THE 20 STATEMENTS TEST - SOME MEASUREMENT ISSUES

This research investigated the issue of the weighting and number of Tw enty Statements Test (TST) responses with samples of 165 Hong Kong sec ondary school pupils and 100 Swedish and 100 Ethiopian university stud ents. Responses were classified into idiocentric, large group, small g roup, and allocentric categories, and scores were weighted by the rank order of the response and 7, 10, or 20 items were considered. Analysi s found that the rank and number of the items made a difference to the proportions assigned to these categories for the Hong Kong but not th e other samples and made little difference to the testing of hypothese s about gender and cultural differences in self-conceptions.