| Sumario: | Reviewing recent developments in the study of organizations and in the use of informants, this piper notes the compatibility of both and the need for further refinements. As an example, it presents methods used in a study of authority conflict in 136 Catholic dioceses in America. These include principles for choosing a small number of informants per unit and two sets of auxiliary methodological propositions. The first, an "instrumental" theory, was part of a strategy to control measurement error. It guided the data collection process, and the creation of "error" control variables. The second set of propositions, devised for evaluating measurement error, presented expected mathematical properties of such error variables. These properties were expected to hold under the condition that the data were sound enough for substantive analysis. For these data, hypotheses were basically supported.
|
|---|