The top portion of the Structural Equation Modeling Results dialog box contains a Summary box with a variety of statistics. You can copy these results to the Clipboard by clicking the button. To reduce or enlarge the Summary box, click the or button, respectively.

Left Side. The
information on the left side of the summary box is designed to enable
you to evaluate, quickly and efficiently, whether iteration was successful.

Method of Estimation. At the top of the window is the Method of Estimation (i.e., the discrepancy function used).

Discrepancy Function Value. This is the numerical value of the discrepancy function.

Maximum Residual Cosine. This numerical criterion should be close to zero if iteration was successful.

Maximum Absolute Gradient. This is the absolute value of the largest element of the gradient.

Maximum Absolute Constraint. In options requiring constrained estimation, i.e., when correlations are analyzed, or when the New Standardization (see Analysis Parameters for details) method is employed, a number of constraint functions are evaluated, and must iterate to zero if the constraints are maintained. If this value is not zero, or a very small value, it indicates iteration was unsuccessful.

ICSF Criterion. This criterion
should be close to zero if the structural model is invariant under a constant
scaling factor (ICSF). Most, but not all, structural models are invariant
under a constant scaling factor.

Boundary Conditions. This criterion indicates the number of inequality constraints that were operative at convergence. This number should be zero, unless there is a boundary case such as a Heywood Case in your model. If this number is not zero, then the Chi-square statistic will not necessarily have the proper distribution.

Right Side. The information on the right side of the window is basic statistical information about the fit of the model.

Chi-square Statistic. Available for all
discrepancy functions except OLS, this statistic has an asymptotic Chi-square distribution if the null
hypothesis of perfect fit is true.

Degrees of Freedom. This is the number of degrees of freedom for the Chi-square statistic.

Chi-square p-value. This is the probability level for the Chi-square statistic.

Steiger-Lind RMSEA. Here the point estimate and 90% confidence interval for the Steiger-Lind RMSEA are given.

RMS Standardized Residual. This is the root mean square standardized residual. Roughly, this number must be less than .05 for the fit to be "good" in a practical sense.