Abstract
There is considerable evidence that the act of participating in a survey can alter participants' attitudes, behaviors, and other outcomes in meaningful ways. Considering findings that this form of panel conditioning also impacts health behaviors and outcomes, we investigated the effect of participating in an intensive half-century-long cohort study on participants' longevity. To do so, we used data from a 1957 survey of more than 33,000 Wisconsin high school seniors linked to mortality records. One third of those people were selected at random to participate in the Wisconsin Longitudinal Study (WLS); the other two thirds were never again contacted. Our survival models show no evidence of panel conditioning effects on longevity: People selected at random to participate in the WLS had the same mortality outcomes as their peers who were not selected. This finding holds for the full sample, for women, for men, for population subgroups defined by family socioeconomic origins and educational experiences, and for treatment compliers.