Estimating retest effects in longitudinal assessments of cognitive functioning in adults between 18 and 60 years of age.