Locating Longitudinal Respondents After a 50-Year Hiatus

Abstract

Many longitudinal and follow-up studies face a common challenge: locating study participants. This study examines the extent to which a geographically dispersed subsample of participants can be relocated after 37 to 51 years of noncontact. Relying mostly on commercially available databases and administrative records, the 2011-12 Project Talent Follow-up Pilot Study (PTPS12) located nearly 85 percent of the original sample members, many of whom had not participated in the study since 1960. This study uses data collected in the base year to examine which subpopulations were the hardest to find after this extended hiatus. The results indicate that females were located at significantly lower rates than males. As expected, sample members with lower cognitive abilities were among the hardest-to-reach subpopulations. We next evaluate the extent to which biases introduced during the tracking phase can be minimized by using the multivariate chi-square automatic interaction detection (CHAID) technique to calculate tracking loss adjustments. Unlike a 1995 study that found that these adjustments reduced statistical biases among its sample of located females, our results suggest that statistical adjustments were not as effective in PTPS12, where many participants had not been contacted in nearly 50 years and the tracking rates varied so greatly across subgroups.

Sixteen going on sixty-six: A longitudinal study of personality stability and change across 50 years.

“…First, a representative subsample of 4,879 participants was randomly selected from a 10% random subsample of the schools that were originally surveyed in 1960. Next, using a wide variety of tracking methods (see Stone et al, 2014), the project team managed to locate 84.8% of the random subsample: 15.5% were deceased, 50.3%…”

DamianSpengler_2020_EJP_Negligible Effects of Birth Order on Selection into Scientific and Artistic Careers, Creativity, and Status Attainment

“…Specifically, participants who stayed (vs. those who dropped) were slightly more likely to stem from intact traditional families, and to come from families with fewer siblings; they were also higher in SES, IQ, maturity, and extraversion (r = .06-.26). These attrition patterns are very common in longitudinal research spanning over decades (Stone et al, 2014).…”

“…Upon these exclusions, we were left with n = 4,513 after excluding non-credible answers, n = 4,140 after excluding twins, and n = 3,763 after excluding only children. Thus, for the present study, the total participant sample used at baseline At the 50 th year follow-up, of the 4,879 representative subsample targeted for data collection, the Project Talent team managed to locate 84.8%, of whom about 56% participated in the 50 th year follow-up (n = 1,952; see Stone et al, 2014). Due to the baseline sample exclusions described above as well as missing data on key outcome variables, our longitudinal sample (using listwise deletion across all study variables) consisted of 1,196 people (52.5% female, 90.4% White/Caucasian, 37% firstborn, Mage = 67.66, SDage = 1.23), though longitudinal samples differed depending on the specific outcome variable, because cases with missing values were non-overlapping (ns = 1,259 to 1,511).…”

Negligible effects of birth order on selection into scientific and artistic careers, creativity, and status attainment

“…At the 50th year follow-up, of the 4879 representative subsample targeted for data collection, the Project Talent team managed to locate 84.8%, of whom about 56% participated in the 50th year follow-up (n ¼ 1952; see Stone et al, 2014). Due to the baseline sample exclusions described above as well as missing data on key outcome variables, our longitudinal sample (using listwise deletion across all study variables) consisted of 1196 people (52.5% female, 90.4%…”