Factor Loading

Factor loading, also known as factor-variable correlation, is a statistical technique used to examine the relationship between an observable variable and an unobservable (also called underlying or latent) variable called factor. The factor loading of a variable provides information about the extent to which that variable is related to an underlying factor and how well it represents that factor.

Loading values are interpreted similarly to correlation coefficients between variables: loading values can range from -1 to +1, with values closer to -1 or +1 indicating stronger relationships between variables. In other words, values closer to -1 and +1 indicate that the factor explains most of the variable’s variance. If the value is closer to 0, the factor exerts a low influence on the variable.