Fig. 2: Data cascade diagram for the study of flight and no-flight astronauts.
From: Viability of internal comparisons for epidemiological research in the US astronaut corps
Of the 350 astronauts originally included for study, 32 were disqualified for inclusion based on dropping out of the training class, dying before they could take their first flight, or having as-yet indeterminate flight status. This left 318 astronauts for potential study of long-term health outcomes. Effective sample sizes ranged from a high of 995.67 astronauts (standard logistic regression propensity model with inverse probability of exposure weighting), to a low of 45 astronauts when performing propensity matching.
