1.8: Step 4: Conduct a critical experiment.

Once we have formulated a hypothesis that generates a testable prediction, we conduct a critical experiment, an experiment that makes it possible to decisively determine whether a particular hypothesis is correct. There are many crucial elements in designing a critical experiment, and Section 1-11 covers the details of this process. For now, it is important just to understand that, with a critical experiment, if the hypothesis being tested is not true, we will make observations that compel us to reject that hypothesis.

In this step of the scientific method, a bit of cleverness can come in handy. Suppose we were to devise a critical experiment to test our hypothesis “Eyewitness testimony is always accurate.” First, we stage a mock crime such as a purse snatching in front of a group of observers who do not know the crime is staged. Next, we ask observers to identify the criminal. To one group of observers we might present six “suspects” all at once in a lineup. To another group of observers we might present the six suspects one at a time. The beauty of this experiment is that we actually know who the “criminal” is. With this knowledge, we can evaluate with certainty whether an eyewitness’s identification is correct or not.

This exact experiment was done, but with an additional, devious little twist: the researcher did not include the actual “criminal” in any of the lineups or one-by-one presentations of the six “suspects.” If eyewitness testimony is always accurate, however, this slight variation should not matter. We would predict that the observers in both groups (the lineup group and the one-at-a-time group) would indicate that the criminal was not present. In the next section, we’ll see what happened. Right now, let’s devise critical experiments for our other hypotheses.

Hypothesis: Echinacea reduces the duration and severity of the symptoms of the common cold. The critical experiment for the echinacea hypothesis has been performed, and it is about as close to a perfect experiment as possible. Researchers began with 437 people who volunteered to be exposed to viruses that cause the common cold. Exposure to the cold-causing viruses was a bit unpleasant: all of the volunteers had cold viruses (in a watery solution) dripped into their noses. The research subjects were then secluded in hotel rooms for five days, and doctors examined them for the presence of the cold virus in their nasal cavity and for any cold symptoms (FIGURE 1-11). (It is actually quite rare to expose human subjects to illness, and they always must be informed of the risks and then must give, by signing a form, what is called “informed consent.”)

Before the cold virus treatment, the volunteers were randomly divided into four groups. In two of the groups, each individual began taking a pill each day for a week prior to exposure to the virus. Those in one group received echinacea tablets, while those in the other took a placebo, a pill that looked identical to the echinacea tablet but contained no echinacea or other active ingredient. Neither the subject nor the doctor administering the tablets (and later checking for cold symptoms) knew what they contained. In the other two groups, the individuals did not begin taking the tablets until the day they were exposed to the cold virus. Again, one group got the echinacea and the other group got the placebo.

Hypothesis: Hair that is shaved grows back coarser and darker. A critical experiment does not have to be complex or high-tech. All that matters is that it can decisively determine whether or not a hypothesis is correct (FIGURE 1-12). This point is illustrated by an experiment published in the scholarly journal Archives of Facial Plastic Surgery. Researchers decided to test the hypothesis that shaving hair causes it to grow back coarser or darker. A group of volunteers had one of their eyebrows, selected at random, completely shaved off. The subjects were then evaluated for eyebrow re-growth over the course of the next six months, with observers also analyzing photographs of the individuals.

In the next section we’ll see how our hypotheses survive being confronted with the results of these critical experiments and how we can move toward drawing conclusions.