Jörgen sandberg. Stockholm School of Economics. A classic managerial problem, страница 7

In collecting data, communicative validity was achieved by: (1) establishing a community of interpretation to ensure an initial understanding between the optimizers and me about their work and my research, (2) using only two principal openended interview questions to encourage the optimizers to identify and describe to me what they themselves conceived of as central in engine optimization, and (3) dialectically using follow-up questions during the interviews to help me further ensure that I understood the optimizers' ways of conceiving of engine optimization. When obtaining data, pragmatic validity was achieved by: (1) observing the optimizers at work and comparing what I had observed with what they said in the interviews, (2) asking follow-up questions that required the optimizers to demonstrate what statements meant in practice, and (3) observing the optimizers' reactions to particular interpretations of their statements. Reliability as interpretative awareness was achieved when obtaining data by being oriented to the ways in which the optimizers were conceiving of their work throughout the observation and interview phase. More specifically, (1) I primarily asked what and how questions in order to encourage the optimizers to focus on describing what engine optimization meant for them, (2) initially strived to treat all the optimizers' statements about their work as equally important, and (3) asked extensive follow-up questions that required the optimizers to elaborate on and be more specific about what they meant by their statements.

In the analysis, communicative validity was achieved by making interpretations of the optimizers' statements about their work that were consistent with both the immediate context of surrounding statements and with the transcript as a whole, Also, the identified conceptions were presented to the optimizers on two occasions, initially to the 20 study participants and then to all 50 optimizers in the department. On both occasions, the optimizers confirmed that the identified conceptions were valid. Reliability as interpretative awareness was achieved in a similar way in the analysis as in the interview phase. Throughout the analysis, I focused on the ways the optimizers conceived of engine optimization. I tried to maintain such a focus by: (1) both trying to hold back my own preunderstandings of competence and continuously checking if my interpretations were grounded in the optimizers' descriptions of their work, (2) initially treating all the statements made by the optimizers as equally important in my interpretations, and (3) checking my interpretation of each optimizer's conception by reading through transcripts expressing one particular conception using a qualitatively different conception.

Besides achieving the above criteria, the results were replicated in two ways. First, I chose a random sample of ten transcripts, three for cohception 1, five for conception 2, and two for conception 3. An independent researcher coded the selected transcripts against the conceptions, achieving agreement of 90 percent with my categorizations. Second, my presentation of the results to the optimizers led to a request from Volvo for a model for competence development based on the conceptions I had identified. To elaborate such a model, Theman (1995) replicated the study using 7 additional optimizers from the group of 50, selected according to the criteria of the original study. The selection of 7 more individuals was primarily motivated by the requirement for a deeper and more detailed description. Theman's study confirmed

the three identified conceptions of engine optimization in the reported study. The distribution of the optimizers across the conceptions was the following: 2 in conception 1; 3 in conception 2; and 2 in conception 3.

RESULTS

Competence in Engine Optimization