5.1.3 Testing Hypotheses

A hypothesis is a proposed explanation for an observed process or phenomenon. Human beings formulate hypotheses because they wish to answer specific questions about the world. Usually, the sciences come to mind when we think of the word “hypothesis.” However, hypotheses can be created on many subjects, and chances are that you have created many hypotheses without realizing it. For example, if you often come home and find that one of your outside potted plants has been knocked over, you might hypothesize that “the wind must have knocked that one over.” In doing so, you answer the question, “Why is that plant often knocked over?” Generating and testing hypotheses engages different forms of reasoning— abduction, induction, and deduction—all of which will be explained in further detail below.

Clearly, simply coming up with a hypothesis isn’t enough for us to gain knowledge; rather, we must use logic to test the truth of our supposition. Of course, the aim of testing hypotheses is to get to the truth. In testing we often formulate if–then statements: “If it is windy, then my plant will get knocked over” or “If nitrogen levels are high in the river, then algae will grow.” If–then statements in logic are called conditionals and are testable. For example, we can keep a log registering the windy days, cross-checked against the days on which the plant was found knocked over, to test our if–then hypothesis.

Reasoning is also used to assess the evidence collected for testing and to determine whether the test itself is good enough for drawing a reliable conclusion. In the example above, if on no windy days is the plant knocked over, logic demands that the hypothesis be rejected. If the plant is sometimes knocked over on windy days, then the hypothesis needs refinement (for example, wind direction or wind speed might be a factor in when the plant goes down). Notice that logic and reasoning play a role in every step of the process: creating hypotheses, figuring out how to test them, compiling data, analyzing results, and drawing a conclusion.

We’ve been looking at an inconsequential example—porch plants. But testing hypotheses is serious business in many fields, such as when pharmaceutical companies test the efficacy of a drug in treating a life-threatening illness. Good reasoning requires researchers to gather enough data to compare an experimental group and control group (patients with the illness who received the drug and those who did not). If scientists find a statistically significant difference in positive outcomes for the experimental group when compared to the control group, they can draw the reasonable conclusion that the drug could alleviate illness or even save lives in the future.