The relationship between prior functioning on cognitive and personality dimensions and subject attrition in longitudinal research.

This study compared the intellectual and personality functioning of continuing participants in a longitudinal study with individuals who were lost from the panel after at least 7 years of involvement. Based on their reason for discontinuation, participants who were lost to the panel were categorized into three attrition groups: deceased, ill, and voluntary dropouts. When the performance of these groups prior to leaving the study was compared with that of the continuing participants, no general differences were found. However, specific differences were revealed, particularly on crystallized-type intellectual abilities and the personality factors. Participants who were lost due to illness or death had functioned at the lowest levels of the four groups prior to attrition. These effects varied by age. In late middle age, ill participants functioned at the lowest levels, whereas in old age, participants who subsequently died scored lowest of the four groups prior to attrition. Individuals who had voluntarily discontinued participation in the study most closely resembled the continuing participants, particularly in late middle age. The results indicate that performance effects associated with continued participation vary by age, the intelligence or personality factor being considered, and the reason for loss of the participant. Furthermore, they suggest that loss of participants from a longitudinal panel does not inevitably bias the validity of developmental findings.