Over the past 15 years, the use of structural equation modeling has become increasingly common in the social and behavioral sciences. Enthusiastic recognition by researchers of the advantages of the structural equation modeling approach and an eagerness to implement this potentially powerful methodology has also brought with it inappropriate use of the technique. One major source of inappropriate usage has been the failure of investigators to satisfy the scaling and normality assumptions upon which estimation and testing are based. The commonly used approaches to estimating the parameters of structural equation models, maximum likelihood and normal theory generalized least squares, assume that the measured variables are continuous and have a multivariate normal distribution. In practice, current applications of the structural equation modeling approach to real data often involve violations of these assumptions.