TPC Journal V8, Issue 4- FULL ISSUE

The Professional Counselor | Volume 8, Issue 4 319 Lastly, average total WAI-S scores for clients were M = 64.63 and SD = 8.0, and CITs’ scores were M = 59.40 and SD = 7.61. A Pearson product two-tailed correlation identified four significant relationships between the variables with effect sizes ranging from small to large (Cohen, 1992). Positive relationships were indicated between clients’ perceptions of CITs’ multicultural competence and the working alliance ( r =.571, p <.05), as well as CITs’ perceptions of their multicultural competence and the working alliance ( r = .623, p < .05), and between the OQ 45.2 pre- and posttest scores ( r = .884, p < .05). Further, a positive relationship was found between clients’ and counselors’ perceptions of the working alliance ( r = .199, p < .05) and between social desirability scores on CITs’ CCCI-R responses ( r = .233, p < .05); however, the effect sizes were small. The positive relationships indicate that the direction of one construct is associated with the direction of the other. For example, how a client rates their CIT’s multicultural competence is associated with the strength (high or low) of the working alliance. Lastly, a negative relationship was found between clients’ social desirability scores with both client outcome OQ 45.2 pretest scores ( r = -.233, p < .05) and OQ 45.2 posttest scores ( r = -.277, p < .05). This negative relationship means that higher scores on one instrument are associated with lower scores on another. Predictors of Client Outcomes In order to assess whether multicultural competence or the working alliance predicted client outcomes, the third-session OQ 45.2 posttest score was the dependent variable and the pretest score of the OQ 45.2 was the control variable. A hierarchical regression is used when the researcher has a theoretical basis to specify the order in which the independent variables are entered into the model (Tabachnick & Fidell, 2013). In the following analyses, social desirability and OQ 45.2 first-session scores were used as control variables. It is common practice within social sciences to use pretest scores as a control variable and posttest scores as a dependent measure in order to reduce error variance and create more powerful tests for data analysis (Tabachnick & Fidell, 2013). Also, social desirability was used as a control variable because of the relationships indicated in the correlation analysis with SDS, OQ 45.2, and CITs’ CCCI-R responses. Further, SDS scores were used as a control variable to minimize potential threat to the study (Drisko, 2013), which can improve the accuracy of the research design (McKibben & Silvia, 2016), because self-report measures have been shown to have a strong likelihood of participants responding in a socially desirable manner (DeVellis, 2003; Gall, Gall, & Borg, 2007). Hierarchical multiple regression analysis was used to explore whether CITs’ multicultural competence (CCCI-R) and working alliance (WAI-S; as perceived by clients) predicted client outcome (OQ 45.2 pretest), while controlling for social desirability (SDS) from clients’ perspective and clients’ outcome pretest scores (OQ 45.2 posttest). Client outcome OQ 45.2 pretest scores and SDS scores were entered in the first block, explaining 78.6% [ F (2, 116) = 213.3, p < .05] of the variance in client outcome OQ 45.2 posttest scores. After entry of clients’ CCCI-R and WAI-S total scores in the second block, the total variance explained by the model as a whole was 78.9%, [ F (4, 114) = 106.80 p < .05]. The introduction of clients’ CCCI-R and WAI-S scores only explained an additional variance of 0.3%, after controlling for client pretest scores and social desirability [ R 2 change = .003, F (2, 114) = .851, p > .05]. In the final model, only one of the four predictor variables was statistically significant, client outcome pretest score ( b = .859, p < .05; see Table 1). The final model indicated a large effect size ( R 2 = .789; Cohen, 1992). Close to 79% of the variance in posttest scores was accounted for by OQ 45.2 first-session scores on client outcomes, after controlling for social desirability response.