TIME PRESSURE AND INFORMATION VALENCE: EFFECTS ON ATTITUDES, CHOICE RULE SELECTION, AND COGNITIVE RESPONS

The research was conducted to test a series of hypotheses based on Wright's research. These hypotheses stated that under high time pressure subjects should simplify their decision making. If the subjects process positive information, then they should simplify their choice rules by using a disjunctive or positively biased choice rule. If subjects process negative information, they should simplify their choice rule by using a conjunctive or negatively biased choice rule. Furthermore, high time pressure should cause the number of supportive arguments and the number of counterarguments to decline. The hypotheses were tested by manipulating time pressure (high vs. low) and information type (negative vs. positive) in a 2 x 2 factorial experimental design. Anderson's functional measurement was used to measure choice rules used in each treatment condition. Subjects evaluated eight profiles for a hypothetical textbook service.;The results showed that subjects under low time pressure-negative information used a differentially weighted averaging choice rule, and subjects under high time pressure-negative information used a linear choice rule. Yet, as predicted the attitudes under high time pressure-negative information were less favorable than under low time pressure-negative information. These results were interpreted to support Anderson's two-stage model of information integration. As predicted, subjects in the low time pressure-positive information condition seemed to use a linear choice rule, and subjects in the high time pressure-positive information condition used a multiplicative rule indicating that a disjunctive or positively biased rule was used. However, a ceiling effect may have prevented subjects in the high time pressure condition from expressing more favorable attitudes than those in the low time pressure-positive information condition. Finally, no differences among treatments were found for the number of counterarguments or supportive arguments. Also, self-confidence was not related to the number of counter- or supportive arguments. These results indicated that the profiles were too simple to generate arguments beyond those concerning the product dimension levels.