Bruce G. Stewart
Faculty - Biological Sciences
Department of Science
Advisor for Wildlife Conservation
Murray State College
1 Murray Campus St.
Tishomingo, OK 73460
(580) 371-2371  Ext. 225


Charter 3:  The Methods and Types of Explanations of Science


Philosophies of Science

Philosophers have debated the question of what constitutes science for centuries. Views have differed throughout the history of science. As scientific knowledge has increased there has been a concurrent evolution of scientific philosophy. For example, there have been the Baconian Philosophy, the Popperian Philosophy, and the Kuhnian Philosophy. All of these philosophies have addressed certain aspects of scientific methods and thought. The great biologist and philosopher, Ernst Mayr, has published two books, The Growth of Biological Thought and Toward a New Philosophy of Biology (1988) which I find particularly thorough and balanced. My ideas of scientific methods stem from his views as they have applied to my own science teaching and research.

Observational Method

Observation of nature is the fundamental basis of science. It may be thought of as the "what is" method. Collection of observational data must precede comparative and experimental methods before these can be applied in science. Mayr (1988) clearly points to the important role of observation in the following statement.

In fact, since the days of Copernicus and Kepler, observation and comparison have been exceedingly successful methods in such physical sciences as astronomy, geology, oceanography, and meteorology. And in biology, where observation and comparison have always been of paramount importance, experimental methods have been incorporated into the methodological repertory of many originally observational disciplines, including ecology and ethology.

Observation has dominated the early stages of most scientific disciplines. But even within established disciplines, such as astronomy, observation continues to produce important knowledge. In health science-related areas such as human anatomy, observation was the primary method used to describe the construction of the body. My own specialty area of ecology has been criticized at times as being a weak science because of the once rarity of rigorous experimental methods. Yet even today, there is a tremendous need for observational data in ecology. Thorough natural history studies, for example, are needed for a vast array of species.

Comparative Method

Comparison of data sets is a powerful method for discovering patterns in nature. Suppose, for example, you wish to determine if smoking tobacco (as the independent variable) is correlated with the prevalence of lung cancer (as the dependent variable) in humans. Since experimental exposure of humans to tobacco smoke for, say, a 30-year periods is both ethically unacceptable and logistically unfeasible, some other method of investigation must be used. The comparative method is ideal. Many humans have exposed themselves to significant amounts of cigarette smoke on a daily basis for more than 30 years. Thus, in effect, they have applied the independent variable of interest in our question about the smoking/lung cancer relationship.

We may hypothesize an answer to our question as follows: Inhalation of tobacco smoke by smoking one or more packages of cigarettes per day for thirty years or more increases the chance of getting lung cancer. Notice that this hypothesis is not a question! It is possible to apply the comparative method to determine if our hypothesis is supported. To do this, we must gather data on two groups of humans: a) an experimental group of people who have subjected themselves to our independent variable (smoking one or more packs of cigarettes daily for 30 or more years) b) a control group of people who have not smoked. By comparing the incident of lung cancer, which is our dependent variable, we can determine if there are significantly more cases of lung cancer in the experimental group. If this is the case, then we may say that our study supports our hypothesis. If we found no difference in the lung cancer rates between our two groups, then our hypothesis would not be supported.

Suppose our control group had been exposed to other potential cancer causing variables or possessed cancer correlated characteristics that our experimental group had not. Second-hand smoke, asbestos fibers, air pollution, stress, age, and many other factors have been associated with the prevalence of lung cancer. We must control the effect of such potentially confounding variables by keeping both the control group and experimental group the same in these respects. Preferably, neither group would have been exposed to these other potential cancer causing variables. In the case of age, both groups should be composed of the same age groups. These kinds of variables that could confuse our results but do not because we control them are called controlled variables. It is very important to understand why they must be controlled.

The comparative method is very good for finding patterns that may later be worthwhile to study in a "cause and effect" experimental study. The value of the experimental method is described in the next section.

Experimental Method

This method is the "typical textbook version" of the "scientific method". As we have seen, other methods are commonly used in science. However, the experimental method is especially powerful for determining cause and effect. Just because a pattern of correlation exists between two variables does not mean that one necessarily caused the other. An often-quoted example of such a "spurious correlation" is the increase in numbers of Baptist ministers that is correlated with the increase in violent crimes during the past several decades! Of course, the correlation does not mean that Baptist ministers cause violent crime! Both increases are also correlated with a third variable, namely, an increase in overall population.

In the case of cigarette smoking and lung cancer, we wish to make sure that correlations discovered by the comparative method are not spurious. Experiments can demonstrate cancer-causing effects of components of cigarette smoke on animal tissues and cells in the laboratory. Manipulating independent variables (nicotine, for example) has done this in controlled experiments. The experimental group is exposed to the independent variable, while the control group is not. Cause-and effect is suggested if cancer-related dependent variables (such as incidence of cancer or death from cancer) increase in the experimental group but not the control group. This is valid, of course, only if we have controlled other potentially confounding variables.

The experimental method, like the comparative method, includes the processes of:

An Example to Illustrate the Use of Observational, Comparative, And Experimental Methods in Animal Behavior

Dr. Douglas Mock of the University of Oklahoma is well known in the study of animal behavior. One of his special interests is a behavior termed "siblicide." Siblicide refers to the phenomenon of siblings (brothers or sisters) killing each other. Why would this behavior occur? Wouldn't it be evolutionarily advantageous for parents to intervene to save their own offspring? Is siblicide a rigid behavior that cannot be altered in the species that practice it? If it can be altered, what factors regulate the animals' decisions?

Dr. Mock has performed some elegant field experiments to answer these questions. In 1980, I had the opportunity to visit one of Dr. Mock's field research sites at Matagorda Bay, Texas. It was there that I first learned the siblicide aspects of the natural history of the Great Blue Heron and the Great Egret. As I tell this story, you should keep in mind the following terms and concepts: hypothesis, control group, experimental group, manipulated (= experimental) variable, controlled variables, and dependent variable.

Before we explore Dr. Mock's research, it would be helpful to learn about "natural history." Natural history is an area of biology defined by Bates (1990) as "the study of life at the level of the individual; of what plants and animals do, how they react to each other and their environment, how they are organized into larger groupings like populations and communities." Bates goes on to explain the problems of natural history.

It is amazing enough to stop and look at a forest or at a meadow - at the grass and trees and caterpillars and hawks and deer. How did all of these different kinds of things come about; what forces governed their evolution; what forces maintain their numbers and determine their survival or extinction; what are their relations to each other and to the physical environment in which they live? These are the problems of natural history, problems that concern us ourselves as animals and that concern us even more as originators of this thing we call civilization - which is, after all, merely a rather special sort of an animal community.

Thus, Dr. Mock's research on the Great Blue Heron and Great Egret was first and foremost based on knowing what these species "do." Observational and comparative data show that these species are water-associated birds with very similar characteristics. They are so similar that they are placed in the same taxonomic family. Both species are tall birds with long legs for wading, long necks for "striking" like snakes, and long sharp bills for spearing and capturing aquatic prey. Both species are colonial nesters, and Dr. Mock found a number of islands in Matagorda Bay on which both species nested in large numbers. This was an ideal location for research on their behavior.

Field observations indicated that both species lay an average of about four eggs per nest. Both species began incubation of the eggs after the first egg was laid. Hatching in both species occurred "asynchronously," meaning that the first egg laid was the first egg to hatch and the others typically hatched in order. This resulted in developmental advantages for the earlier-hatching chicks since the first chick (designated "A" chick) would be four to five days old by the time that "D" chick hatched.

A most interesting observation was that there was an amazing difference in the behaviors of Great Blue Heron chicks versus Great Egret chicks. Great Blue Heron chicks showed low levels of aggression toward each other. Great Egret chicks were highly aggressive to the point that in over 30% of the nests, one or more younger chicks were killed by their older siblings! Herein lay a general question, "Why would two otherwise similar species have nestlings that exhibit such differences in behavior?"

How do you imagine that aggressive behavior could be put into precise mathematical terms? Hundreds of aggressive encounters were observed. Dr. Mock and his research assistants distinguished specific categories of aggression and related behaviors. For example, a "strike" occurred when one chick struck another chick with its bill. A "retaliation" was a strike back by the second chick. Retreats occurred when a chick attempted to move away from its aggressor. These and other categories allowed mathematical quantification of data that could later be statistically analyzed and compared. In a sense, the observers were "judges" such as in a human boxing match.

One question of interest was "Is food type related to the magnitude of aggressive behavior?" Dr. Mock and his research assistants observed a difference in the size of food items that parent birds fed to their nestlings. One species brought their nestlings smaller numbers of larger individual food items while the other fed larger numbers of smaller items. Dr. Mock suspected that it would be possible for an older individual nestling to monopolize larger food items by driving younger nestlings away at feeding times. He suggested that a scattering of smaller food items could not be easily dominated.

A prediction based on the "food type" scenario is that nestlings should fight when the benefits (i.e. greater caloric intake) outweigh the costs (e.g. injury or less feeding success). Allow me to use an analogy to illustrate this logic. Suppose you were in a room with a group of three other students. Your professor walks in and throws $100 to the floor for anyone to take. Imagine your strategy for getting the most money if it were in the form of a single bill. How might this differ if the money were in the form of 100 scattered $1 bills?

To test the food type hypothesis, Dr. Mock conducted an ingenious field experiment. Great Egret eggs where switched with Great Blue Heron eggs in clutches at the same stage of incubation. Then the aggressive behavior of the chicks that resulted was quantified. The outcome was clear. When raised on food items of their close evolutionary "cousins," the chicks reversed their typical behaviors. Great Blue Heron chicks exhibited high levels of aggression and siblicide; Great Egret chicks became more peaceful!

This example of field experimentation is really two experiments in one. Behavior of Great Egret chicks was compared in a manipulated condition (i.e. raised on Great Blue Heron food type) to that in a natural condition (i.e. raised on Great Egret food type). The same was true for the Great Blue Heron chicks.

Some Literature:

Futuyma, Douglas. 1982. Science on trial: the case for evolution. Pantheon Books, New York.

Overton, William R. 1982. Judgment, injunction, and opinion for McLean v. Arkansas Board of Education. US District Court. Arkansas. 5 January 1982.

Schadewald, Robert J.  1992.  Looking for lighthouses.  Creation/Evolution.  National Center for Science Education.   12(2), Issue 31, Winter 1992.

© 1999 Bruce G. Stewart