Is a Hypothesis Necessary to do Science?

Forrest M. Mims III

When my children were doing science fair projects, one of the most frustrating challenges they faced was fulfilling the requirement in the rules for stating a hypothesis about the outcome of their project before they started their research. This step was claimed to be a key part of the scientific method, and students were generally required to include a hypothesis on their project poster and in their report. They still are.

So what is a hypothesis? Is stating a hypothesis an essential step in doing science?

Science Service, administrator of the Intel International Science and Engineering Fair (ISEF), includes this explanation of hypothesis in its copyrighted definition of the scientific method:

Hypothesis

Are these hypothesis steps essential for doing a science project? Do scientists follow these steps?

I've found that the answer is no much more often than yes, and a classic example is my daughter Sarah's discovery of living microbes in smoke arriving in Texas from Yucatan (Fig. 1). Sarah was looking for microbes in Asian dust that reaches Texas during spring. Her finding of microbes in Central American smoke was pure serendipity that would have never been predicted in advance by a hypothesis. Sarah didn't include a hypothesis on her display board unless it was absolutely required, and she went on to earn many major awards for her discovery.

While some of the science I've published in peer-reviewed journals of science was based on a hypothesis, most was not. Instead, I simply asked a question and then planned and executed experiments to provide answers.

For example, does a solar eclipse cause waves to form within the ozone layer? Some scientific papers said yes and others said no. So I went to a solar eclipse with a homemade instrument to measure the ozone layer and find out for myself. My son Eric stayed home to make observations off the direct track of the eclipse using a second ozone instrument I built.

It never occurred to us to predict the outcome of the experiment before it was conducted. So we were as surprised by finding that the eclipse did indeed produce waves in the ozone layer as we would have been had no waves been detected. The results of this experiment-without-hypothesis were published in a leading peer reviewed journal (F. M. Mims III, and E. R. Mims, Fluctuations in Column Ozone During the Total Solar Eclipse of July 11, 1991, Geophysical Research Letters , 20, 5, 367-370, 1993).

The editor of the journal that published our paper never asked us about our minimal academic credentials. Nor did he ask us to add a specific hypothesis to the paper. In fact, only a few of my scientific publications have invoked a hypothesis, and I have never been asked by a journal editor or peer reviewer to include one. That's because a hypothesis doers not appear in most scientific papers.

So why are students usually required to develop a hypothesis for their science fair projects? And why does Science Service, the largest science fair promoter in the US, advocate the hypothesis step?

Richard D. Storey and Jack Carter asked and answered the hypothesis question 14 years ago in a superb article they called "Why the Scientific Method?" (The Science Teacher, December 1992, pp. 18-21). Recently I remembered this article and reread it.

The copyrighted article begins, "Why are teachers the only members of the scientific community who remain closely tied to THE scientific method? Why do they require their students to memorize the steps of THE method...What, then, is outdated about teaching THE scientific method? First and perhaps foremost, a hypothesis is rarely stated today, because many scientists dislike its formality and consider its rigidity questionable ands restrictive to their creativity. In addition, it is unrealistic to expect a hypothesis to always originate in a pure and unbiased manner, from a scientific mind free of preconditioned notions. Such hypotheses can seem feigned, spurious, and perhaps even haughty in some cases.

"It follows, then, that we have seldom encountered a formally stated hypothesis during our years of attending scientific meetings, serving as referees for manuscripts and acting as peer reviewers of research proposals for various funding agencies. We underscore, nonetheless, there is really nothing unscientific about stating a formal hypotheses, it just is not done very often, especially as a first step in research....

"What, then, do scientists do? Almost exclusively, we begin or continue a research project with a question instead of a formal hypothesis. This questioning is probably the most important part of scientific creativity....[This] is the basis of research in many disciplines, yet seems to be largely ignored in science textbooks and classrooms."

Storey and Carter were totally on target. That's because they were real scientists, not science fair administrators and rule writers.

Unfortunately, a generation of students has passed through the schools since their article appeared, and the hypothesis requirement is still in place.

I've found that a hypothesis was an essential step in several major research projects that I conducted and published. But most of my research has not required a hypothesis.

What do you think? Does the science you do or study require a hypothesis in advance? Or do you simply ask a question and think about or actually design experiments and observations to find the answer? Send your thoughts here, place "SAS Hypothesis" in the subject line and we'll publish your opinion in Backscatter so you can have your say.