A. Causal uncertainty and information processing

The main thrust of my research currently concerns causal uncertainty. Causal uncertainty is a perceived failure to understand clearly the causal conditions for events (Weary & Edwards, 1994; 1996). People who are causally uncertain believe that they do not adequately understand the causal structure of the environment. They feel that their knowledge of causation is flawed or inadequate. Although everyone has available such beliefs, people differ in how easily such beliefs are activated. When activated, such beliefs motivate perceivers to attempt to regain an understanding of social causation. This motivation has a variety of effects on the manner in which people seek out (Weary & Jacobson, in press) and process information (Edwards, 1996).

Our initial work on causal uncertainty concerned its measurement and antecedents. Gifford Weary and I recently have constructed and validated a measure of causal uncertainty (Edwards, Weary, & Reich, 1996; Weary & Edwards, 1994). In addition, using a prospective design we have found evidence that chronic causal uncertainty often has its origins in perceived lack of control (Edwards & Weary, 1996). Because of this link to lack of control, causal uncertainty is related to depression.

My more recent work on causal uncertainty concerns its effects on the attribution process. One set of studies is based on the premise that causal uncertainty leads perceivers to lack confidence in their conscious inferential reasoning processes. This causes perceivers to prefer categorizations and other sources of inferences that they believe do not result from their conscious processing efforts. One such source is the results of more automatic processes (e.g. spontaneous trait inferences, unconscious assessments of covariation). People generally do not recognize that the results of these processes are inferences; they believe them to be innate characteristics of the target. One set of studies has shown that because of such preferences, causally uncertain perceivers tend not to adjust dispositional attributions initially based on a behavior to take into account situational information relevant to the behavior (Edwards, 1996). I am currently conducting studies designed to show that the perceived diagnosticity of the situational information will moderate this effect (causally uncertain individuals should be sensitive to and use highly diagnostic information in making judgments but not use information of moderate or questionable diagnosticity).

B. Depression and impression formation

A second line of research concerns the effects of mild and moderate levels of depression on social judgment. With Gifford Weary and others, I have investigated how mildly depressed perceivers' chronic feelings of lack of control and causal uncertainty affect the processes by which people form impressions of others. We have found that depressed subjects tend to generate more inferences in response to receiving additional information about a target other than do nondepressed subjects (Edwards,Weary, & Gleicher. 1991). Results such as these have previously been assumed to result from effortful, integrative processing of information, stemming from depressives' feelings of lack of control over life events. However, this assumption had never been confirmed experimentally. I recently (Edwards &Weary, 1993) verified this processing assumption. Mildly depressed individuals do indeed form impressions in an effortful, piecemeal fashion, even in situations where nondepressed subjects typically base their impressions on a simple categorization of the target.