TPC Journal-V6, Issue 4- FULL ISSUE

The Professional Counselor /Volume 6, Issue 4 349 academic issues.” Some sample items from the curriculum subscale are “Conduct classroom lessons addressing career development and the world of work” and “Conduct classroom lessons on conflict resolution.” Scarborough (2005) examined the validity by investigating the variances in score on the actual scale based on participant grade level and found that participants’ grade level had a statistically significant effect across the scales with small to large effect sizes (e.g., ranging from .11 to .68[ω 2 ]), which supported the convergent validity of the SCARS. Additionally, construct validity was supported using factor analysis. In prior research using the SCARS, the internal consistency of the counseling and curriculum scales was strong with Cronbach’s alphas of .93 for the curriculum actual scale and .85 for the counseling actual scale (Scarborough, 2005). The internal consistency coefficients of the SCARS actual subscales in this investigation were good (Streiner, 2003) with Cronbach’s alphas of .77 for the counseling scale and .93 for the curriculum scale. PSS . The PSS (Cohen et al., 1983) is a 10-item self-report measure that examines the participants’ appraisal of stress by asking about feelings and thoughts during the past month. The PSS uses a 5-point Likert scale that ranges from 0 ( never ) to 4 ( very often ) and includes four positively stated items that are reverse coded. Some sample items include, “In the last month, how often have you felt that you were on top of things?” (reverse coded), and “In the last month, how often have you been upset because of something that happened unexpectedly?” The PSS has been shown to have acceptable internal consistency with Cronbach’s alphas ranging from .84 to .91 (Chao, 2011; Cohen et al., 1983; Daire, Dominguez, Carlson, & Case-Pease, 2014). The internal consistency coefficient of the PSS in this study also was acceptable (Streiner, 2003) with a Cronbach’s alpha of .88. Results Preliminary Analysis Initial screening of the data included the search for outliers (e.g., data points three or more standard deviations from the mean) using converted z-scores (Osborne, 2012), which resulted in identifying 21 cases that had at least one variable with an extreme outlier. To accommodate for these outliers, the researchers utilized a Windorized mean based on adjacent data points (Barnett & Lewis, 1994; Osborne & Overbay, 2004). Next, the assumptions associated with structural equation modeling (SEM) were tested (e.g., normality and multicollinearity; Hair, Black, Babin, Anderson, & Tatham, 2006; Tabachnick & Fidell, 2007). Multicollinearity was not present with these data; however, the data violated the assumption of normality of a single composite variable (e.g., devaluing clients scale on the CBI). Researchers conducted descriptive analyses of the data using the statistical software SPSS. Table 1 presents the means, standard deviations and correlations for the study variables. Model Testing This correlational investigation utilized a two-step SEM method (Kline, 2011) to examine the research hypothesis employing AMOS (version 20) software. The first step included a confirmatory factor analysis (CFA) to inspect the measurement model of burnout and its fit with the data. Then, a structural model was developed based on the measurement model. The measurement model and structural model were appraised using model fit indices, standardized residual covariances, standardized factorial loadings and standardized regression estimates (Byrne, 2010; Kline, 2011). Modifications to the models were made as needed (Kline, 2011). Both the measurement and the structural models employed the use of maximum likelihood estimation technique despite the presence of non-normality based on recommendations from the literature (Curran, West, & Finch, 1996; Hu, Bentler, & Kano, 1992; Lei & Lomax 2005; Olsson, Foss, Troye, & Howell, 2000).