CONTRAST EFFECTS AND SUBTRACTION IN PART-WHOLE QUESTIONS

Responses to a question about the state economy changed when the quest ion followed a similar item about the economy of the respondent's loca l community. The overall distribution of the answers changed, as well as the correlation between the two items. Prior work has suggested tha t respondents may exclude specific aspects of a general issue from con sideration when a question about the general issue follows a question about the specific aspect of it. Responses to an open-ended item indic ate that respondents may have excluded a salient characteristic of the community economy when subsequently assessing the state economy, even though this characteristic figured prominently when respondents evalu ated the state economy first.