ABSTRACT
Exploratory factor analysis (EFA) is the name given to a range of statistical techniques to evaluate the dimensionality of items in a questionnaire. EFA explores and uncovers the smallest number of underlying constructs in a questionnaire. A latent variable is inferred from observed variables in the data set. A numerical addition of Likert scores in the questionnaire gives the researcher a rank order. Factor analysis is used to uncover the underlying constructs and identify associated items. Principal factor analysis has the advantage of having no distribution assumptions. When performing EFA the communalities output is generated and provides information about the shared and proportional variance between items. Kaiser-Meyer-Olkin is generally considered to be the best measure of sampling adequacy for carrying out factor analysis. Rotation of factors in factor analysis is an important mathematical procedure in order to help to produce a solution that clarifies its interpretation. Factor analysis can also be used as a data reduction technique.
