TPC Journal-Vol 10- Issue 3-FULL ISSUE

The Professional Counselor | Volume 10, Issue 3 383 Supports and Practices factor was: “ Collaborating with the school’s IEP team to implement student’s individualized education programs . ” This four-factor model served as our preferred model, but competing models were explored using CFA on the first subset from the sample. CFA Using First Subset Sample The competing models were examined to determine the best data–model fit by conducting a CFA using MPlus. The following models were tested: (a) one-factor model, (b) four-factor model, and (c) second-order four-factor model. Model modifications were allowed during the exploratory phases. The results of the CFA are reported in Table 3. Table 3 Results of the Confirmatory Factor Analyses for the Exploratory Phase Competing Models Chi-square df SRMR RMSEA 90% CI, RMSEA TLI CFI AIC Exploratory Analyses 1 One-Factor (initial) 8,518.75 495 .057 .084 [.083, .086] 0.79 .80 168,279.20 One-Factor (modification) a 4,465.60 478 .048 .060 [.059, .062] 0.89 .90 162,654.94 2 Four-Factor (initial) 5,619.01 489 .058 .068 [.066, .069] 0.86 .87 164,253.62 Four-Factor (modification) b 3,866.27 481 .048 .055 [.054, .057] 0.91 .92 161,407.48 3 Four-Factor second order 5,632.53 491 .058 .068 [.066, .069] 0.86 .87 164,270.54 Four-Factor second order (modified) b 3,866.27 483 .048 .055 [.054, .057] 0.91 .91 161,432.63 Confirmatory Analysis Four-Factor second order 5,424.82 490 0.058 0.066 [.065, .068] 0.87 .88 164,999.81 Four-Factor second order (modified) b 4,468.62 483 0.051 0.060 [.058, .062] 0.89 .90 163,723.53 Note. All chi-square tests were statistically significant at < .001. a Seventeen correlated error variances were estimated. b Eight correlated error variances were estimated. The initial one-factor model did not fit the data (chi-square = 8,518.75, df = 495, p < .001; RMSEA = .084, 90% CI [.083, .086]; CFI = .80; SRMR = .057), but after modification (i.e., 17 correlated error variances between observed variables), the one-factor model had an adequate fit (chi-square = 4,465.60, df = 478, p < .001; RMSEA = .060, 90% CI [.059, .062]; CFI = .90; SRMR = .048). Reasonable data–model fit was obtained for the modified models in both the four-factor and four-factor second-order models (see Table 3). Modifications included freeing eight correlated error variances between observed variables. A content expert reviewed the suggested modification to determine the appropriateness of allowing