The Purpose of Experimenting

The purpose of any experiment is to collect data, which, in turn, needs to be analyzed or processed to derive inferences. There are four ways of collecting data: observation, surveys, computer sim­ulation, and experiment. Though analyses of data using these methods share common features, our interest here is confined to the collection of data from experiments. An astronomical obser­vation, for example, is not an experiment, though it entails data collection. This is so because there is little we can do to control the independent variables, in this case, the arrangement of astro­nomical objects. A laboratory experiment, which is our concern in this book, is understood to be such only when the one or more independent variables, which act upon the dependent variable, can be controlled to the required levels.

Through the medium of the experimental setup, we alter one or more independent variables within the possible, or required, range, and we observe, measure if need be, and record the out­come, which either is itself or leads to the determination of the dependent variable. In a situation like this, if the independent variable is A and the dependent variable is B, it is customary to say that “A caused B,” where A is referred to as the cause and B as the effect. The change made in quality or quantity (of a measur­able property) on the cause, and the corresponding change in quality or quantity (of a measurable property) in effect, observed or measured and recorded, constitutes data. In most experiments in the physical sciences (including technology), to connect one independent variable, the cause, with one dependent variable, the effect, is fairly common. Can there be situations wherein causes A1, A2, A3 . . . can all participate together to produce an effect B? The answer is yes, and this indeed is, more often than not, the case in biological, as well as some industrial, experi­ments. In this chapter, we restrict the discussion to one cause-one effect relations, but only after noting that although convenient and often adequate, these may fall short of correctness. Discus­sion of the combined effects of many causes will be dealt with in Chapters 7, 8, and 9.

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