A series of pilot studies were conducted to operationalize and further develop hypotheses about cognitive aspects of organizational reporting which were formulated based on earlier work related to proxy reporting within households. Since universities and university departments are themselves organizations, although of a special kind, and were easily accessible to us, we started with them. PRIOR RESEARCH Cognitive Aspects of Proxy Reporting In order to gain a better understanding of the processes people use in answering survey questions, researchers are drawing on theories of cognitive psychology and social information processing (Bradburn, Rips and Shevell, 1987; Feldman and Lynch, 1988; Hippler, Schwarz, and Sudman, 1987; Jabine, Straf, Tanur and Tourangeau, 1984; Loftus, Fienberg and Tanur, 1985; Schwarz and Sudman, 1991, 1994). This theoretical work can be used to increase our understanding of proxy reporting. Sudman, Bradburn and Schwarz have summarized much of this work in their new book Applications of Cognitive Science to Survey Methods (Jossey-Bass, in press.) Individuals go through four stages when answering survey questions (Strack and Martin, 1987; Tourangeau, 1987; Tourangeau and Rasinski, 1988). These include (1) interpreting the question, (2) retrieving relevant inputs or a prior judgment from memory, (3) integrating information to form a judgment and (4) reporting a response. Organizational respondents are most likely to differ from self reporters in stages 2 and 3, retrieving information and integrating this information. Retrieval and judgment processes are affected by how information is learned or encoded and how it is stored in memory. Encoding Processes The context in which encoding takes place will affect the cues which activate retrieval of information from memory at a later time. One's own behaviors provide a rich set of experiences, including information about what one wanted to do, what one actually did, how one felt while doing it, etc. Thus, the episodic representation is likely to include information relevant to the event, such as the location and emotional responses (Tulving, 1972, 1983). In contrast, when reporting about organizational events or issues, informants are typically answering questions about reported events, events that may be learned "second-hand." These events are likely to be represented as episodes which relate to the occasion of receiving or learning about the event (Larsen and Plunkett, 1987). Encoding methods have implications for strategies used to answer questions about organizations. Cues related to the event, itself, will be more effective in enhancing recall for organizational informants who participated in the event than for those who merely heard of it. Since reported events are not necessarily encoded in chronological order, organizational informants who did not participate in the event should be less likely than participants to use a chronological pattern (a backward or forward time search of memory) when reporting about events in a specified period. Finally, this suggests that organizational participants in an event should have similar reporting strategies while non-participants strategies will vary. There are also several studies that show that information that is relevant to important others receives increased elaboration (Bower and Gilligan, 1979; Kuiper and Rogers, 1979). This suggests that relative accessibility of semantic versus episodic stores of organizational information will be a function of the respondents' positions within the organization as well as involvement in the behavior.