 Coryn A. L. Bailer-Jones

Coryn A. L. Bailer-Jones developed a statistical treatment of estimating stellar Euclidean distances from noisy stellar parallaxes. His 2015 analysis showed why simply taking the reciprocal of a measured parallax can behave poorly when the fractional uncertainty is appreciable.^[ref]

A Bayesian inference combines the parallax likelihood with a physically chosen distance prior. A positive-distance prior can assign sensible posterior probabilities even to negative measured parallaxes; a negative noisy measurement does not imply a negative physical Euclidean distance.