Acquiescence as a Function of Test Type and Subject Uncertainty

Abstract

Empirical studies have shown that attitude items are loaded more highly on a general acquiescence factor than are personality-type items. These results have been explained by postulating that two different psychological processes underlie acquiescent responses to attitude and personality measures—the first, a general proclivity for agreeing, and the second, an impulsive tendency to endorse item content as being self-descriptive. A more parsimonious explanation, consistent with the conceptualization of acquiescence as a tendency to agree when uncertain, suggests that these results are artifacts of the measurement procedures used, and that content may confound the comparison of responses to attitude and personality items by unequally reducing subjects' uncertainty across measures. This alternative conceptualization also suggests that instructions reduce uncertainty. To test these questions, 160 undergraduate students at the University of Washington were given a "content free" passage to read. At random, students were asked to respond to one of two sets of directions and to one of two types of tests (Select the answer that best or least represents the kind of student you are or that represents your opinion about students). Thus, best-least formed one main effect for positive and negative directions, whereas kind of student you are-your opinion about students formed a second main effect for personality and attitude, respectively. Students receiving positive instructions were more likely to respond true on a content-free test than were students receiving negative instructions, F (1,156) = 8.87, p < .003. As hypothesized, students receiving attitude test instructions were no more likely to respond true than were students receiving personality test instructions. These results suggest that agreeing more on personality than on attitude items (as was found on previously published research) results from the confounding effect of content.