TPC Journal-Vol 9 Issue 3-FULL

232 The Professional Counselor | Volume 9, Issue 3 Results CFA In our sample, the CFA fit indices for the single-factor model yielded a poor model fit for single- factor solution. Specifically, although non-significance is desirable (Tabachnick & Fidell, 2013), we obtained a significant result for the chi-square test of model fit. This test is known as sensitive to sample size (Lent et al., 1996), and to account for this, we continued with examining other criteria for our model fit. We also obtained initial values for RMSEA (.12) and CFI (.90) outside the recommended criteria for fit (RMSEA < .06, CFI > .95; Hu & Bentler, 1999). The SRMSR was the only index meeting the recommended fit criteria (.04 < .08; Hu & Bentler, 1999). After this initial review, we followed Cole and Maxwell’s (2003) recommendations on examining modification indices. As a result of conducting necessary modifications, we removed nine items from the CSS, and the CFA results revealed a better fit for an 8-item version of the instrument the authors called the Consultation Skills Scale-Short Form (CSS-S; χ 2 (28) = 86.21, p = .00, CFI = .98, RMSEA = .075, 90% CI [.06, .90], SRMSR = .02). See Table 2 for means, standard deviations, and factor loadings of the eight items. Table 2 Means, Standard Deviations, and Factor Loadings of the Items of the CSS-Short Form Item # Item labels M a SD Factor Loadings 1 I know how to help consultees improve programming issues for work with identified clients. 4.31 1.63 .86 2 I know how to develop a consultation contract. 3.71 2.21 .83 3 I know how to apply established problem-solving models to address consultee concerns. 4.23 1.76 .89 4 I am familiar with systems consultation. 3.77 2.10 .87 5 I know how to operate as an external consultant. 3.97 2.05 .87 6 I know how to operate as an internal consultant. 4.16 1.85 .87 7 I know how to assess the culture and climate of consultee organizations. 3.97 1.92 .85 8 I can address theme interference effectively with consultees. 3.88 1.98 .85 a Means are based on a scale of 1 to 5. Validity Analysis We obtained initial evidence for the construct validity of the CSS-S through our CFA results. Convergent validity of the CSS-S was established through the obtained significant correlation coefficients between the CSS-S and SWAI-CF ( r = .50), SWAI-R ( r = .46), and SWAI-I ( r = .46). Indicating divergent validity for the CSS-S, the correlation coefficients between the CSS and the MIIC, gender, and ethnicity were .34, .05, and -.03, respectively. The results of a one-way ANOVA indicated concurrent validity for the CSS-S with significant differences between the three groups of participants without consultation training, those with one to two consultation training experiences, and those with three or more consultation training experiences: [ F (2,368) = 28.27, p = .00]. Participants with three or more