Chapter 2 – Literature review
3.7 Interview as method
Mason (2002) cites the interview as the most common modality deployed for collecting data as part of a qualitative methodology. The acquisition of what Easterby-Smith et al.
(2012: 126) refer to as “natural language data” involves the collection of information, or data, from organizational members in order to gain insights in to social and organizational realities. Via a series of semi-structured, in-depth interviews, the worldviews, perceptions and opinions of individuals and/or groups may be discovered through the language they use.
In their review paper “How many qualitative interviews is enough?” Baker & Edwards (2016) discuss adequate sample size when conducting research interviews. The authors concluded that, in the light of transcribing thousands of hours of interviews in the ‘publish or perish’ world in which researchers live, between ‘a dozen and sixty interviews’ may be taken as a guide, with thirty being the mean.
The qualitative data collected in support of this study was represented the views and experiences of some 15 distinct interview respondents, each of whom required a number of follow-up sessions to clarify certain aspects of data that was felt to be missing or was in need of further elaboration. When these follow-up sessions (in person, by telephone, via Skype etc.) are taken in to account, the actual number of interview sessions exceeds 50. Where an interviewee’s story spanned several sessions, transcripts and/or the field notes for each session were concatenated to form a ‘master narrative’ for subsequent analysis.
3.7.1 Interviewee Sampling
Spradley (1979: 25-26) outlined a number of preferred interviewee qualities via three principles pertaining to the ‘excellent informant’; the second of which states that “it is necessary to locate ‘excellent’ participants to obtain excellent data.” Morse (2010: 231) adds to this by outlining what constitutes the ideal respondent with specific respect to a grounded theory approach: “An excellent participant for grounded theory is one who has been through, or observed, the experience under investigation.” Participants must therefore be experts in the experience or the phenomena under investigation, and fulfil some or all of the following criteria:
• Who has the relevant experience?
• Who is accessible?
• Who is willing to provide the relevant information?
• Who is most able to provide the information?
Given the considerable difficulty in securing respondents as part of a targeted (purposive) sampling approach, it was clear that I would need to rely on existing professional contacts as part of a sampling effort initially based on convenience (Easterby-Smith et al. 2008:
217). The Silicon Valley based (‘EXemplar’) respondents were professional executives, generally regarded as being ‘at the top of their game,’ and would have been impossible to book had it not been for the prior relationships and rapport that I had nurtured and maintained over the years with Silicon Valley based entrepreneurs in my field of specialisation (telecoms). The power of personal recommendation was critical to securing interview time with all U.S. based participants. Indeed, I was reminded on more than one occasion that the time allocated to my interview was probably worth x-thousand dollars, and had it not been for the personal introduction received from a highly-respected mutual acquaintance, my request for interview time – along with the many others doubtlessly
received each year from hopeful researchers – would have been summarily refused. For these reasons, I felt especially privileged to have been granted an audience at all.
UK-based respondents were much easier to deal with. Again, sampling was based initially on a convenience approach, where associates from my existing business network identified as possible contenders were invited to take part in the study. Many of those initially approached declined, but on several occasions offered potential candidates from within their network of associates; and in the event that those referrals were not able to assist either, would also put forward contacts from their network of contacts. The cascading power of a so-called ‘snowball’ approach to sampling (Atkinson & Flint 2001;
Biernacki & Waldorf 1981) was therefore recognised not only for its ability to identify hidden populations of otherwise inaccessible respondents, but when coupled with the power of personal introduction, represented a formidable means of securing interviewees that may otherwise have been inaccessible.
3.7.2 Respondent selection
Research participants with past and present experience in the creation and ongoing management of high-technology start-ups were identified specifically on account of their familiarity and experience operating in high-velocity technology environments, and it was the recollections of these respondents that were used to build a natively sourced, empirical narrative to feed-in to a comprehensive grounded theory study.
Unless created by way of merger, acquisition or other management buyout, most-all of the larger firms in existence today have a historical timeline that can be traced back to their formative days as a start-up entity. Each of these companies began life as the idea of a founding individual or group of founding members, typically beginning with little in terms of starting capital or other resources. It is on the strength of this common background that the idea of collecting a group of ‘exemplar’ interviewees was based, where so-called ‘expert entrepreneurs’ who now occupied senior positions within established corporations, were sought as interview respondents.
In making the distinction between exemplar/expert (EX) and novice ‘start-up’ (ST) entrepreneurs, Sarasvathy’s (2009: 21) definition of the expert entrepreneur as “a person who, either individually or as part of a team, has founded one or more companies, remained a full-time founder/entrepreneur for 10 years or more, and participated in taking at least one company public,” was adopted during the selection process.
Availability played a crucial, yet unavoidable role during the respondent selection stage, as interviewees were understandably averse to spending any time that did not directly contribute to the trading activities of their company. It was only via a protracted period of careful negotiation with an existing network of colleagues and acquaintances that the author was able to secure a series of 45-minute slots with each of the respondents.
Other than the need for individuals to have been engaged with the business of venture creation long-enough to be able to constructively reflect on the operational ‘highs-and-lows’ of the hi-tech start-up during its formative years, the criteria for respondent selection were not particularly specific. It was expected that such individuals would have (by now) reached a senior executive position, able to rationally contrast between then and now, and would able to offer a longitudinal view of the various milestones and break-points that had been pivotal to the successes (and failures) of the firms with which they have been involved. The selection criteria could therefore be summarised thus:
i) Founding member(s) fitted the entrepreneurial profile outlined earlier in this section, as distinct from non-entrepreneurial employees;
ii) They represented a typical example of an established, successful Silicon Valley tech company;
iii) The interviewee had survived the early start-up years. Whether the earlier venture(s) were still operational was not at issue; it was the fact that the respondent had successfully navigated the processes and demands of a challenging start-up environment that mattered.
3.7.3 Approaching Respondents
The approach to research is based upon a set of original empirical data which is broadly longitudinal in nature and is expected to address the birth-to-grave range of relevant experience encountered by respondents during their active years in the field. As part of the initial selection process respondents were asked to complete a pre-screening questionnaire (See Appendix Q), which was used to assess early experience gained in surviving the formative years of one or more start-up ventures (for ST and EX respondents), and also any later experience gained as the (EX) entrepreneur matured in to more executive roles.
The screening process was used to ensure that all data, as well as any derivatives concluded or theorised from it, would relate only to those who could be continuously identified as acting entrepreneurially in terms of their current and/or historical business
activities, and who had participated in the early formation of one or more ventures, either alone or as part of a small founding group.
While the 10-year break-point relating to degrees of experience was applied in accordance with Sarasvathy’s (2009) expert selection criteria (see previous section), care was taken during the pre-screening process to ensure that the start-up/novice (ST) respondents – especially those claiming significant levels of experience – had nevertheless also participated directly in the early stages of venture formation, as opposed to simply taking-on the management of a pre-existing firm, or having been appointed by investors as a professional board member to guide the early direction of a start-up.
Not all candidates were accepted. For example, a potential respondent indicated that he had successfully managed a family firm for a number of years, and then made the decision to establish an independent consultancy, described as a ‘spin-off’ from his previous activities in the family business. While this individual clearly had owner/manager experience as an employee in a family firm, he had yet to fully experience the formative stages of business generation, and so was excluded from the study. While the data from such respondents was extremely attractive from a convenience sampling perspective (Etikan et al. 2016; Robinson 2014), the potential for their data to corrupt the emergent concepts – as well as any subsequently ensuing theory – from immersion in anything other than an ab-initio start-up situation was too great a risk to take.