Why your research may be out of order

This article discusses how age-sensitive context effects may lead us astray. As Western societies age, the information needs about the life conditions, consumer behaviours, attitudes and routines of older people are steadily increasing. Many surveys are now conducted with elderly respondents or include a substantial proportion of them. Recent research has shown that younger and older people are differentially affected by features of the research instrument. Such differential influences can lead researchers to draw dramatically different conclusions about age differences in opinions and behaviours, depending on the way in which the specific question is asked. This article highlights some of the relevant findings. When age differences in attitudes or behaviours are observed – unexpectedly or expectedly – one is well advised to carefully examine whether they could be a methodological artefact. Any observed difference may reflect a true difference in the attitudes or behaviours being studied, a differential influence of features of the research instrument, or an unknown mix of both. 'Age-sensitive' questionnaire design requires that researchers think through the issues at hand for every question asked and minimise the differential impact of memory on survey responses. (PsycINFO Database Record (c) 2007 APA, all rights reserved)