|TITLE:||The Effect of Mentoring on the Leadership Practices of Principals in Large Unit Districts in Illinois|
|RESEARCHER:||Susan B. Wolfe
Department of Leadership, Educational Psychology and Foundations
Northern Illinois University
Doctoral Dissertation: December 2005
The purpose of this study was to determine whether practicing principals were mentored prior to or during their present assignment and see what difference this had on their leadership practices.
A random sample (N = 180) of the population – consisting of all public school Principals in Illinois districts which are members of the Large Unit District Association (N = 1419) – yielded 106 responses (59% response rate). Males comprised 54 percent of the sample (N = 57) and 44 of them had been mentored. Of the female respondents, 41 had been mentored. About half of the respondents had from 1 to 10 years of administrative experience, 88 percent were Caucasian, and 54 percent were at the elementary school level. Most of those mentored (71%) had been mentored by another school principal. Respondents completed the LPI and provided demographic information. Internal reliability in this study ranged between .69 and .87.
Using MANOVA, and matching the sample for mentored and non-mentored principals (N=19 each) revealed slightly higher scores on each leadership practice for the mentored principals, although this did not reach statistical significance. Eta-squared, the effect size measure used in this study, however, in measuring the magnitude of mentoring on the dependent variables indicated a large effect size on the overall leadership practices (meaning that “overall the two groups differed in level across the subscales” p. 70). The lack of statistical significant, the author suggests, was likely due to the small sample size.
Comparing male mentored principals with female mentored principals revealed no significant differences and gender had a small effect on the overall leadership practices. Years of experience (less than 10 versus over 10 years) for mentored principals did not account for statistically significant variance although the eta-squared statistic indicated that it had a moderate effect on the overall leadership practices. Matching the comparison samples by ethnicity (white or non-white) reduced the effective sample to 13 cases each and failed to reach statistical significance, despite a large eta-squared. Profile plot of LPI subscale means revealed more frequent use by non-Caucasian mentored principals than Caucasian mentored principals.