Differences between Duplicate Assays of Human Urinary Gonadotrophins

Sheps and Munson 1 have shown how the internal statistics of a single hormone assay may underestimate the variation actually observed on replication of that assay. Inter-assay errors may therefore be important when demonstrating qualitative differences between two hormone preparations by the simultaneous use of two different methods of assay. Such differences have been shown in human urinary gonadotrophins 2 , and the errors between duplicates of the assays used have therefore been investigated.