The selection process used was to ask CEOs, one by one, to participate, until the desired number of four was reached. Five CEOs who were asked did not participate in the study: one resigned to run for political office, two were fired, and two declined. That two of the CEOs I chose were subsequently fired may raise questions about my perceptions of effectiveness. I would respond that effective CEOs have been fired-and more than once. I do not claim that these four chosen are effective CEOs, only that they are CEOs of large health care organizations who have held their positions for a number of years and that I believe them to be effective.
The four participating CEOs have been in their present positions for five years or more and work in large not-for-profit health care organizations with gross revenues in 1984 of $85 million or more and with 500 employees or more.
Prior to the study I knew three of the nine CEOs very well, two slightly, and four not at all. Of the four CEOs who participated, I knew two well and the other two by reputation. My beliefs as to the effectiveness of the nine CEOs were based upon the following: my perceptions of their local reputation (this was not systematically derived, but I did ask others about the CEOs whom I did not know well); the size and perceived achievement of their institutions relative to others in the area; and their longevity in their jobs, which is unusual in large health care organizations.
After a preliminary meeting, I interviewed each of the four CEOs twice, for a total of three hours. Several of the questions were adopted from Levinson and Rosenthal (1984, 292-94). As part of the study, each CEO was asked for the names of eight or nine close associates who had direct experience with his work. I interviewed each of these 34 individuals for 60 to 75 minutes.
Obviously, allowing the CEOs to choose the persons to be interviewed introduces a bias. I asked to speak with close associates representing a range of occupations, inside and outside the health care organization. I specifically wished to include trustees, other managers, and physician leaders. Close associates were not entirely laudatory of the CEOs with whom they worked, but all of them believed their CEO to be effective, although not in every aspect of the job.
Between the two interviews, I followed each CEO for five consecutive days during a week of his choice. One CEO had to be out of town unexpectedly for a day, so I spent one day of the next week with him instead. In every case, I spent most of each long working day with the CEO. There were some confidential meetings and some other meetings after work that I did not attend, but every CEO said he reported to me what these meetings were about and with whom they were held. In the cases of two of the CEOs, there were no confidential meetings during working hours from which I was excluded.
I also reviewed annual reports, articles written by or about the CEO and the organization, and financial analyses of all four organizations. I took copious but not verbatim notes during the interviews and observations; I did not use a tape recorder. I summarized the data daily and had all my notes typed to facilitate analysis. I then summarized the data and isolated themes.
