The purpose of this study was to investigate the psychometric properties of the LPI and determine its reliability and validity with a population of potential educational leaders.
Participants were graduate students enrolled in the Innovative Leadership Program at The University of Alabama enrolled in the Summer, 2001 (N = 105). Respondents voluntarily completed the LPI-Self and provided demographic information. The response rate was 90 percent (N = 94) and most of the respondents were between 31-40 years of age (51%), Caucasian (79%; with 18% African-American), and female (55%).
Analysis by the author concludes: “The items appear to provide evidence that supports the sub-dimensions related to leadership, as purported by the creators of the instrument” (p. 90). The overall Cronbach alpha reliability coefficient of .91 was “strong, indicating evidence of reliability for the instrument” (p. 91). The Cronbach alpha reliability coefficient for Modeling was .60, Inspiring was .80, Challenging was .73, Enabling was .67 and Encouraging was .79. “Together, the overall reliability and the item-to-total correlation plus the item-to-dimension total correlations provide evidence that the LPI has internal consistency, thus supporting the theoretical construct” (p. 91).
The author notes that the distribution of scores from the sample population was similar to those of the normative population, and this “finding provides some evidence of construct validity” (p. 96). The most frequently engaged leadership practice was Modeling, followed by Encouraging, Enabling, Challenging and Inspiring. The average scores from this sample were higher than the average scores of the normative database, which the author suggests may be due to the high leadership propensity of the sample population.
One-way ANOVAs were computed for the six demographic variables and the five leadership practices. Of the possible 30 relationships, only seven reached statistical significance. Females scored higher on Challenging than their male counterparts but not differently on the other four leadership practices. More credentialed teachers reported engaging significantly more in Challenging and Enabling than their less credentialed counterparts. Teachers in Middle School rated themselves more highly engaged in Challenging, Inspiring, Modeling and Encouraging than did their counterparts in Elementary Education or in High Schools. No statistically significant differences were found among the subgroups by age, race, or years of experience teaching.
“The findings and conclusions of this study regarding the
psychometric properties of the LPI did suggest some support for the
belief that the LPI-Self report was a reliable and adequately valid
instrument when used to measure the self-perceived leadership practices
of graduate students in the educational leadership program at the
University of Alabama” (p. 123).
Download a Printer-Friendly Version of the Abstract (PDF)