Initially Held Hypothesis Does Not Affect Encoding of Event Frequencies in Contingency Based Causal Judgment

Abstract

It has long been known that the event types of the standard 2 x 2 contingency table are used differentially in making contingency judgments. The present experiment sought to investigate the possible role of initially held hypotheses about the relationship between two binary, causally related events on subsequent causal judgments about those events and further, to investigate the role of encoding and/or retrieval processes. Subjects were given one of three hypotheses suggesting a positive, negative, or an indeterminate relationship between application of a chemical and plant growth. Subjects then received either 24 or 72 learning trials, with ∆P = 0.5 for all groups. Subjects then gave a causal judgment as to the relationship between the events and then were then asked to provide frequency estimates of each event type. We found that subjects’ initial hypothesis did affect subsequent causal judgments, with subjects given a positive initial hypothesis providing significantly higher causal judgments than subjects given a negative initial hypothesis. However, no effect of trial number was found on subsequent causal judgments. These results seem to suggest that, while subjects’ initial hypothesis about the causal relationship between two binary events did affect subsequent causal judgments of the relationship between those events, this effect was not mediated by differential encoding and/or retrieval of specific event type frequencies. Implications for the mechanism underlying differential cell use as well as possible future directions are discussed.