ABSTRACT
When it comes to studying change, experts have pronounced that “cross sectional data-so easy to collect and so widely available-will not suffi ce” (Singer & Willett, 2003, p. 3). Such approaches ignore some basic but critically important caveats. As noted by one team of prominent developmental methodologists:
It is never possible to know [from cross-sectional research designs] whether observed differences between people within different age groups refl ect the aging process or the fact that the people being compared have lived through different historical periods. The best that the researcher can hope for is that as many alternative explanations as possible have been eliminated through careful selection of respondents. (Cavanaugh & Whitbourne, 2002, p. 88).
