A Robustness Report of "Do We Become More Lonely With Age? A Coordinated Data Analysis of Nine Longitudinal Studies"

The original study by Graham et al. (2024) investigated whether loneliness changes with age across the adult lifespan, synthesizing data from nine longitudinal studies via meta-analyses. The primary fi nding was that loneliness follows a U-shaped trajectory: decreasing from young adulthood to midlife and increasing in older adulthood (estimated Age2 regression coeffi cient of 0.07 with 95% confi dence interval from 0.02 to 0.13, age centered at 60 years). We computationally reproduced the reported meta-analyses. We assessed the robustness of the main fi nding with respect to alternative analytic decisions regarding the estimation of the heterogeneity variance and inclusion/exclusion of individual studies. We fi nd that the main claim from Graham et al. (2024) is robust regarding these decisions.