Days of their lives: a mixed‐methods, descriptive analysis of the men and women at work in the principal’s office

Abstract This study examines the work of US school principals from the perspective of their workday using a distributed perspective to frame the investigation. Using data on 38 school principals in one mid‐sized urban school district in the US, it describes school principals' work practices, examining both the focus of that work and how it is accomplished. Cluster analysis is used to analyse data from an experience sampling method (ESM) log, identifying three patterns of practice: administration‐centred, solo practitioners, and people‐centred. To explicate these patterns, qualitative interview and observation data were combined with quantitative survey and log data to construct mini‐cases of three principals, each representing one of the three patterns of practice. The study concludes with a discussion of how this analysis both confirms and challenges popular portrayals of the school principal in the literature. Keywords: instructional leadershipmixed methodsprincipalsschool administrationschool leadership Acknowledgements Work on this paper was supported by the US Institute for Education Sciences (Grant # R305E040085) and the Distributed Leadership Studies funded by the National Science Foundation (RETA Grant # HER—0412510). We are grateful to our colleagues on the study for their help with data collection and data analysis: Carol Barnes, Eric M. Camburn, Lisa Dorner, Ellen Goldring, Jonathon Supovitz, Jason Huff, Henry May, Beth Sanders, James Sebastian, James Pustejovsky, and Amber Stitziel‐Pareja. All opinions and conclusions expressed in this paper are those of the authors and do not necessarily reflect the views of any of the funding agencies. We are grateful to Henry May and Jelani Mandara for comments on an earlier draft of this paper. Notes 1. See, e.g. Bryk and Driscoll (Citation1985), Newman and Wehlage (Citation1995), and Rosenholtz (Citation1989). 2. See, e.g. Coelli et al. (Citation2007), Hallinger and Heck (Citation1996), Leithwood et al. (Citation2007), and Sheppard (Citation1996). 3. All names used in this paper are pseudonyms. 4. While these authors used data generated by the EOD log, our paper uses data generated by the ESM log. While both the EOD and ESM log collected data on the type and duration of activities, the ESM log also gathers data on other dimensions of practice, including whether the principal takes a leadership role in activities, whether they lead alone or co‐lead, and with whom they co‐lead. 5. According to the ecological validity perspective, outcomes (behaviours, beliefs, emotional responses, etc.) are assumed to respond to environmental stimuli. Instruments are considered ecologically valid to the extent that they capture a representative sample of stimuli and subsequent responses existing in an environment. 6. We chose to partition the data using Ward's method, rather than a random partition, because the random method may result in distorted partitioning (Aldenderfer and Blashfield Citation1984). We selected the Squared Euclidean distance as the measure of dissimilarity at this step as it is the most common and recommended measure (Mandara Citation2003). 7. The pseudo‐T‐squared values, which are presented with the Duda and Hart Je(2)/Je(1) index, also indicate a 3‐cluster solution in that smaller values indicate more distinct clustering. 8. The k‐means procedure attempts to minimize the distance between cases within each cluster and maximize the distance between clusters.