The Gay Party Question of the Eysenck Personality Inventory: An Example of Effects of Item Alteration on Direction and Latency of Response 1

Abstract

This study examined the effects of making an apparently innocuous change in the wording of an item on the Eysenck Personality Inventory to improve face validity. A microcomputer was employed to administer the inventory (Form A) to 90 female and 68 male volunteer respondents from college, university, and shopping center settings. Response direction and latency (in milliseconds) were recorded for each item. For every second subject the word ”gay” in item 25 (”Can you usually let yourself go and enjoy yourself a lot at a gay party?”) was replaced with the word ”lively,” a word judged by a panel of raters as equivalent in connotation to the word ”gay” as employed in the test. Data support warnings with respect to making even apparently minor changes in the wording of the items in a personality inventory. Respondents were slower to respond to the ”gay” item and endorsed it significantly less frequently. in addition, respondents’ post ”critical-item” response latencies to the ”gay” item differed inconsistently from those of respondents to the ”lively” item.