Online Program

Variability, Redundancy, and Reduction Using Principal Components Analysis (302998)

Keywords: Principal Components Analysis, Data Reduction, Item Reduction

Exploring variability and redundancy in a set of variables is helpful to practicing statisticians and academic researchers when identifying the extent to which a larger set of observed variables may be reduced into a smaller set, called principle components (PCs). A principle components analysis (PCA) is a practical and appropriate statistical methodology to reduce the number of items in a questionnaire, for example, when the items measure the same or nearly the same construct. Several methods may be used within the PCA procedure to identify principle components for a given set of variables. Each of these methods is illustrated using a large data set with a 21-item questionnaire. Components derived from a PCA may be used in fur ther analyses. A major drawback of PCA is the requirement of a large data set to yield trustworthy results. However, the exploratory nature of the procedure permits applications to many social, behavioral, business, and industrial environments.