Describe basic statistical methodology and concepts
Describe how basic statistical methodology may be used in pharmacokinetic and pharmacodynamics study design
Describe how basic statistical methodology may be used in critically evaluating data
Describe how basic statistical methodology may be used to help minimize error, bias, and confounding, and, therefore, promote safe and efficacious drug therapy
Provide examples of how basic statistical methodology may be used for study design and data evaluation
Several types of variables will be discussed throughout this text.1 A random variable is “a variable whose observed values may be considered as outcomes of an experiment and whose values cannot be anticipated with certainty before the experiment is conducted” (Herring, 2014). An independent variable is defined as the “intervention or what is being manipulated” in a study (eg, the drug or dose of the drug being evaluated) (Herring, 2014). “The number of independent variables determines the category of statistical methods that are appropriate to use” (Herring, 2014). A dependent variable is the “outcome of interest within a study.” In bioavailability and bioequivalence studies, examples include the maximum concentration of the drug in the circulation, the time to reach that maximum level, and the area under the curve (AUC) of drug level-versus-time curve. These are “the outcomes that one intends to explain or estimate” (Herring, 2014). There may be multiple dependent (aka outcome) variables. For example, in a study determining the half-life, clearance, and plasma protein binding of a new drug following an oral dose, the independent variable is the oral dose of the new drug. The dependent variables are the half-life, clearance, and plasma protein binding of the drug because these variables “depend upon” the oral dose given.
Discrete variables are also known as counting or nonparametric variables (Glasner, 1995). Continuous variables are also known as measuring or parametric variables (Glasner, 1995). We will explore this further in the next section.
TYPES OF DATA (NONPARAMETRIC VERSUS PARAMETRIC)
There are two types of nonparametric data, nominal and ordinal. For nominal data, numbers are purely arbitrary or without regard to any order or ranking of severity (Gaddis and Gaddis, 1990a; Glasner, 1995). Nominal data may be either dichotomous or categorical. Dichotomous (aka binary) nominal data evaluate yes/no questions. For example, patients lived or died, were hospitalized, or were not hospitalized. Examples of categorical nominal data would be things like tablet color or blood type; there is no order or inherent value for nominal, categorical data.
Ordinal data are also nonparametric and categorical, but unlike nominal data, ordinal data are scored on a continuum, without a consistent level of magnitude ...