Index to this page

Scientific Methods

There is nothing mysterious or even particularly unusual about the things that scientists do.

There are many ways to work on scientific problems. They all require common sense. Beyond that, they all display certain features that are especially — but not uniquely — characteristic of science.

For example:

Skepticism. Good scientists use highly-critical standards in the judging of evidence. They approach data, claims, and theories (ideally, even their own!) with healthy doses of skepticism.
Tolerance of uncertainty. Scientists often work for years — sometimes for an entire career — trying to understand one scientific problem. This often involves finding facts that, for a time, fail to fit into any coherent pattern and that even may support mutually contradictory explanations.
Sometimes, as one listens to scientists vigorously defending their views, their confidence seems absolute. But deep in their hearts, they know that their views are based on probabilities and that a new piece of evidence may turn up at any time and force a major shift in their views.
Although they certainly have no monopoly on hard work, their willingness to work long hours and years pursuing a problem is the mark of all good scientists. For science is hard work.
Before undergoing the frustrations — tempered by occasional joys — of wresting more secrets from nature, you must learn the foundations on which your subject is based.

Although scientific methods are as varied as science itself, there is a pattern to the way that scientists go about their work.

Scientific advances begin with observations.

A census of the members of a species in some habitat is an observation.
The readings on the display of a laboratory instrument are observations.

But science is more than a catalog of facts.

The goal of science is to find an explanation for why the facts are as they are. Such an explanation is a hypothesis.

Link to a case study illustrating the scientific method at work.

Testing Hypotheses

A good hypothesis meets several standards.

It should provide an adequate explanation of the observed facts.
If two or more hypotheses meet this standard, the simpler one is preferred.
It should be able to predict new facts.

So if a generalization is valid, then certain specific consequences can be deduced from it.

One of the most exciting events in science is to

predict the results of an experiment not yet performed if the hypothesis is valid and then to
perform the experiment.

Link to an example.

Experimental biology often involves setting up an experimental treatment and — at the same time — a control. Then one compares the results of the experimental treatment with the results in the controls. If there is a difference, what is the probability that it is due to chance alone; that is, the experimental treatment really had no effect?

The hypothesis that the experimental treatment had no effect is called the null hypothesis.

Most workers feel that if the probability (designated p) of the observed difference is less than 1 in 20 (p = <0.05), then the null hypothesis is disproved and the observed difference is significant.

Link to discussion of statistical methods.

But significance is not proof. In fact, hypotheses can never be proven to be absolutely "true" is the sense that a theorem in geometry can. The most we can say is that there is a high probability that the hypothesis provides a valid explanation of the phenomenon being studied.

Hypotheses that are supported by many observations come to be called theories.

So, in contrast to some areas of human thought, science can never prove that a theory is "true". But it can show that a theory is false.

Lest the tentative nature of science cause you to lose confidence in it, think of what science has produced. The many achievements of scientific methods, despite the absence of absolute certainty, has been well-expressed in the sonnet "Paradox" by the late mathematician Clarence Wylie, Jr.

Not truth, nor certainty. These I foreswore
In my novitiate, as young men called
To holy orders must abjure the world.
"If..., then...," this only I assert;
And my successes are but pretty chains
Linking twin doubts, for it is vain to ask
If what I postulate be justified,
Or what I prove possess the stamp of fact.
Yet bridges stand, and men no longer crawl
In two dimensions. And such triumphs stem
In no small measure from the power of this game,
Played with the thrice-attenuated shades
Of things, has over their originals.
How frail the wand, but how profound the spell!

Reproducibility of Scientific Work

The single feature that is most characteristic of science is its reproducibility. If scientists cannot duplicate their first results, they are forced to conclude that these were invalid. This problem occurs often. Its cause is usually some unrecognized, and hence uncontrolled, factor in the experiment (e.g., unrecognized variation in the properties of different batches of the materials used in the experiment). With luck, the inability to reproduce experiments will be discovered by the same scientists who did the first experiments. This is why scientists generally repeat their experiments several times before reporting them in a scientific paper.

Link to a description of the format of scientific papers.

On other occasions, workers in another laboratory fail to secure the same results when they

repeat experiments that have been published or, more often,
perform experiments designed to carry the study into new areas, but these fail because of a flaw in the original experiments.

When this happens, all the parties concerned should get together to see if they can find out why their results differ.

Often it is simply a matter of not using precisely the same materials and methods.
Sometimes, however, a serious flaw may be discovered in the design and/or execution of the original experiments.
And sometimes it proves impossible to find out why experiments that once seemed to work no longer do so.

In any of these cases, the failure to confirm the experiments must be reported.

Although this is acutely embarrassing for the original investigators, it represent one of the great strengths of science: its built-in system for self-correction.

Scientific Fraud

In the vast majority of cases, irreproducible results in science are caused by honest errors.

On rare occasions, however, laboratory reports cannot be confirmed because they are fraudulent. This is distressing to all concerned. If such a fraud becomes widely known, it is also likely to cause a great deal of excitement among the general public.

I believe, however, that rather than casting a cloud over the scientific enterprise, these rare aberrations reveal its great strength.

There is probably no other area of human activity where error is detected and corrected more rapidly. I am confident that you can think of a number of other fields of human study and activity where errors have been made that went uncorrected for years and caused widespread harm.

Dishonest scientists usually harm only themselves. They are disgraced; their careers often at an end.

But the progress of science usually moves forward as fast as (sometimes faster than) before.

Building on the Work of Others

Only rarely does a scientific discovery spring full-blown on the scene. When it does, it is likely to create a revolution in the way scientists perceive the world around them and to open up new areas of scientific investigation. Darwin's theory of evolution [Link] and Mendel's rules of inheritance [Link] are examples of such revolutionary developments.

Most science, however, consists of adding another brick to an edifice that has been slowly and painstakingly constructed by prior work. In fact, it is possible to construct a genealogical tree that traces the historical development of any scientific discovery (even, to a degree, Darwin's and Mendel's). The way in which science builds on the work of others is another illustration of what a communal activity science is.

The development of a new technique often lays the foundation for rapid advances along many different scientific avenues. Just consider the advances in biology that discovery of the light microscope and, later, the electron microscope have made possible. Throughout these pages, there are many examples of experimental procedures. Each was developed to solve a particular problem. However, each was then taken up by workers in other laboratories and applied to their problems.

In a similar way, the creation of a new explanation (hypothesis) in a scientific field often stimulates workers in related fields to reexamine their own field in the light of the new ideas. Darwin's theory of evolution, for example, has had an enormous impact on virtually every subspecialty in biology (and in other fields as well). To this very day, biologists in specialties as different as biochemistry and animal behavior are guided in their work by evolutionary theory.

Basic Versus Applied Science

The distinction between basic and applied science is more one of goals than of methods. The same rules and standards apply to each.

However, the motivation behind the work is somewhat different. Researchers in applied science have before them a practical problem to be solved. Much of the research that goes on in medicine and in agriculture is applied.

The researcher in basic science, on the other hand, is primarily driven by curiosity - the desire to find out more about how nature works.

Both types of research are not only honorable and demanding professions, but they are mutually dependent as well.

Applied science repeatedly loses momentum without periodic infusions of fresh ideas and discoveries from basic research. (The light bulb would never have been discovered in the research and development (R and D) department of a candle manufacturer!)
On the other hand, much basic research has depended on the development of new tools and instruments and, more often than not, these have been developed in laboratories devoted to applied research.

Welcome&Next Search