Abstract
This article reports on differences between expert and aspiring principals. Following the work of Leithwood and colleagues, we asked expert and aspiring principals to respond to ill-structured written problem scenarios. Our sample of 44 included 20 expert principals and 24 aspiring principals. The aspiring principals were from a cohort of participants in a professional development program for aspiring principals in a large Midwestern urban school district. Drawing from that same school district, the expert principals were selected based on an analysis of longitudinal teacher survey data and student achievement data. We found statistically significant differences between expert and aspiring principals on five problem-solving processes, three of which were in line with prior hypotheses generating research on principal expertise.
Notes
1. The rate is the mean of each problem solving process across observations, which is equivalent to the percentage of observations coded as 1.
2. Since power is a concern in this analysis, we considered alternative adjustments such as the Šidák-Holm adjustment, which is less conservative than the Bonferroni (CitationWestfall & Young 1993), and adjustments using resampling techniques that account for both dependence among dependent variables and the binary distribution of the data. There were no qualitative differences, however, among these adjustments. We chose to present the Bonferroni adjustment because it is the most well-known and simple.
3. Discussion results are based on columns (1)–(4) of , which present unadjusted mean differences and p-values from chi-square tests.
4. We thank Ken Leithwood for bringing this interpretation to our attention.