Michael E. Tarter
Data transformations modify measured values systematically. Epidemiologists often transform measured values so that they conform more closely to a distribution germane to a statistical method that the epidemiologist would like to apply. For instance, many common statistical techniques assume that a data set consists of approximately normally distributed values, and if this is not the case, a transformation may be applied to the data before analysis. For example, measured data might be used to calculate the heart rate (HR) ratio variate, HRR = (HR work − HR rest)/(HR predicted maximum − HR rest), then this data transformed to the new variate arcsin(√ HRR p). In terms of the matchup between, on the one hand, the statistical methodology applied to study arcsin(√ HRR p) and, on the other, the assumptions that underlie this methodology, a variate such as arcsin(√ HRR p) is often a preferred transform of a variate such HRR ...