Leadership can be understood as a process of influence that drives, guides and unifies collective action. Typically, this influence is regular and repeated with the understanding that “any member of a community exerts influence from time to time, but some influence the course of events more often than others” (Gather Thurler, 2000, p.151). This conceptualization that leadership is not an attribute of an individual but more of a process has also been highlighted by organizational sociologists such as Crozier & Friedberg (1979) and Friedberg (1993) drawing on the concept of power.

A common criticism of educational leadership research is the relative lack of methodological variation (Showunmi & Fox, 2018). Research methods for educational leadership tend to be comprised in large part of survey and interview approaches. Observational approaches like shadowing which involve following leaders or leadership actions in situ can be used to gather data that are closer to the processes themselves (Gather Thurler, Kolly Ottiger, Losego & Maulini, 2016; Progin, 2017; Tulowitzki, 2019).

In short, shadowing is a research technique in which the observer follows the observed like a “shadow”, acting as a constant, discrete observer and attempting to gain a better understanding of the observed subject's professional reality (Progin, 2017).

While shadowing has been deemed to have the potential to document the nature and context of professionals in leadership positions such as principals (Pollock & Hauseman, 2017), some have argued that it “needs to be reflected, positioned against similar methods based on a systematization of its use so far and further developed” in order for it to evolve and mature as a research method (Tulowitzki, 2019, p.105). A first step could therefore consist in systematically reflecting shadowing-type approaches in educational leadership research vis-à-vis other, more common methods. A second step could be to analyze actual shadowing practices to explore their advantages but also their limits. Finally, a third step could be to compare the merits and limitations of shadowing with those generally attributed to standardized surveys and semi-structured interviews.

This study which is part of a larger project to make leadership processes visible aims to explore how shadowing is conceptualized in educational leadership research and what kind of data it generated. It sought to answer the following research questions: 

• How does data gathered through shadowing-type observation compare to survey data in terms of proximity to the experience of school leaders? 
• What are methodological constraints and challenges of shadowing-type approaches?
• What are special considerations for using the method for intercultural research (for example for conducting comparative research in two different countries or in two different linguistic regions)? 

We answer the research questions by drawing on existing research and confronting our shadowing experiences in different contexts. The overall design consists of three steps: First, an analysis of studies making use of shadowing-type approaches is used to identify points in common as well as variations to shadowing as well as advantages and limitations specifically associated with this approach. Second, for additional depth and detail, three leadership studies that relied upon shadowing for data collection from the past five years in which we were principal investigators are used as points of systematic reflection. We revisit the methodological parts of the studies with the aim to make underlying assumptions and methodological parameters transparent. We attempt to make explicit the theoretical foundations and related fields of the shadowing approaches (for example ethnography vs. structured observation). Additionally, we look for elements tied to culture in those shadowing approaches, such as the proximity of the researchers to the observed parties. The confrontation of our experiences is recorded and transcribed. These transcriptions are analyzed in a logic of grounded theory (Strauss & Glaser, 2008). Thus, the data is coded by cross-induction based on the identification of regularities and variations present in the material (Maulini, 2013). In other words, we attempt to identify - in a systematic way - the similarities and differences existing in our shadowing approaches. Finally, the insights into shadowing as a method for educational leadership research are posited against merits and limitations commonly attributed to standardized surveys and semi-structured interview approaches.

Compared to surveys or semi-structured interviews, shadowing eliminates the risk of memory bias. Participants do not need to remember certain key situations (recall can thus be considered as a reconstruction process). Instead, the key situations are observed in situ, i.e. at the very moment they take place. However, shadowing is very time-consuming (thus limiting the number of cases studied) and generates a very large amount of data which needs to be managed. Data seems to be very much contextually bound, making intercultural designs challenging. The final results will more specifically address the benefits but also the limitations of such an approach beyond the elements succinctly presented in the paragraph below as well as contexts that might lend themselves more and less to this approach. Moreover, it is important to note that the originality of our contribution resides in the confrontation of our experiences of shadowing within different linguistic and cultural contexts (French-speaking Switzerland, German-speaking Switzerland, France). The differences between linguistic regions and countries exist and it is a matter of identifying them and then analyzing them in order to be able to work with them. Finally, to approach real work in a non-normative approach (which does not immediately assert the best way to do one's work but first tries to understand it as it is actually done) is a major challenge for educational and social science researchers more generally. Much of the research in our field approaches leadership from a best practice perspective. In a complementary way, we believe it is essential to also offer an understanding of these practices as they really are in all their complexity. From this perspective, we consider it important to study the methodological approaches to achieve this, while at the same time having the opportunity to exchange with other researchers on this subject internationally.

Crozier, M., & Friedberg, E. (1979). Macht und Organisation. Königstein/Ts.: Athenäum. Friedberg, E. (1993). Le Pouvoir et la Règle. Dynamiques de l’action organisée. Paris: Seuil. Gather Thurler, M. (2000). Innover au cœur de l’établissement scolaire. Paris: ESF. Gather Thurler, M., Kolly Ottiger, I., Losego, P., & Maulini, O. (2016). Les directeurs au travail. Bern: Peter Lang. Maulini, O. (2013). Penser les pratiques éducatives par l’induction croisée de leurs régularités et de leurs variations : Une méthode de recherche ancrée dans les observations. Geneve: Université de Genève, Faculté de psychologie et des sciences de l’éducation. Consulté à l’adresse https://www.unige.ch/fapse/SSE/teachers/maulini/publ-1313.pdf Pollock, K., & Hauseman, D. C. (2017). Observational Research on the Work of School Principals : To Time or Not to Time. In L. Ling & P. Ling (Éds.), Methods and Paradigms in Education Research (p. 88‑107). Hershey: IGI Global. Progin, L. (2017). Devenir chef d’établissement : Le désir de leadership à l’épreuve de la réalité. Enquête sur l’entrée dans un métier émergent. Bern: Peter Lang. Showunmi, V., & Fox, A. (2018). Exploring research methods for educational leadership. Management in Education, 32(1), 3‑5. https://doi.org/10.1177/0892020617748139 Strauss, A. L., & Glaser, B. G. (2008). Grounded theory (2e éd.). Bern: Huber. Tulowitzki, P. (2019). Shadowing school principals : What do we learn? Educational Management Administration & Leadership, 47(1), 91‑109. https://doi.org/10.1177/1741143217725325