Steven J. Dick, Ph.D.
The scientific method is difficult to understand. It seems to be filled with double negatives (e.g., failure to support the null hypothesis) and convoluted language. It is easy to skip to the bottom line and assume a research project “proves” the theory.
Understanding scientific research can be difficult. Sometimes a simple example is a great place to start. To look at the investigation of a cause/effect relationship, I propose the following simple situation.
I wake every morning with a high-pitched tone in at least one ear (tinnitus). Within a few minutes, the tone is gone or substantially reduced.
Theory: My audiologist maintains the theory that the tone reduction comes from the rather expensive hearing aides I insert. She postulates that the distraction of a fuller range of sounds frees my mind of the tone.
The scientific method is not about proving my audiologist’s theory. Experienced researchers attempt to show that alternative explanations are less likely to be the cause. What are some alternatives?
- Morning coffee: The chemical changes caused by caffeine result in a change that quiets the tinnitus.
- Positional differences: At night, I am primarily horizontal. In the morning, I am primarily verticle. A change in position causes a reduction in the tone.
- Daybreak effects: As a body prepares for the new day, the tone is reduced as a side effect.
Now that three alternatives have been identified, it is easy to imagine test construction. With each test, the researcher is not trying to prove the causal effect of the hearing aide but rather the causal effect of the alternatives. The researcher tests the alternative (null) hypothesis that one of the other causes is correct. Each time an alternative is shown unrelated (or less likely), the primary hypothesis is supported (but not proven). Of course, there are still many alternative explanations. The hearing aide works for a different reason or a combination of effects.