Types of Experiments
Countless experiments have been performed in a variety of contexts and for a variety of reasons in the laboratories of the world. In general, in terms of goals, experiments may be put into three broad categories.
The first may be called the mere measurement type. An experiment may be performed solely for the purpose of measuring a certain physical quantity. For example, one may wish to determine the density of a newly found substance, the distance of a star, the heartbeat of an animal, the intensity of the earth’s magnetic field at a place, etc. Innumerable experiments of this kind are carried out every day all over the world.
Experiments of this type are among the most common and valuable types of experiments in the scientific enterprise, because measurement is the life-blood of modern science. In discussions on science-religion one seldom recognizes that is a crucial matter where the two simply do not overlap. Religion has little to do measurement, but there can be no science, in the modern sense, without measurement. This is also an important respect in which modern science differs from ancient science where measure played, if at all, only a minor role. As Ernest Nagel pointed out “It is by discovering the recurrence of certain constants in different numerical laws that the ideal of a unified science is progressively realized.”
Measurements have revealed aspects of the world that we would never have known otherwise. It was the precise measurements of wavelengths from burning hydrogen that led to the formulation of the atomic theory, and thence to the exploration of the microcosm. It was the measurement of the recession of galaxies that gave rise to the big-bang theory of cosmology. Another matter of extreme importance, not always recognized, is that measurements or limitations on them have been responsible for giving a jolt to scientific epistemology. For example, Heisenberg’s famous principle, which sets a limit on the precision in our measuring efforts microcosmic entities, has had enormous impact on how we view ultimate nature of reality.
The second type of experiment has as its goal verification or falsification. Here one tries to verify a statement that follows from theoretical considerations or from a fellow scientist’s report. For example, we may be told that when dropped from a height, a feather and a ball of lead will both fall to the ground at the same rate. To see if this so or not, one can do an experiment. That is the only way to determine the correctness or otherwise of the proposition. For many centuries, it was thought, as Aristotle had stated, and as intuition might suggest, that the heavier of two bodies would fall faster.
Or again, when a scientist reports that he or she has discovered a new phenomenon, such as cold fusion, others try to replicate it in their laboratories to establish the truth or the falsity of the claim. The notion of such a procedure did not exist at any methodological level in the framework of ancient science. This is one reason why a good many mistaken views persisted for long centuries in the ancient world. One has only to glance into the Natural History of Caius Plinius Secundus (also known as Pliny) to see how many errors, mistaken views, and superstitions of the ancient world used to be taken as scientific facts or theories. This encyclopedic compendium of ancient errors, which discusses topics from anthropology to zoology, recommended, for example, sexual union for such ailments as impaired eyesight, hoarseness, and melancholia; and asked women not to sneeze after intercourse for this would abort the fetus. Because the notion of verification was not ingrained in the methodology, Plinyâ€™s book served as a major source of knowledge during the Middle Ages. Modern writers who are very eager to establish that the ancients knew all about modern science should peruse Plinyâ€™s volumes which is a treasure chest of ancient views.
The third type includes experiments whose goal is to uncover relationships. In the incessant transformations taking place in the physical universe, innumerable physical quantities are involved. Most of these are variables. One of the interests of science is to explore and discover relationships among them. For example, we may wish to know in what specific manner the current in a wire depends on the voltage difference between its terminals, how the pressure of an enclosed amount of gas changes with temperature, or what the relationship, if any, is there between the amount of food one eats and the number of hours spent in sleep. Relationships of this kind are at the basis of the general laws of nature.
The above categorization is by no means watertight. It refers only to the initial plan or purpose that the experimenter has in mind. Sometimes the goals of experiments may overlap. For example, some aspects of an experiment of the second or third kind may be mere measurement. Sometimes while performing an experiment of one kind, the investigator may change it into an experiment of a different kind. Thus, for example, in trying to measure the energy with which electrons (beta rays) are emitted from radioactive materials – a measurement type of experiment – one may discover that this is not, as one would expect, uniform. In fact, this is what led to the discovery of the neutrino. Thus, it turned out to be an experiment of the third type. Or again, in trying to verify that the current through a conductor is proportional to the potential difference between its ends, the experimenter may discover that this is not quite true at higher temperatures. Thus an experiment of the second kind is transformed into one of the third.
There is one important characteristic of all scientific experiments: replicability, sometimes called repeatability. A result following from an experiment is considered to be valid only if the experiment, when done again, yields the same result. Many stray observations and alleged facts have turned out to be non-repeatable occurrences, if they had occurred at all in the first place. Much controversy between science and non-science arises from the importance that science attaches to replication.
The term repeatability, though commonly used in this context, needs to be clarified. One does not mean that every phenomenon and process, or even experiment, considered in science must be capable of being repeated in practice, but rather that, if repeated, the results must be the same. Science attaches great significance to many non-repeatable conceptual experiments. Many theories pertaining to geology, astronomy, astrophysics, and evolutionary biology belong to this category. In these instances it is physically impossible to repeat the experiments that are implicit in the theories. What is implied, however, is that if the experiments were repeated the (long range) results would be as stated (or observed). Thus what one actually means by repeatability is consistency of results when the experiment is repeated.
Then again, replication, whether actual or in principle, is more easily done in the physical sciences. But when we treat psychology and sociology as science, the matter becomes enormously more complex. Even in physics, one may argue on purely logical grounds whether two situations can be precisely replicated so as to verify a theory.