0.9 Koonce, g., & Kelly, m. (october 2011). effects of utilizing  (Page 5/7)

Figure 1. three-way principal internship assessment

A Repeated Measures Analysis Of Variance (ANOVA) was performed to examine the null-hypothesis of no difference between the means (Lund, 2010a) for the three internship assessments under each of the six ISLLC standards. Prior to conducting each Repeated Measures ANOVA, researchers first determined sphericity using Mauchly's Test of Sphericity. Sphericity is the condition where the covariances between all levels of the within subjects variable are equal (Lund, 2010b). Because ANOVAs with repeated measures are particularly susceptible to the violation of the assumption of sphericity with violation causing the test to become too liberal (i.e., an increase in the Type I error rate), sphericity must be tested in order to determine the proper repeated measures test (i.e., univariate vs. multivariate; Lund, 2010b).

When the statistical analysis determined that sphericity could be assumed, a univariate test was used to analyze the within subjects effect; if sphericity could not be assumed, a multivariate test using Wilks’ Lambda was used. If the appropriate within subjects test revealed a difference significant at the 0.05 level, paired samples t- tests were used as a post hoc analysis. To reduce the possibility of a Type I error, the Bonferroni correction to α was implemented (Simon, 2008).

Findings

Through the statistical analysis of the data, it was determined that there were no significant differences between the means in three of the six ISLLC standards with regard to the assessments completed by the mentors, university supervisors, and the students themselves. For these three standards (ISLLC 1, 3 and 6), the researchers failed to reject the null hypothesis that there will be no significant difference in the means of three assessments. Table 1 provides the descriptive statistics for all responses to items under ISLLC standards 1-6.

General linear model comparing responses for isllc standard 2

In the analysis of ISLLC Standard 2 responses, results of Mauchly’s Test of Sphericity on the covariance matrix indicated a significant difference of less than 0.05 ( p = .000). As such, sphericity could not be assumed for this standard and multivariate testing was again needed. The Wilks’ Lambda multivariate test revealed a level of significance less than 0.05 (p = .001). As such the null hypothesis was rejected and paired t- tests were then used as a post hoc analysis.