There is no single, rigid method that scientists use in studying the natural world, but all scientists use a similar investigative process. The key elements of this process of science are the following:

1. Observations


2. Questions


3. Hypotheses


4. Predictions


5. Tests

In a scientific study, the initial observations take a variety of forms. They frequently come from previous studies. For example, an observation that has often been made about flying squirrels is that, when they land on a tree, they quickly scramble to the other side. This observation raises a question: Assuming this behaviour benefits the squirrel’s survival, how does it do so? A scientist who wishes to answer this question generates a testable hypothesis, a tentative explanation for the phenomenon. One hypothesis is that by moving quickly away from a landing site, a squirrel prevents an owl or other nearby predator for catching it when it lands. This hypothesis sounds reasonable, and it is often cited in articles about flying squirrels, but scientists are skeptical. Flying squirrels are nocturnal and glide very fast, making them extremely hard to observe. Also possibly relevant to our question is the fact that these squirrels are temporarily blinded by bright light. Do these observations make you wonder how many times flying squirrels have actually been seen landing? What if most observations have been made of squirrels landing on the bright sides of trees on moonlit nights? Critical thinking like this leads to alternative hypothesis; for example, maybe scrambling around the tree enables flying squirrels to avoid bright moonlight.

Having formed one or more hypothesis, and with future observations or experiments in mind, scientists then use deductive reasoning to predict the results of new observations or experiments. A prediction states the results expected from observational or experimental tests if the hypothesis is correct. Deductive reasoning is the use of “if… then” logic. In this case, if a hypothesis is correct, and we can perform a test of the hypothesis, then we can expect a particular outcome. For example, if flying squirrels scramble around a tree simply to avoid bright light (hypothesis), and we devise a way to observe them landing in both shadow and light (test), then we would expect that squirrels landing on the dark sides of trees would not immediately move to the other side (prediction).

Suppose we make this hypothesis and prediction and then observe numerous landings of flying squirrels, perhaps by taking infrared photos with remote cameras, while monitoring light conditions. We find that, in all light conditions, the squirrels scramble to the other side of the tree after landing. This result does not agree with our prediction and is therefore evidence that our hypothesis is false. However, our result neither supports nor falsifies the alternative hypothesis that scrambling around the tree enables squirrels to avoid predators. (Can we think of a way to test this alternative hypothesis?)

Our discussion of flying squirrels has brought up several features of the scientific process, in addition to the roles of observations, questions, hypothesis, predictions, and tests. We have seen that science involves critical thinking at every step. Furthermore, science is cumulative, with the results of earlier studies often serving as the initial observations for new studies. We have also seen that the scientific process can falsify or support hypothesis. Note the key phrase “support hypothesis”. It is important to realize that science does not pretend to prove hypothesis with absolute certainty, because it is not possible to make enough observations or to repeat an experiment enough times to be absolutely certain results will always be the same.