The lower the chi-square value gained from the accumulation of these residuals, the more
likely the theory represents the reality.
However, such tests are sensitive to sample sizes, and the probability of rejecting
any model increases when a sample size increases even with minimal deviation from the
model. Therefore, Bentler and Bonnett (1980) suggested that X2/df ratio (df. degree of
freedom) as a more appropriate measure of model fit. This ratio should not be bigger than
5 for models with a good fit (Bentler, 1990). In addition to the chi-square statistic, the
root mean square error of approximation (RMSEA) and the non-normed fit index (NNFI)
are used for evaluation for the significance of the relations between the constructs of all
models. According to Bentler and Bonett (1980), these two indices were determined to be
robust with relatively small sample sizes and non-normal or skewed distributional form
(see also, Browne & Cudeck, 1993; Neuhaus & Swank, 2002). The second step in model
estimation is to examine the path significance of each association in our research model
and variance explained (R2 value) by each path. The LISREL reports raw and
standardized estimates for all specified paths, along with standardized errors and test
statistics for each path.