Date of Award
Doctor of Philosophy in Psychology
Charles E. Collyer
Previous research suggested that subjects tend to investigate a potentially multiply caused event by initially exploring the most likely cause. Such search was said to be "truncated" in that it did not go on to explore additonal causes if the involvement of the initially hypothesized cause was demonstrated. It was suggested that this search strategy led to subjects' encountering a narrow range of information which would not allow them to adequately understand such an event.
The present study allowed subjects to investigate three hypothesized causes of an event by sequentially selecting three separate pieces of information. Subjects could choose to investigate the same cause or a different one at each question selection. Perceived likelihood of all hypothesized causes was measured before and after each new piece of information was encountered. After all information was selected, additional measures were taken, two of which called for the subject to interpret the information received.
The information available was confirmatory with respect to all three causes, although the nature of the task only allowed subjects to be certain of the involvement of one. Depending on the three pieces of information selected, subjects' feedback ranged between solid proof of one cause's involvement with no information about others, to proof of one cause with information making other cause likely, but not certain. Results indicated that there was a tendency for subjects to truncate search. In addition, most subjects, regardless of the scope of information encountered, agreed with a statement which implied that only a single cause was operating. Subjects also felt that they completely understood why the hypothesized event occurred, even though it was impossible to have thoroughly explored all three causes.
These result are seen to be the result of an initial set of "sufficiency" assumptions. Specifically it is suggested that subjects seek only to account for the event and encountered information, and to do so as parsimoniously as possible.
It is argued that the appropriateness of this strategy largely depends upon the cost of an incomplete understanding of the event
Ginpil, Stephen A., "Investigation and Interpretation of an Event with Multiple Causes" (1982). Open Access Dissertations. Paper 991.