested by detractors of science).
In short, it is essential for students to visualize the processes they are studying.
The goal of science is to understand how the world works. Scientists ask questions about the world -- questions to which they do not already know the answer. To ask these questions, scientists make observations, do experiments that provide more observations, and then use logic and reasoning to explain those observations. The explanation is, in essence, an "answer" to a scientific question. But is it the right answer? How can we tell if it's right?
Science uses an operational procedure for determining the "correctness" of its explanations. When we first develop an explanation, we call it a model, or hypothesis. These terms remind us that the model is tentative, unproven. The model/hypothesis is tested experimentally by gathering more information, more observations, more data. If new data contradict the model, then the model must be modified or even discarded and a new one developed. If a model is tested many, many times and continues to withstand the tests, then it looks like the model might be close to correct. But, we still can't know it is correct, because there is always a possibility, however remote, that new data may come to light and require that we modify the model yet again. To distinguish a new model from a well-tested model, and to distinguish model from fact, we refer to a new model as an hypothesis, and a well-tested model as a theory.
Unfortunately, Conversational English uses "theory" to mean "guess," so it is very easy to misinterpret the strength of scientific information. It may be best to summarize it this way:
Scientific knowledge is reliable in that it is based on reproducible evidence. Scientific theories have been tested many, many times and have been shown to have great explanatory power.
Scientific knowledge is tentative in that future observations or experiments may indicate that we may need to revise our understanding. Not surprisingly, highly-reliable theories approach the possibility of being true, but we nonetheless refer to them as theory rather than fact, just in case. Hypotheses that have been tested less often, usually because they are new ideas based on recent data, are more likely to be revised as new observations come to light.
Scientists try very hard to be objective about their work. The process of peer-review has been developed to strengthen objectivity (i.e. scientists submit reports for publication, and those reports are sent to several other scientists who review the data and the interpretations critically. These reviews can be very critical, since reviewers are often competitors. This fact is under-appreciated by non-scientists, who sometimes suggest that peer-review exists to exclude opposing viewpoints; it actually exists to minimize subjectivity.) The "rule" of building explanations based on the data, without incorporating concepts for which there is no evidence, also helps to maintain objectivity. The fact remains, however, that every scientist lives in a culture and a time, and holds certain views. Experimental data are inevitably seen in the light of existing models, which themselves form a kind of "societal norm" within the scientific community.
Nonetheless, science is a global endeavor, involving people from nearly every country, every race and ethnicity, both male and female. The diversity of participants helps identify culture-specific subjectivity. Scientific progress itself helps identify other forms of subjectivity, as new data come to light and old models are replaced with new ones. Although a certain degree of subjectivity is inescapable, the effects of subjectivity become minimized with time.
In the short list above, items 4 and 5 are similar: scientific information is grounded in evidence, and unavoidably involves a combination of observation and inference. These are a shorthand summation of the following process:
Observe something interesting
Think about it, develop a tentative explanation
Develop a method for investigating further -- experiment, additional observation, etc
This provides new, additional observations (new data, new evidence, etc)
Think about the data/evidence/observations and refine the explanation
* science is a process of obtaining data, then developing inferences from the data
* the data, observations, and evidence are facts
* the Scientific Knowledge developed from the observations is inference
The common misconception of science -- to which Alan Leshner refers in the quote above -- is that scientific knowledge is Facts. It is not, and it cannot be. Scientific knowledge is the inferences that scientists draw from the data; it is the models for how things work. [For example, no one has ever watched the earth orbit around the sun. We infer that it does so because this is the only explanation that fits all of the available data.]
What we have said above should, we hope, make it clear why scientific knowledge is tentative: it is our current best explanation for the currently available data, and is subject to revision if new data become available.
The discussion should also make it clear why scientific explanations are naturalistic explanations, and do not invoke supernatural causes. The data are observations. They are verifiable facts from the natural world. Explanations that explain those observations must be consistent with other observations of the natural world, and -- to maintain objectivity -- cannot go beyond the data that are available. For any scientific explanation, we must always ask, "what are the data that lead to that conclusion?"
There are, however, branches of philosophy that challenge science by asking, "what if reality does not actually exist?" If there were no objective reality, then "science" would have no objective basis, and scientific knowledge would be tentative because it is the subjective imagination of scientists. This philosophy shows up in science-education textbooks, which may define "constructivist teaching" not as students "constructing their own knowledge," but "constructing their own reality." Each of us must construct our own knowledge, as we think about observations and fit them into our understanding. But science is based on the ability of each of us to make the same observations on the same aspects of the natural world -- the reality of that world -- which we all work together to describe and understand.
The idea that students must construct their own reality as they go through school leads to the idea that scientific knowledge is tentative because scientists make personal guesses at what the world is like. This is quite different from scientists working together from common, shared observations and data.
Therefore, the most reasonable explanation for the tentative nature of scientific knowledge is that most of this knowledge is in the form of working models derived from the study of factual data.
How often do we hear about people receiving a "wrong diagnosis"? It goes both ways -- there's the cancer patient who was mis-diagnosed initially, and might have had a better chance of survival had the first diagnosis been correct; and there's the aneurism patient who is rushed to surgery, and the surgeons discover during the operation that there was no aneurism after all. Shouldn't the doctors have gotten it right the first time?
Making a diagnosis is an exercise in scientific thinking. The doctor is presented with some observations, and from these, must try to develop a working model. Just like any other scientific question, the answer is not written down somewhere; it requires problem-solving skills, inference, and imagination to put the facts together and arrive at an explanation. Like all scientific explanations, a medical diagnosis is tentative, and may need to be revised as new information comes to light.
Fortunately, most human diseases have been seen before, and most individuals present similar symptoms. But there is always variation, so sometimes a diagnosis turns out not to be correct.
Sadly, US culture has evolved toward expecting doctors to be Right all of the time. People seem to think that diagnoses, and science in general, is simply a body of facts, and that mis-diagnoses are like getting the wrong answer on some kind of multiple-choice test. We need to understand the nature of diagnosis, and the Nature of Science -- and, perhaps, even learn a bit about scientific and medical problem-solving.
You want to lose weight. Should you cut back on fat? There are a great many fat-free foods available. Are they what you want? Or, should you cut back on carbohydrates? Low-carb diets are said to work. Or should you buy some of those fat-burning pills that contain ephedra? How do you decide?
Unfortunately, most Americans seem to decide on the basis of what a friend-of-a-friend said, or on the basis of a testimonial from a total stranger. Few pay attention to the science. Partly, this is historical. Consider the example of fats. As we have learned more about the class of chemicals called "fats," we have been forced to revise our understanding. Nutritionists have tried to make the public aware of the latest information, but with the frequent revisions, many people have given up. First, we were told not to eat butter, but to eat margarine. Then we were told butter was OK. Then we were told margarine was bad, unless it was in a tub and not in a stick. Then we were told it's not the margarine, but something called trans-fat. When are those scientists going to get it right?
If we fail to understand the Nature of Science, then we may fall into the trap of thinking that a nutritional recommendation is a statement of Scientific Fact. It's not. It's a recommendation, based on our current understanding. Our current understanding is merely the current interpretation of the currently-available data, and is subject to change.
Of course, the data themselves remain valid; the data are the facts. If we can keep the Nature of Science in mind, then we can look at the nutritional recommendations in the light of the data that are available. In this case, the same "butter series" would look like this:
What We Should Conclude
Saturated fat correlates with heart disease
Don't eat butter; instead, eat margarine (which had just been invented)
Eating lots of butter may raise our risk of heart disease, but eating margarine should not
Cholesterol correlates with heart disease even more strongly than saturated fat
It's OK to eat butter, just don't overdo it
Eating lots of butter may raise our risk of heart disease, but eating a moderate amount is probably fine; margarine should also be OK.
Margarine that has been hardened so that it can be formed into a stick correlates with heart disease, but margarine in a tub does not because it is emulsified rather than hydrogenated.
Avoid stick margarine
Eating lots of butter may raise our risk of heart disease, but eating a moderate amount is probably fine. We can use margarine if we want, but use the tub varieties.
The process of hardening margarine is "hydrogenation" of unsaturated fats; a side-reaction of this is to convert cis-fats into trans-fats. Trans-fats correlate very strongly with heart disease, and are not natural products besides.
The reason to avoid stick margarine is the trans-fats, which should be avoided in all products
Eating lots of butter may raise our risk of heart disease, but eating a moderate amount is probably fine. We can use margarine if we want, but make sure that it does not contain "partially hydrogenated vegetable oil;" also make sure to avoid anything else with partially hydrogenated vegetable oil. Simplifying this seems to suggest that we should use olive oil or other liquid plant oils for most things, but use modest amounts of butter if needed.
In this "butter series," the recommendations seem to flip flop. But, in the context of the facts (the data), our conclusions simply become more complete. That is, if we focus on the recommendations, we get confused; if we keep track of the data, we develop deeper and deeper understanding.
Global warming is a good example here. The science shows remarkable consensus that atmospheric CO2 needs to be limited, and that environmental change is under way. But, because there are always alternative explanations at the forefront of scientific knowledge, it is always possible to find reports that question some of the newer consensus. Politicians who do not want to act on measures that might be unpopular with their constituencies often put off action, arguing that "we should wait until the science is clear -- until the conclusions are no longer tentative."
If we look back through history, we will probably find that nearly every US administration has used this argument at one time or another. It is a political tool. In our opinion, the real problem is that the general public seems not to notice. Scientific knowledge is always tentative. However, it is the best we have at the time, and should be taken into account in legislating public policy.
While we can argue that the Evolution Battles are, at heart, a conflict between a religious way of knowing and a scientific way of knowing, we suggest that the conflict also grows out of misunderstanding the Nature of Science. Put simply, if a child goes to school and her teacher says "this is how it is," and then goes to church and the minister says "this is how it is," then the child is put into conflict. Who is right? We are guaranteed to set up this conflict every time we present the conclusions of science as if they are fact.
However, if we present science as a process of reasoning from data, and present students with the data rather than with the conclusions, then students have the opportunity to work through the reasoning themselves. They may well conclude that evolutionary theory makes sense as an explanation of the data, or they prefer their religious viewpoint instead. Either way, they will see that the data are the facts, and that the evolutionary model is one explanation of those facts. They are free to develop alternate explanations, with the constraint that a scientific explanation must be consistent with all of the available data, and not invoke processes for which there are no data.
In our experience, using this approach in the classroom lessens the controversy significantly. Science is about interpreting the data, not memorizing facts.
The most fundamental aspect of science is that it develops explanations based on data. Thus, the most fundamental aspect of the Nature of Science is that the knowledge of science is tentative, because future data may change our understanding. How are these two Core Issues dealt with at different levels of education?
Our analysis of this issue is based on the historical trends in education, and the recognition that the Educational Enterprise has always tried very hard to do what is best. But, like science itself, our understanding changes as more information comes to light.
At one time, the tradition was for students to stand and recite their lessons. Learning was thought to be reflected in what one could memorize. Later, after the Soviets' launch of Sputnik, science education took on a particular urgency, in which there was a major push to "bring students up to speed" in math and science, so that they could become scientists and engineers and ensure US leadership in science, engineering and technology. To "bring students up to speed," of course, requires presenting them with as much current knowledge as possible, so that they can extend this knowledge after they graduate and become scientists. This put an emphasis on The Facts, and required only a modest change in methodology from the reciting of lessons. To a large extent, it worked. The US became pre-eminent in science and technology. Even so, the large majority of students did not learn math and science particularly well. Now, at the beginning of the 21st century, we have thought to look at the scientific learning of the majority of students, rather than the small percentage who become scientists -- and we find that the cram-them-full-of-knowledge model is not very effective. Learning, and particularly scientific reasoning, may not be so easy to measure simply by reciting lessons (or its mass-delivery equivalent, the multiple-choice test).
In the following discussion, we explore the current traditions and consider their impact on students' understanding of the Nature of Science. To the extent that we are critical of teaching at other levels than our own, we are equally critical of our own teaching.
Although there are certainly exceptions, a study of Elementary science lesson plans reveals that the majority of them follow a common pattern: an introductory exploration, entry of information into a Science Notebook, and an explanation by the teacher of what the students learned. Whether we use a Learning Cycle model or an Inquiry model of teaching, we generally do not engage students in reasoning from the data. This seems to stem from a mix of Tradition and careful, research-based planning. Tradition is mentioned above: it emphasizes factual knowledge over scientific reasoning. The research basis is summarized elegantly in Taking Science to School. Much of the current tradition in Elementary science teaching is based on Piagetan analyses, which suggested that young students are not capable of complex patterns of thought -- that they are restricted to linear, concrete thinking. According to this understanding, science teaching should focus on demonstrating scientific phenomena, and on presenting the basic facts, but not on making inferences from observations.
Taking Science to School updates our understanding. As with other scientific theories, the theory upon which we have based Elementary science teaching is subject to modification as new data come to light. Recent data indicate that, contrary to prior thinking, young students are capable of, and engage in quite sophisticated reasoning. Even preschool children can distinguish knowledge that they were told from that which they have inferred from their own observations. There is no longer a compelling case for excluding from the Elementary curriculum the type of scientific thinking that is required for students to make sense of data. We argue that there is a compelling case for putting in such reasoning.
Whatever future teaching strategies should be [and here, we need additional research to test alternative models], the fact remains that current methods tend not to provide Elementary students with very much experience in the fundamental core of scientific reasoning: developing explanations for observations/data. Current methods tend to allow students to infer that Scientific Facts are presented by the teacher.
There are occasional lessons that engage students in complete inquiries, in which they do, indeed, wrestle with data and develop understanding. However, these are often oriented around scientific questions that are divorced from the Big Questions of science. For example, they may address the question, "which paper towel is best?" rather than "how do living things acquire and use energy?" Apparently, it is difficult for students (and teachers) to internalize the Nature of Science fully when "science" is taught in some lessons, and "the Nature of Science" is taught in different lessons. The Nature of Science becomes merely additional Facts to learn, but facts that are often at odds with the other facts of science.
There are also "inquiry-based lesson plans." These more-closely approach the process of scientific investigation. But, many suffer from a lack of direction, such that students primarily explore materials and record their thoughts, and then are done, with no further discussion. To the extent that science learning is thought to be a process of constructing one's own reality, this makes a kind of sense. If each student's reality is different from every other student's reality, and from the teacher's reality, then it would be most improper for the teacher to impose her reality on the students. By this logic, students should be allowed to infer whatever they want from their observations.
But science is the examination of the single, natural world in which we all live. Learning science requires consideration of observations, data, and facts that are available to all. These are part of a single, common reality, which science seeks to understand. Therefore:
- Inquiry-based lessons should not stop after data-collection. They should not leave students on their own to interpret their findings. They should involve students in a common discussion of the findings, with the teacher guiding the discussion. The discussion should illuminate the best-available explanation for the findings.
Secondary science teaching tends to follow a different paradigm than Elementary science teaching. Most commonly, the focus is on transmitting fairly large bodies of information to students. The information-transmission focus is emphasized, and codified, by High-Stakes Testing that can assess factual recall much more easily than it can assess higher-level reasoning. In this information-transmission teaching model, laboratories, or "hands-on experiences," tend to be used as confirmation of the basic scientific principles.
A typical lesson sequence begins with a discussion of the basic information and a reminder of the terminology, to ensure that students "have the background they need" to understand the lab. Then, the laboratory exercise follows. In general, students know what answer they are "supposed to get" in the lab, and also know that the teacher already knows this answer (despite attempts by the teacher to provide an atmosphere of discovery). In the end, the lab serves to confirm what the teacher had said previously. The data support this explanation, unless, of course, they are "wrong."
This approach tends to emphasize the conclusions of science, and minimize the role of data. Students are able to form the misconception that "science" is a collection of facts. They are able to form the misconception that the role of data is to support conclusions that the scientist has already developed, rather than to develop conclusions initially. That is: if one has a pet theory (say, evolution), then one looks around for some data that seem to support that pet theory, and then voila, the theory stands. Needless to say, this is backwards: in science, one collects data, then tries to figure out how to explain it. There are usually several alternative explanations that must be evaluated before one can reach a (tentative) conclusion.
Again, it is traditional to offer several lessons that directly address the Nature of Science. Again, these tend to be separate from the lessons that address the science itself. Again, these separate lessons tend not to dispel the conceptions that students form over months of science teaching.
To a large extent, traditional college/university science teaching parallels secondary science teaching. Typically, background is presented in large lectures; the background is confirmed by laboratories. There are notable exceptions to this trend, but the tradition remains remarkably common. It is particularly common among non-majors courses, which are often populated by students who were unenthusiastic about science in high school, and who might be most in need of alternative teaching strategies. Distinct from secondary science teaching, however, is a greater effort to describe some of the more significant experiments, thereby to illustrate to students how we know what we think we know.
It is significant that college/university science teaching, like primary and secondary, tends to describe the Nature of Science separately from the science itself. It is also significant that students who are destined to become primary and secondary teachers take their science courses from science departments, separately from the teaching courses that they have in departments or schools of Education. The Nature of Science is likely to be discussed explicitly in Education courses, but may not be discussed in science courses. Thus, in the college/university setting, pre-service primary and secondary teachers are likely to experience a reinforcement of the conceptions about the Nature of Science that they formed in prior grades.
We mention above that college/university science teaching parallels secondary science teaching, because we want to emphasize the fact that people tend to teach the way they were taught. If college/university faculty would like to see students entering their classes with a deeper understanding of the Nature of Science, and with experience in scientific thinking, then they must model the teaching styles that develop this depth of learning.
There is an extensive literature on teachers' beliefs about the Nature of Science. An unjustly brief summary is that little has changed in decades, if not over the last century (1907-2006 to be precise). One thread in the Nature of Science literature is the question of whether it is better to present Nature of Science concepts through explicit teaching, or through embedding them in the science lessons ("implicit" teaching of these concepts). In general, explicit teaching has been more effective. The implicit approach can be nearly as effective if the instructor calls attention to the aspects of the Nature of Science that each lesson illustrates. And yet, the research indicates that neither approach seems to lead to lasting understanding of the Nature of Science, or to transmission of understanding to students.
In the light of the discussion above, it is fair to ask the following question: were these studies performed using traditional teaching methods? If implicit instruction was embedded in science lessons that do not illustrate the nature of science because they do not involve significant reasoning from data, then we would not expect an understanding of the Nature of Science to emerge. After all, the fundamental reason that scientific knowledge is tentative is that scientific knowledge is a collection of inferences from data. Explicit instruction should work better, but if explicit instruction is an add-on, distinct from the teaching of science content, then we might expect that learning about the Nature of Science would be transient: enough learning to pass the next test, but not enough to remember it six months later.
How do scientists themselves learn about the Nature of Science? Oddly, many practicing scientists do not recognize the phrase "The Nature of Science." They do science, and they know that its conclusions are tentative because conclusions are interpretations of data. They understand the Nature of Science because these concepts are an integral part of their work, even if they do not speak about the Nature of Science explicitly. One exception may be evolutionary biologists, who often do use this term. Because evolution appears to conflict with some religious beliefs, evolutionary biologists often initiate teaching sessions with a discussion of the Nature of Science. The conflict between science and religion is lessened when one emphasizes the Nature of Science, and points out that evolutionary theory is an explanation of data, rather than Absolute Truth handed down by an Infallible Source.
If scientists in general understand the Nature of Science, but do not use the phrase and have not been explicitly instructed in it, how do they know? They learn through doing real science. This is a form of implicit instruction, but it differs from implicit instruction used in the classroom in that it is built around authentic scientific reasoning. Scientists do not do activities to confirm that facts are true. They do investigations to gather evidence, then reason from the evidence to reach conclusions. Because many investigations have alternative explanations, and because later investigations so frequently overturn earlier hypotheses, scientists repeatedly experience the tentative nature of scientific "knowledge."
Based on the above discussion, we recommend the following:
1. Embed Nature of Science instruction in normal, content-oriented lesson plans. That is, use an Implicit Instruction model, rather than separating the Nature of Science from the science itself.
2. As suggested by the literature, call attention to the concepts of the Nature of Science as they arise in the course of instruction. That is, make the important concepts explicit.
3. Make sure that every science lesson (or at least a majority of them) includes the fundamental core of science: reasoning from the data. This recommendation challenges tradition, so it will probably be difficult to do at first. In brief, we suggest these types of changes:
*Elementary Science: After students engage in an initial exploratory activity, making observations and (as appropriate at different ages) recording data, students should reflect on their findings. They should be given the opportunity to think about what the data imply, and to discuss their thoughts with each other. The teacher should guide this discussion and lead it toward a useful endpoint. Rather than telling students The Answer or leaving students entirely on their own to guess at answers, the teacher should help students derive their own answers from the data.footnote
* Secondary Science: The belief, for decades, has been that students need to know the background in order to understand the lab exercise. This turns out not to be entirely true. Put the labs first, before the discussion (or even short lecture) on the background information. Give students the opportunity to do what scientists consider to be the fun part of science: wrestling with the findings. Will students know everything if they do the lab first? No. Will they understand everything? No. Will they show signs of confusion? Yes. Will this mimic real research in real laboratories by real scientists? Yes -- and that's the point. As students work through the procedures, making mistakes and getting results that are not immediately clear, and talking amongst themselves about what the data tell them, they are engaging in authentic scientific reasoning. When the teacher subsequently leads a discussion and/or gives a short lecture that ties the lab experience together, the students have a background of personal experience that helps the discussion make sense. Too often, when we go over the facts and terminology first, students find that it's all new, and the cannot associate the information with anything, and therefore, can neither make sense of it nor remember it. Putting the lab first gives them a context for the information.
It is our experience, from the Summer Research Institute, that even this simple change makes a difference. It greatly diminishes the frequency with which students ask, "why do we have to learn this?"
The next step is to develop lesson plans that engage students in a scenario of some kind, and discuss possible approaches to solving the problem. Then, give the students real data (or have them find data on the web or from some other source). Again, give them time to wrestle with the data and make sense of it. Then start a whole-class discussion in which the students' interpretations come to the fore, and work toward a description of the conclusions as we know them today. This is one of the few strategies that can enable students to reason from data and develop understanding of complex scientific concepts, such as plate tectonics, transcription/translation, etc. Even if it is not possible for students to discover certain information on their own, in an open-ended inquiry, it is possible to examine some of the critical data and learn from it. As an example, consider the analysis of the genetic code. footnote
* College/University Science: Although it is challenging to develop strategies that foster student collaboration and discussion in a 300-seat lecture hall, it is possible. One solution is described in the preceding paragraph -- essentially short lead-in scenarios that give students the opportunity to think about the topic in particular ways. Other strategies involve the "think/pair/share" approach of asking a question, and having students think about it, discuss it pairwise, and then have a classroom-wide discussion. Where possible (mathematics and physics are good examples), it can be effective to ask students to solve particular problems collaboratively as a part of discussing problem-solving strategies. The biggest challenge facing college/university faculty is that the tradition of lecturing is so deeply entrenched that it is very hard to break out of this mold. We are particularly interested in open discussions among and contributions from faculty who have wrestled with this issue. footnote
Frequently, different student groups will propose somewhat different interpretations, when we ask them to think about data. This provides an ideal starting point for discussing the Nature of Science. "Scientific Knowledge" is comprised of scientists' interpretations of data. While studying the same problem, different scientists may reach different conclusions. When this happens, it is necessary to think about how to determine which interpretation is more likely to be correct. Without an "answer book" to check in, we must develop operational criteria for "reasonableness" of interpretations. These include:
- are the data reproducible? Were there technical errors in gathering the data? Sometimes, we can rule out alternative interpretations on the basis of operational difficulties that render the data invalid. But, in a classroom setting where everyone is examining the same data, this option is not available.
- is an interpretation consistent with other information? If valid information from other studies contradicts an interpretation, then the interpretation seems unlikely to be correct. In considering other information, there is a wonderful opportunity to ask "what did we learn last week or last month?"
- what are the implications of different, alternative explanations? Each explanation is an hypothesis, and as such, makes particular predictions. It is often possible to distinguish among alternatives when one of them makes unrealistic predictions.
- how complicated is the explanation? Scientists typically opt for the simpler explanation.
- what aspects of an interpretation are assumed to be true? Assumptions often color what we think, but assumptions are not always correct.
In some cases, possibly many cases, it may not be possible to distinguish between two alternative interpretations. This provides the ideal opportunity to remind students why scientific information is tentative. If we cannot distinguish between two alternatives, then we must consider them to be equally likely until we obtain additional data. Indeed, even if we have agreed upon a single likely interpretation, there is still the possibility that new data may indicate that our interpretation was incomplete.
With respect to the Nature of Science, we argue that students must experience it to understand it. They must wrestle with data. They must wrestle with data from which our scientific knowledge derives. It is not enough to present a lesson that demonstrates complete inquiry, and that illustrates the Nature of Science, but that uses an unimportant and unrelated topic (such as which type of paper towel is best). Scientific thinking should be built into our science teaching across the board -- and that means we must use strategies that lead students to develop understanding by reasoning from data. As they do so, teachers can explicitly point out those aspects of this reasoning that cause science to be what it is.
last updated:August 25, 2010