In an example in an earlier chapter, the “risk” of benign prostate hyperplasia was considered to depend on age, self-reported sexual activity, and the type of diet. In another example, the birth weight of a child was considered to be dependent on the mother’s weight and the father’s weight. In both these examples, several variables were considered together, but only one was considered a dependent variable. The others were predictors. But there are other situations where the dependent variable is a multifactorial entity in the sense that it is based on several measurements. Thyroid function is evaluated by simultaneous consideration of triiodothyronine (free T3), thyroxine (free T4), and thyroid-stimulating hormone (TSH). These might be dependent on age, diet, exercise, stress, and other factors. In another setup, the outcome of a treatment is evaluated not just in terms of the extent of recovery but also in terms of the time taken in recovery, the nature of side effects, the magnitude of discomfort experienced by the patient, the convenience in administration of the regimen, and so forth. The predictors, in this case, could be treatment regimen, care provided, cooperation of the patient, severity of the condition at the time of admission, and such other considerations. An analysis of the extent of recovery alone as a dependent variable, ignoring the other aspects of outcome, or on each aspect separately, would not provide a holistic view. All aspects of outcome should be considered together for a total picture. Similarly, if severity of disease is an outcome of interest, this could be measured in terms of intensity and magnitude of complaints, extent of abnormality in laboratory and radiological evaluations, extent of disability, etc. All of them should be considered together. Thus, severity too is intrinsically a multivariate entity. It can be considered to depend on baseline physiological measurements of the patient, magnitude of infection, injury or toxicity, personality traits, and such other factors.