The OLA possesses strong psychometric properties and can be trusted to accurately measure the characteristics of the healthy servant-minded organization as perceived by top leaders, managers and the workforce.
Construct Validity was determined by the use of an expert panel to determine the Necessary and Essential characteristics of servant leadership that became the 60 items used within the instrument. A Delphi process was utilized to bring these experts to a clear consensus on the constructs that represent the servant-minded organization.
Reliability – The OLA shows high reliability. In the original field test, the OLA obtained a reliability score of .9802 using the Cronbach-Alpha coefficient. Horsman (2001), Thompson (2002) and Ledbetter (2003) also conducted reliability tests on the OLA showing scores equal or higher verifying OLA reliability. Miears (2004) study revealed equally high reliability scores for the Educational version of the OLA.
[Insert Data Table Here]
Item Analysis – In the original field test of the OLA (Laub, 1999), the lowest item-to-item correlation was .41 and the highest was .77, showing that all of the items have a strong correlation with the instrument as a whole. Ledbetter’s study (2003) found that the lowest item-to-item correlation was .44 and the highest was .78
Test – Retest – Ledbetter (2003) conducted a test-retest study on the OLA showing “the means and standard deviations between the Test and the Retest for this study remained consistent.” “The correlation between the Test and Retest were significant and the findings indicate that the validity of the OLA remains consistent over time.” Both the Test and Retest were significant at p<.01
Face Validity tests were run on the perceived accuracy of the six organizational descriptions utilizing over 100 adult graduate students. There was a consistently high perception of the accuracy across all six of the extended full-page descriptions. The descriptions are therefore, seen as accurate in describing, on average, the various organizational levels. This also served to confirm that the scoring break-points for the six organizational levels were place properly.
The Job Satisfaction Scale obtained an estimated reliability, using the Cronbach-Alpha coefficient, of .81. A correlation of Job Satisfaction to the OLA scores was run utilizing a Pearson correlation and a significant (p<.01) positive correlation of .635 existed, accounting for 40% of the variance in the total instrument score. Horsman (2001) and Thompson (2002) also found this strong correlation between the OLA score and the Job Satisfaction scale. A validity study was conducted and validity established by Thompson (2002) on the Job Satisfaction scale.