Planning for the Experiment

Experiments are often time-consuming, if not also expensive. It is reasonable, then, to plan the goal, extent, and step-by-step course of experiments before embarking on their execution. Since sci­ence is not a product of a manufacturing process, the unforeseen and the unavoidable are to be expected. But the element of con­tingency should be reduced to the minimum possible. An unplanned, haphazard beginning can be rather wasteful, both in time and money, besides possibly leading to a dead-end. In pre­paring to conduct experiments, one needs to attend to several details, some obvious, some not so obvious.

A large number of experiments consist in correlating an inde­pendent variable, a cause, with a dependent variable, an effect. Between the cause and the effect, speaking in terms of strategy, is located a device, usually called the experimental setup, whose function is to receive the input in the form of cause and to pro­duce the output in the form of effect. The input and output, either one or both, may involve some materials or objects that need to be chosen and may not be arbitrary. The required means is sampling. It may also happen that the input and output, either one or both, may not be tangible objects but measurable quanti­ties, such as pressure, temperature, and voltage. In either of these cases, planning the experiment is relatively straightforward. On the other hand, in such areas of study as agriculture, patient care, and industrial quality assurance, in which several variables, or causes, simultaneously influence the desired output(s), or effect, the planning of experiments can become complex and confusing. While a considerable amount of literature has grown up around the various aspects of preparing to conduct experiments, an anal­ysis now known as “Design of Experiments” is indispensable for planning such many-cause experiments as mentioned above. The earliest attempts at design seem to have originated with the scien­tific study of agriculture; the guesswork of Bacon relative to crop yields from seeds subjected to different treatments (one of those was to soak seeds in urine) is an example. Systematic work based on sound principles of experimentation, again in agriculture, began in the early part of the twentieth century. Some of the basics involving more than one cause are dealt with in Chapters 7, 8, and 9. The few, basic principles of planning for experiments discussed in this chapter are somewhat general in nature and can be applied with benefit to experimental work in many diverse areas of study.

Source: Srinagesh K (2005), The Principles of Experimental Research, Butterworth-Heinemann; 1st edition.