GROUP GENDER COMPOSITION EFFEC'TS IN GENDER-ROLE ATTITUDE MEASUREMENT: TWO FAILURES TO REPLICATE

Some years ago an experiment was reported (Shomer & Centers, 1970) indicating that male col- .lege students respond differently to gender-role attitude questionnaires depending on the gender composition of the group in which the questionnaire is administered. The study found-in addition to the usual effect of respondent gender (women more egalitarian on this topic than men)--that the egalitari- anism scores of men (but not women) were higher in mixed-gender than in single-gender groups) If the finding is reliable, it suggests either (a) that the presence of women in the room completing the questionnaire at the same time raises the l~vfl of gender-role egalitarianism expressed by the men, (b) that the presence of other men in the room complet- ing the questionnaire at the same time lowers the level of gender-role egalitarianism expressed by the men, or (c) both. 2 Either way, the results raise the question of potential bias in the measuring of a particular variable (gender-role attitude) in the cir- cumstance of other people being present--a circum- stance that varies somewhat with the data collection mode but is apparently not uncommon when ques- tionnaires are administered in an organizational setting).