Describe path analysis and its relationship to SEM

What will be an ideal response?

Path analysis (PA) can be viewed as a special case of structural equation modeling (SEM). We could think of PA as SEM with only single indicators for each of the variables in the causal model. In other words, path analysis is SEM with a structural model, but no measurement model. Path analysis may also be viewed as an extension of the regression model. The PA model is depicted in a rectangle-and-arrow figure in which single-headed arrows indicate causation. A regression is done for each variable in the model as a dependent on others, which the model indicates are causes. The regression weights estimated by the model are compared with the observed correlation matrix for the variables, and a goodness-of-fit statistic is calculated. Path analysis calculates the strength of each relationship using only a correlation or covariance matrix as input.