A summary of the featuresof the twenty papers is shown in Table 2. The articles reviewed are classified according to data type (for example. cross sectional versus time -series), SEM modeling strategy, estimation technique used, assessment of normality, sample size, reporting of a path diagram and reported fit indices.