POLITICAL KNOWLEDGE, ATTRIBUTION, AND INFERRED INTEREST IN POLITICS: THE OPERATION OF BUFFER ITEMS

Bishop (1987) observed that being unable to answer a politicalknowledge question decreased self-reported interest in publicaffairs. This effect was unaffected by the introduction of upto 101 unrelated buffer items. In contrast, a single bufferitem that provided respondents with an external explanationfor their lack of knowledge greatly reduced the context effectin the present study. Implications for the operation of bufferitems are discussed.     

CONFIDENTIALITY ASSURANCES IN SURVEYS: REASSURANCE OR THREAT?

Over the last three decades, the public's willingness to takepart in surveys has gradually declined, and the decline hasbeen attributed in part to increasing concern about the confidentialityof the data requested. This paper reviews the early literaturebearing on confidentiality assurances and willingness to respond,and then reports on three experiments designed to investigatethe effects of confidentiality on the expectations of respondentsand on their willingness to take part in a survey. The resultsof all three experiments confirm our expectation that confidentialityassurances are not always perceived as reassuring, and do notnecessarily increase the public's willingness to respond.