The Scientific Method

Science is a way to gain knowledge about natural phenomena using empirical observation and testing (Jurmain, et al 2013).

While there are different protocols used in science, it is performed using a set of rules called the Scientific Method that guide scientific practice (Fancher 2000). The method stresses the need to develop a testable hypothesis, the use of objectivity and rationality, and the circularity of scientific research. This does not mean that science is infallible, but in using the scientific method, particularly having testable hypotheses and tests that are replicable by other researchers, rigorous conclusions are reached.

Why do students of biological anthropology need to understand the scientific method? Because anthropology is a science.

Steps of the Scientific Method

1. Define the problem

This is based on observation — either something you’ve observed from nature or from something that’s already been written. How many times have you seen something or read about something and thought up a question about it? If you have, then you’ve done the first step of the Scientific Method!

2. Develop Hypothesis

Propose an explanation for the observed phenomenon. For a hypothesis to be a GOOD hypothesis, it must be testable. In other words, you must be ABLE to test it to see if it’s supported or not supported. A hypothesis is still good, even if it’s unsupported, as long as it’s testable. There are lots of questions out there that are not testable, such as, “Somewhere on the planet, a pink elephant is dancing the can-can in a tutu.” Because it’s impossible to explore every inch of the world looking for our dancing elephant, this would not be feasibly testable. Along the same lines, a testable hypothesis would be, “Student group A, who have read the material on The Scientific Method, will have a higher average score on their quiz than will Student group B, who have not read the material on The Scientific Method.”

3. Collect Data

The experiment should be specifically designed to test the hypothesis. The experiment will provide data as to whether the hypothesis is supported or not. Experiments must be replicable by other researchers. For our example above with ever-so-fortunate quiz-takers, you’d have Group A read the material (Group B would not) and both groups would take the quiz. This experimental design is replicable by other researchers because they too could find two groups of students, and follow the same protocol (one group reads the material, one not, and both take the quiz).

4. Organize/analyze data:

Once the experiment is completed, you must organize the data and analyze the results. For our example, you would grade the quizzes and calculate the average score per group.

5. Conclusion

You develop a statement that sums up what the data collected during the experimental phase says about the hypothesis. For our example, our concluding statement would be something like, “Group A, who read material on The Scientific Method, performed better on a Scientific Method quiz than did Group B, who did not read the material.” If your hypothesis is supported, then you would move on to the next step in the process. If your hypothesis was not supported, you would rework your hypothesis and start the process over.

6. Share the knowledge

It is important to share the results of your work even (and maybe even especially) if your hypothesis was not supported. This hasn’t been a step in the old school version of The Scientific Method, but remember — information is only good if it’s communicated to others!

Things to keep in mind:

Science does not “prove” anything. Hypotheses are falsified/unsupported or not falsified/supported. For a hypothesis to be accepted as a theory, which is a generally accepted explanation of specific phenomena, it undergoes rigorous testing (Larsen 2008: 16). It can take decades for a hypothesis to become a theory.

While we discuss various topics in biological anthropology, keep in mind that each piece of information was once a hypothesis that was tested and supported by the data. Contradictory supported-hypotheses are possible; some refer to this as equifinality. This just means that the data collected supports two or more hypotheses — in most cases, there is not sufficient data available to support one over the other (especially when we get to the information within human evolution). Hopefully, in these cases, one day we’ll have enough evidence to overwhelmingly support one particular hypothesis, but at the present that’s not possible. So, for the moment we have to open our minds to accept more than one potential conclusions.

The Scientific Method in Action!!! Watch the following clip and identify the steps of The Scientific Method. Was their hypothesis testable? Was their experiment replicable? What was the conclusion of their “scientific endeavor”? Clip: The Scientific Method in Action.

The Scientific Method: Experiments

Experiments test relationships between variables. In other words, does a change in one variable cause changes in other variables?

There are three types of variables:

1. independent variables: condition or event being studied
2. dependent variables: condition or event that might change because of the independent variable
3. control variables: conditions or events that stay (or at least attempted to make stay) the same

There are two types of experiments:

1. controlled experiments: generally designed for a laboratory setting; researcher has a great amount of control over the variables
2. natural experiments: researcher has little to no control over variables (think about how difficult it would be to control the behavior of chimpanzees as you observe them in the wild)

Assumptions in Science

Scientists work under several assumptions:

1. universalism: hypotheses are testable and replicable by others
2. objectivity: testing without bias
3. rationality: arguments are drawn from the data at hand
4. veracity: researchers are telling the truth

Obviously, there are problems with the assumptions. Sometimes results are falsified; scientists are human, therefore, they struggle with the same foibles that other humans do. This is where universalism comes into play–if someone else were to run the same experiment using the same or similar data set, they should come up with the same results. If the results are different then there was some problem with the original experiment or problems with the test. In that case, researchers generally run their tests again to determine whether it was a problem with the original test or the retest.

Another “problem” is equifinality–a large number of alternative hypotheses can be devised that are equally unfalsifiable. A researcher may get to a point where there are two or more plausible explanations for observed phenomenon based on existing data. I put quotation marks around “problem” because equifinality does not necessarily have to be viewed as a problem. For instance, when we look at human evolution later in the quarter, we will see there are several equally plausible phylogenies of human evolution. What the situation is doing is driving more research. As more data is collected, we will be able to start falsifying phylogenies. More research is a good thing.

References

Fancher LJ. 2000. The Great “SM.” < http://www.cod.edu/people/faculty/fancher/scimeth.htm>. Accessed May 7, 2015.

Jurmain R, Kilgore L, Trevathan W. Essentials of physical anthropology, 4th edition. Belmont (CA): Wadsworth, Cengage Learning; 2013. 437 p.

Kaziek CJ, Pearson D. 2014. Ask a biologist: using the scientific method to solve mysteries. http://askabiologist.asu.edu/explore/scientific-method>. Accessed May 7, 2015.

Larsen CS. Our origins: discovering physical anthropology. New York (NY): W.W Norton & Company, Inc.; 2008. 430 p.