Effects of temporal and sequential stimulus presentations on logistic models from Sniff Olfactometer (SO) data

dc.contributor.authorLin, Jiehua
dc.description.abstractAn unanswered question in smell perception is "How does the brain represent a stimulus intensity?". A recent study (Ni, et al. 2022) showed that when human subjects are presented with odorant puffs in a sequentially increasing or decreasing order, as would be the case in the usual "stair-case method" (e.g.,1-2-3 and then 4-5-6), the data produced questionable models with enormous confidences intervals (CIs). However, when the concentrations were interlaced (e.g.,1-3-5 and then 2-4-6), the models were more robust and reproducible with CIs < 10% of the stimulus intensity. One consequence of interlacing is that the difference between adjacent odorants is much greater. We hypothesized that the brain is "recording" the difference between adjacent pairs, rather than their difference from a fixed standard. If true, this may have two consequences; 1) A noticeable difference is required for the brain to relate two stimuli, and 2) The interstimulus difference (ISD) should affect the CIs. The experiments reported here were designed to determine the effect of varying the ISDs between stimuli on the CI of the resulting logistic functions. Six concentrations ranging across the subject's threshold were grouped in two triads, and the probability of perception of an odor was used to model a threshold function from the six concentrations. The ISDs were increased from 1 minute to 1 day, 2 days, and 7 days; however, increasing the ISDs did not affect the threshold's robustness, reproducibility, or CIs.
dc.titleEffects of temporal and sequential stimulus presentations on logistic models from Sniff Olfactometer (SO) dataen_US
dc.typedissertation or thesisen_US


Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Lin_Jiehua_MFS thesis.pdf
4.17 MB
Adobe Portable Document Format