The Professional Counselor | Volume 13, Issue 3 230 Round 2 Participants (n = 19) completed the CAPS pretest in January 2022 and then were assigned to the treatment group (n = 9) or comparison group (n = 10). The treatment groups started in February 2022 and met weekly for 8 weeks. The last group session occurred in April 2022, and all participants completed the CAPS posttest. Participants completed the final posttest data (DASS-21 and SEHS-S) in May 2022 as a part of the school’s counseling program. Preliminary Analysis Researchers use an analysis of covariance (ANCOVA) when they want to compare two or more groups on one dependent variable at the same time (Heppner & Heppner, 2004). To determine if it is appropriate to use a one-way ANCOVA, the data needs to meet the necessary statistical assumptions. First, Bendit and Paul Peluso determined that the dependent and covariate variables were all measured on a continuous scale and the independent variable consisted of two or more independent groups. Next, Bendit and Peluso determined that independence of observation was met, with the treatment and comparison groups having different participants in each group with no participant being in more than one group. There were two outliers detected (DASS-21 pre and CAPS pre), but we opted to keep these values included. Next, we used Shapiro-Wilk tests to determine that the data met normal distribution. Homogeneity of variances was assessed by a visual inspection of a scatterplot, and there was a linear relationship between pre and post for each intervention type. Data Analysis To address the research questions in this study, an alpha level of .05 was set and a series of one-way ANCOVA was used to analyze the effect of the independent variable (CBT-P group; Shafran et al., 2002) on the dependent variables (perfectionism, negative affectivity, social–emotional well-being) between the treatment and comparison groups. Results were reported using the F statistic and associated p-value (alpha .05), indicating statistical significance (Heppner & Heppner, 2004). Effect size was also calculated using the partial eta-square statistic (ηp 2), with benchmarks set forth to determine small (.01), medium (.06), and large (.14) effect size strength (Maher et al., 2013). We did not include participants’ data in the analysis if they did not attend a minimum of four group sessions or if survey data was missing at posttest. Results The treatment and comparison groups’ means, standard deviations, and change scores for the study variables at pretest and posttest are provided in Table 4. To control for differences prior to treatment, the participants’ pretest scores were used as covariates and group as a factor. The first research question explored the impact of the CBT-P group intervention on levels of perfectionism. Results from the one-way ANCOVA revealed a statistically significant difference [F(1, 38) = 4.94, p = .03; ηp 2 of .11] in Self-Oriented Perfectionism between treatment and comparison groups and no statistically significant difference [F(1, 38) = .04, p = .83; ηp 2 of .00] in Socially Prescribed Perfectionism between treatment and comparison groups. This significant medium effect finding revealed that using the modified CBT-P small group intervention (Shafran et al., 2002) can positively influence levels of self-oriented perfectionism among early college high school students. Participants in the treatment group (M = 39.15, SD = 5.78) reported lower scores for Self-Oriented Perfectionism at posttest as opposed to participants in the comparison group (M = 43.20, SD = 6.50), a mean difference of −4.05. The second research question assessed if the students who received the CBT-P group intervention would show a decrease in negative affectivity compared to students in the comparison group who did not receive the intervention. Results from the ANCOVA revealed a statistically significant difference [F(1, 37) = 10.35, p = .003; ηp 2 of .22, a large effect] between treatment and
RkJQdWJsaXNoZXIy NDU5MTM1