The interview technique was selected as the primary approach to collect field data because of its ability to assist in interpreting the significance of particular variables (Borg & Gall, 1989; Richardson, Dohrenwend, & Klein, 1965). It was chosen over other techniques and "objective" instruments because of its flexibility and adaptability (Kerlinger, 1986). The interview this study used can, in part, be labeled as moderately structured (Stewart & Cash, 1985). This is because the interview contained open-ended and probing questions that followed each major question. The probes were a mixture of open-ended and closed-ended questions. This funneling technique allowed the freedom to probe into and to adapt to different interviewee answers like the nonstructured interview does, but it also provided a schedule for the interview. As with the structured interview, this technique can be replicated fairly easily; it produces data that can be analyzed and compared, and it does not require the use of a highly trained interviewer. In effect, the moderately structured interview was selected based on a decision to utilize the strength and not the weaknesses of both the structured and nonstructured interviews (Stewart & Cash, 1985).
Though Kerlinger (1986) and Borg and Gall (1989) describe the face-to-face interview as "perhaps the most powerful and useful tool of social scientific research" (Kerlinger, 1986, p. 379), it is also costly, time consuming, and is subject to compromises made to improve the social context of the interview (Frey, 1983). Consequently, this study utilized the telephone to minimize the cost and time.
A key element of the interview schedule was a Behavioral Event Interview (BEI). The BEI was developed by David McClelland (1978) and colleagues at McBer and Company. It is based on the Critical Incident Technique that was created by Flanagan (1954). Flanagan had job incumbents write behavioral descriptions of critical incidents they experienced in their work. However, a problem with this technique was that the written incidents tended to not be detailed enough to determine what the job incumbent was thinking, feeling, and specifically doing. As a result, the BEI was developed so that a particular critical incident could be explored until behaviors, thoughts, and feelings were adequately reported (McClelland, 1978). Though BEI respondents may initially only discuss behaviors they believe are critical, additional probing can reveal other relevant behavior that occurred in the event. Klemp (1979) maintained that through the use of extensive probing, the interviewer can elicit descriptions of behaviors that were actually performed in the event, rather than more selective recollections of behaviors.
The BEI served to rectify some of the weaknesses that have been identified in studies of leadership (Boyatzis, 1982; Campbell, Dunnette, Lawler, & Weick, 1970; Murphy, 1988). The BEI has had a very successful history of use in a variety of settings, including business, industry, education, and the military (see Goleman, 1981; Huff, Lake, & Schaalman, 1982; Schmidt, Finch, & Faulkner, 1992; Spencer, 1979). Campbell et al. (1970) describe the Critical Incident Technique (upon which the BEI is based) as one of the most effective methods for assessing managerial behavior. In summary, the BEI was chosen because of its ability to focus on meaningful dynamic behaviors demonstrated by leaders that they judge have had an impact on their development as leaders.
Each CVA was asked to describe two on-the-job incidents that had the most impact on their professional development as leaders. The specific probes to be used by the interviewers were developed through group discussion by the research team as a whole. The intent of these probes was to gain as much description as possible about the incident and to obtain specific examples to capture what happened and what the administrator was thinking and feeling during the experience. The ultimate goal was to gain an understanding of how this experience was developmental and how its developmental effects might be replicated for others.
The eight-part Interview Protocol used the BEI technique. The following were the initial probes: