Over the years, many different labels have been used to describe nontraditional students. This study examined whether terminology used in surveys could affect respondents' answers. Two forms of a questionnaire were designed and distributed randomly to college students. One form consistently employed the terms gay, lesbian, and bisexual while the other form employed the term homosexual. On 29 of the 30 items, there were no significant differences between the two groups. However, on one item, the difference was significant. In the true-false statement, "I have had at least one homosexual experience during the last year," only 2.5% of the respondents marked it true. However, for the statement, "I have had at least one sexual experience with someone of my gender during the last year," 10.4% of respondents indicated this as true. It is speculated that the term "homosexual" seemed to generate some negative reactions and that respondents did not want to see themselves as homosexual. It is concluded that if researchers wish to measure behavior, it may be preferable to avoid any labeling and to be as operational as possible. Contains 11 references. (RJM)