Item-pair measures of acquiescence : the artificial inflation of socially desirable responding

Abstract

The purpose of this study was to evaluate the role of socially desirable responding in an item-pair measure of acquiescence from the Big Five Inventory. If both items in an item-pair have desirable content, the likelihood of agreeing with both items is increased, and consequently, the type of responding that would be taken to indicate acquiescence. In Study I, item content desirability was evaluated for each of the 32 items belonging to the item-pairs in two samples of 214 and 68 university students. The item-pair desirability was then correlated with the percentage of respondents who agreed with both items in a separate sample of 895 students. Results showed a substantial correlation between item-pairs’ desirability and the percentage of estimated acquiescence, indicating an inflation of acquiescence when item-pairs have desirable content. The finding was further supported by Study II, in which acquiescence and item difficulty, assessed with cognitive interviews, were unrelated.