Sample Selection

The most important criterion for building a “participant pool” is to locate potential participants who have experienced or lived the phenomenon (Moustakas, 1994;

Polkinghorne, 1989). For this study, potential participants were individuals who were heirs apparent before becoming CEOs. The purpose of this study was to understand the heir’s apparent experience independent of age difference. I therefore adopted the definition of an heir apparent previously discussed in Chapter 2 – that is, any manager in a company who was a senior executive, including the COO and/or president, before becoming the CEO. This design allowed me to understand how the designation of executives as heirs apparent affected their learning journey from designation to assuming the CEO role. I continued recruiting participants until I reached sufficient variance in the degree of change in the suggested properties of sensemaking among participating heirs apparent and namely in the three hypothesized enactment forms. The design helped understand the impact of change in these properties on executives’ learning outcomes. It also allowed me to document how closely each participant’s experience conformed to the sensemaking process and to associate the degree of conformity with the degree of “success” of the heir apparent after assuming the CEO role. While this sampling strategy allows me to comment on the learning of heirs apparent, it does not allow me to explain the link between designation and actual company performance. The interviews I conducted were all in large successful organizations.

In order to build my participants’ pool, I downloaded the names of all of the companies headquartered in Ontario, Quebec, and Alberta from the Lexis Nexis Corporate Affiliations database. Individual information on each company included (1) company name, (2) complete address, (3) phone and fax numbers, (4) website address and e-mail, (5) number of employees, (6) sales turnover, (7) industry, and (8) current CEO name and e-mail address. To identify whether the current CEO had an heir apparent

experience, additional information on whether the current CEO was a senior executive in the company before becoming the CEO was gathered from other databases, namely (1) Mergent Online – Executive Search, (2) Thomson ONE Banker – Director Search, and (3) the Lexis Nexis people directory and subdirectories such as Marquis Who’s Who and the Dun and Bradstreet guide. The resulting names constituted my potential participants’ pool.

Canada is an important country for research in this area because it is the ninth largest economy in the world based on its gross domestic product (GDP) (EconomyWatch, 2010). Following the 2008 crisis, the Canadian economy proved to be one of the strongest advanced economies worldwide. I selected the provinces of Ontario, Quebec, and Alberta for several reasons, most importantly their economic and financial importance in North America and worldwide. Together, these provinces constitute around 75% of the GDP of Canada, with Ontario contributing around 40%, Quebec 20%, and Alberta 15% (Statistics Canada, 2012). Ontario’s GDP alone is higher than the GDPs of many major countries such as Belgium, Sweden, Switzerland and Ireland (Ontario Government, 2012). Furthermore, Ontario, Quebec, and Alberta host more than 80% of the companies headquartered in Canada.

Random sampling makes sense in the quantitative realm, where the researcher is interested in the generalizability of the research findings to a population (Denzin & Lincoln, 2011). Random sampling does not make sense in the qualitative realm or when adopting the phenomenological approach such as in the case of this research, where the purpose is to generalize back to theory (Merriam, 2002; Yin, 2011). Participants in a phenomenological study constitute the main data sources for the researcher. The

researcher carefully chooses participants based on their ability and willingness to describe and share their experiences with the researcher (Polkinghorne, 2005) and their accessibility to the researcher (Seidman, 2006). Qualitative research uses sampling strategies to provide a better understanding of the research questions at hand and to contribute to the existing knowledge of the phenomenon under study (Creswell, 2013). The theoretical sampling in this research included heirs apparent who were appointed CEOs from within organizations (to control for new learning related to industry or organization knowledge) and those who became CEOs not more than eight years ago. I limited the years of CEO experience to eight based on a study by Sherlock and Nathan (2008) which showed that CEOs can reliably remember their experience if it was within this range. CEOs were chosen from large organizations in different industries to control for the impact of company size on the quality and quantity of the learning that may take place during the transition.

Though very rare in large organizations, some of the organizations in the participants’ pool were family run businesses. Family members were appointed heirs apparent and became the CEOs such as in the case of Linamar Corporation where Ms. Linda Hasenfratz succeeded her father, Mr. Frank Hasenfratz, as the CEO. These cases were excluded from my sample as they may constitute a confounding factor to my ability to generalize back to the theoretical framework.

Even large family businesses are different from other enterprises in many ways including family aspects, management, and succession practices (Filser, Kraus, & Märk, 2013; Lansberg, 1999; Pearson, Carr, & Shaw, 2008). Hence the designation of heirs apparent in the case of family businesses is a different process. Lambrecht and Lievens

(2008) found family businesses prefer to have boards of advisors rather than boards of directors since they do not necessarily want to be directed. This dynamic is different for large companies where the boards of directors are actively engaged in directing the organization and who have CEO succession as their second priority after financial performance.

While succession in family businesses remains an important topic (Filser et al., 2013; Strike, 2013), these businesses usually seek advice primarily on strategic issues, followed by organizational issues and CEO priorities, and last about succession (Strike, 2013; Ward & Handy, 1988). The leadership in family businesses is different from that in publicly held organizations for several reasons (Stewart & Hitt, 2012). First, while leaders enjoy long tenure and are entrenched in family businesses, they experience high turnover and are subject to market discipline in publicly held organizations (Oswald & Muse, 2009). Second, successors who are family members are trained on the job and hence accumulate different, more personalized, knowledge about the business while successors in large organizations are trained in educational institutions only before joining the organization (Jorissen, Laveren, Martens, & Reheul, 2005). Third, leaders in family businesses are drawn from a limited kinship pool, while successors in large organizations are drawn from a large pool (Pérez-González, 2006). Goldberg (1996) found that effective successors in family businesses are assured ascension to the top job very early in their life, even as early as high school age. Therefore, the designated family members and others in the company know they are likely the heirs apparent even before officially joining the organization. For the above reasons, I excluded heirs apparent who were

family members from the sample as their identity and social context may have existed prior to joining the organization.

Appendix A describes the sampling strategy based on learning needs. It highlights the possible origins of the heir apparent as being from within the organization, different organization but same industry, or coming from a different organization from a different industry. Heirs apparent recruited from different organizations and different industries may need to spend their transition period learning about the organization and industry specific characteristics. Those recruited from same industry but different organization may spend some of their transition period learning about the organization’s specific characteristics. In order to focus on learning the requirements of the CEO role as much as possible, I only chose participants who were designated as heirs apparent and promoted from within their organization (the shaded area in the table).

Sample size is a similarly important issue (Creswell, 2013). The aim of phenomenology is to inform the research question and the phenomenon, rather than to generalize the findings to the population (Moustakas, 1994). Attaining large sample sizes alone is not the goal. In phenomenology, research sample size has been recommended to be between five and twenty-five (Polkinghorne, 1989). I was able to access and interview 22 participants. Two of my interviewees were female constituting 9% of my sample size. While the number is low, the number of female CEOs for large organizations is less than 5% in the top 1000 public companies in Canada (Eisenschmid, 2010). The female representation in the sample is therefore approximately representative of the national representation in Canadian public companies.

The sample size of 22 was high compared to the sample sizes in similar previous research, which range from 10 to 20 elite participants (Gabarro, 1987; Smerek, 2011; Welch, Marschan-Piekkari, Penttinen, & Tahvanainen, 2002). The large size of my sample provided me with theoretical saturation and enough variation based on the enactment argument (see Figure 4.1) (Creswell, 2013). The sample size also provided an acceptable representation from the three provinces (see Figure 4.2). Almost half of the respondents were from the finance and insurance or the oil and gas industries (see Figure 4.3). All the findings and discussions provided in chapters 5 and 6 respectively should be interpreted given the characteristics of this sample especially given that the respondents are currently CEOs of large organizations and 21 of the 22 respondents followed successful incumbent CEOs.

Figure 4.1: Sample Composition by Type of Enactment

5 23% 8 36% 9 41%

Figure 4.2: Sample Composition by Province 9 41% 4 18% 9 41%

Ontario Quebec Alberta

Figure 4.3: Sample Composition by Industry

Services 3 14% Finance and Insurance 6 27% Oil and Gas

4 18% Transportation 2 9% Agriculture 2 9% Others 5 23%