ABSTRACTA testing procedure was developed to determine the concentration range including an absolute or difference threshold for a compound added to a background medium. Samples were arranged into three groups, each containing a labeled control and three randomly arranged samples. Random samples in each group included an unlabeled control and two samples containing adjacent concentration levels from a series of six increasing concentrations of the added substance. Groups were presented in ascending order by concentration of the added substance. Within each group, panelists rated each sample on a nine‐point intensity scale of difference from the labeled control. Group specific psychological errors in judgement were reduced by subtracting the unlabeled controls from the sample scores. Analysis of variance was applied to these data yielding separate estimates of variance for sample scores within groups. Sets of t‐tests from paired comparisons were used to construct a confidence statement regarding a threshold inter