Factor analyses were conducted on raw scores for all participant
responses on the MSI (N = 252). Although an initial factor analysis of the
data produced eight factors with eigenvalues greater than 1, parallel analysis
suggested that a maximum of six factors be retained. The six-factor solution
produced a nonsensible result, with a large number of cross-loadings. In
addition, one of the factors contained only two items (a minimum of three
items is necessary for a factor). As a result of these initial findings, and
because the MSI’s original developers intended it to consist of five factors, a
confirmatory factor analysis was re-run, specifying five factors and using
varimax rotation, as Barbuto and Scholl (1998) reported the factors of the
MSI to be orthogonal