Answers are grounded in this paper and cite their sources.
Four methods of panel intensity calculation were compared for 13 compounds, studied on five concentration levels using gas chromatography–olfactometry. All the methods produced sigmoidal dose–response functions or fractions of sigmoid, when the missing observations were replaced with zero. In contrast, panel intensities calculated only from the reported percepts provided predominantly linear plots. The linearity of such calibration curves can be additionally improved, either by employing a panel of similar dynamic range or by replacing missing observations with predicted (censored) values. Strong scale differences were observed. The panel intensity of methyl benzoate changed rapidly over a very small range of concentrations. Conversely, a very small change of panel intensity was observed for sotolon in a concentration range of three orders of magnitude. Such differences relate closely to the Stevens' slope, in the form of exponential growth function. A systematic decrease of panel intensity was observed for some volatiles at the highest concentration levels. Some of these observations could be explained with the help of current neurobiological knowledge.
No takes yet. Share an insight, caveat, or question.
Pet’ka et al. (2005) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: