Nick Byrd, 11 Nov 2020 "the reliability of individual [intelligence] assessments was poor [but] aggregate ratings showed good to excellent reliability.