This proposal addresses the emergence of age-sensitive context effects in self-reports of opinions and behaviors and proposes experiments aimed at identifying the underlying cognitive processes. Preliminary analyses of survey data show that older adults are more likely to be influenced by the order with which response alternatives are presented within a question than younger adults (""""""""response order effects""""""""), but less likely to be influenced by the content of preceding questions (""""""""question order effects""""""""). In both cases, the age-sensitivity of the emerging context effects may lead to erroneous substantive conclusions about age differences in opinion and behavior, depending on features of the questionnaire. To avoid misinterpretations of methods effects as substantive effects, one needs to understand their age-sensitivity. Drawing on relevant theorizing in cognitive aging, social cognition, and survey methodology, the age-sensitivity of context effects in self- reports is traced to age-dependent differences in cognitive resources and a set of complex predictions is specified. These predictions are tested in laboratory experiments and an experimental survey. Combining an experimental and an individual difference approach to cognitive functioning, experiments 1-6 address the age-sensitivity of response order effects and experiments 7-12 the age-sensitivity of question order effects, cognitive resource is assessed through a number of individual difference measures as well as manipulated. For the final year, an experimental survey is proposed to assess the stability of the laboratory results under regular survey conditions.
