Nonparametric Statistics

Marjorie A. Pett

Some of the most popular statistical inferential techniques in epidemiological research are those that focus on specific parameters of the population such as the mean and variance. These parametric statistics share a number of common assumptions: There is independence of observations except when data are paired. The set of observations for the outcome (i.e., dependent) variable of interest has been randomly drawn from a normally distributed or bell-shaped population of values. The dependent variable is measured on at least an interval-level scale of measurement (i.e., it is rank ordered and has equidistant numbers that share similar meaning). The data are drawn from populations having equal variances or spread of scores. Hypotheses are formulated about parameters in the population, especially the mean. Additional requirements include nominalor intervallevel independent variables, homoscedasticity, and equal cell sizes of at least 30 observations per group. Examples of commonly used parametric statistical tests include the independent ...

Users without subscription are not able to see the full content on this title. Please, subscribe or login to access all content on this website.