A general approach to testing multifaceted personality constructs

Some personality characteristics are composed of multiple, distinct subcomponents (e.g., Type A, hardiness, attributional style, self-monitoring). The advantages and disadvantages of 3 typical approaches to testing these constructs are reviewed. An alternative approach based on structural equation modeling is then offered. This approach has many advantages over its alternatives, including the provision of an explicit test of the structure of the multifaceted construct, the simultaneous test of the effects of this general construct and the unique aspects of its subcomponents, and the explicit consideration of measurement error. Although the modeling approach does have limitations, these limitations are equally applicable to all of its alternatives. Indeed, the principal disadvantages of the modeling approach seem to be its statistical complexity and the lack of education regarding its proper use.