BACKGROUND

A research issue that is commonly encountered is to assess the effect of one variable (called treatment) on another, called an output. It is alternatively known as ‘to measure or evaluate the effect of a treatment on an output variable’, that is, to measure the change in an output created by the treatment. When a change in the output is exclusively due to treatment and not because of any other variable, it is termed as its causal effect. Evaluation designs facilitate understanding the causal effects.

A treatment, whose effect is to be determined, is generally considered to represent two categories denoted by the presence or absence of a characteristic. For obtaining a treatment effect, two sets of observations are needed, one having the ‘presence of a treatment’ (denoted as T) and the other called control would be ‘without the treatment’ (denoted as C).

It is not mandatory for a treatment to have only two categories. It is possible, for example, to measure the effect of more than one medicine for a disease in a single study. To gauge the effect of two medicines, we need at least three sets of observations on C, T1 and T2, where the first is the control (i.e., without any medicines) and the other two with the two medicines in evaluation. The focus can also be to comprehend the potency of a medicine or the optimum amount of a fertilizer to use. In this case too, one needs to have at least three categories, C, T1 and T2. Here, T1 and T2 would represent two different levels of potency for the medicine or the two different amounts of fertilizer.

As the variety of treatments to be included in a study widens, the scope for applying newer designs also arises. We, however, confine our discussion on development of designs to understand the causal effect of a treatment having only two categories, T and C. A treatment can be a medicine or a vaccine used to understand its effect on a disease. The study can be designed to compare two groups of persons, one having the medicine or the vaccine (T) and the other consisting of those without it (C).