More specifically, statistical significance decisions did not differ for the final measurement and structural models reported except for one structural path in the final model reported (between comfort discussing nonpharmacy approaches and behavior) becoming marginally significant when measurement errors were allowed to correlate.As can be seen in Table 1 , all loadings on the presumed factors were statistically significant and as can be seen in Table 2 , all measures of internal consistency were acceptable (α ≥ .77, ω Raykov ≥ .78).
1
—
—