Hey /sci/,I was doing some research on Moth olfactory sensitivity, and then I pivoted over to canine olfactory sensitivity as one point of comparison for perspective.I saw the results from pic related, and it stated a range of 1:10^17 and 1:10^21.Am I wrong in assuming that the range presented there is fucking massive, and isn't exactly anywhere close to a definitive experimental result? I'd assume that whatever funding they got to perform that experiment and deliver those results would be considered a failure due to the sheer variability of the results, or rather a result of excessively flawed methodology?
>>169019364 orders of magnitude isn't too badAt least it's not 5
>>16901936There may just be one result which is specific and just happens to be between those numbers which they highlighter for unrelated reasons. You probably should read the method and the actual data instead of a simple summary if you want to find out what they actually mean by that and if the result is particularly bad or good. Simply diluting things to 1:10^17 is already difficult enough for amateurs and it could just be an estimate.
>>16901956I did. Their starting point was 1:10000. And keep diluting the solution.But they didn't use the same source of Eucalyptus Hydrolats, and called out that the concentration levels varied, even before diluting it. That's why I brought up the Methodology. That doesn't sound like even attempted to control for concentration levels before diluting it. Which probably contributed to the range they specified.But I would argue that's pretty shitty all things considered, and in my opinion kind of invalidates the findings altogether; especially given their starting dilution and the range they settled on.
>>16902006I assume they used a different bottle for each run of the experiment, and the 1:10^17 to 1:10^21 variance is due to their differing concentrations and freshness out of the bottle. ie, some bottles they could dilute the oil to 1:10^17 before they stopped detecting it, and others all the way to 1:10^21>invalidates the findingsThe findings were that dogs can be trained to find very low concentrations of eucalyptus hydrolat, and that consistency in the scent source is crucial. Seems well supported by the results (and also sort of obvious without doing any experiment.
>>16902017The variance is due to the minimum detection threshold varying between dogs.Here's the article since OP is a fag and didn't link it https://www.mdpi.com/2076-2615/14/7/1083
>>16901937scientists actually believe this
>>16902030yeah op should have linked that but he's a retard. I don't get how he can be confused if he read the whole thing.
The differences between oils was well analyzed.
>>16902017Right, so they have an uncontrolled value. They couldn't stop detecting something at 1:10^17 if they continue 4 magnitudes further. That doesn't make sense.And to the second point, you say "dogs" as a plural; and the study doesn't even control for the same dog through the variability of the tests of variable concentrates before dilution. How is that a solid methodology? Explain that. >>16902030I didn't read any article. I read the paper.>>16902092What about the OP made you confused? I was debating the accuracy of the study given the massive variability of the estimated ranges the study claimed. Even the snapshot you snipped shows a massive amount of variability in the findings. That was the whole point of the thread. How massive gaps in experimental ranges are considered empirical for the purposes of publication of studies that then get cited as fact.
>>16901936First, welcome to biology, it's crap. In no small part because variation is the norm. Some people have vision of 100:20, some are 20:100. Bio sensors are not all equal. Second, learn to read> the study doesn't even control for the same dog through the variability of the tests of variable concentrates before dilution. yes they did, they used an oil as stock for the tests and dilute it, (Methods, 2.1), the other ten solutions were just for NMR measures (Methods 2.9).> They couldn't stop detecting something at 1:10^17 if they continue 4 magnitudes further. That doesn't make sense.those are results for different dogs, one stops at a different threshold than another, how is that confusing?Obviously this study is underpowered and I haven't even heard of this journal, but that the range of sensitivity for a chemical receptor is large is not at all shocking, and congruent with much larger studies in a number of species
>>16903295>different dogs,Or different tests. Individuals vary as well and don't always perform the same. Personnel best are not averages
>>16901936Instrument sensitivity, ability to create a consistent smell density, variation in the smell density due to animal breathing and the animal's physical distance from the smell emmitter and geometry of the animal's olfactory organ. Plus animal varriation is a huge factor. Even in model organisms each individual will have a slightly different response and this can be magnified by the animal's age.You wouldn't be surprised to learn that two children who are the same age had different smell sensitivity, so you would expect with a large number of test subjects the smell sensitivity would be a predictable range in a bell curve or more complex curve that suggests there may be different types of smellers.
>>16901936Just diluting 1:10^21 is really shaky. You could measure up 1 cm^3 fairly reliably with an uncertainty of 1:10^3, perhaps 1:10^4.Dilute that in 1 m^3 (water, I guess) and you have 1:10^9 - but that assimes no uncertainly in measuring up 1000 liters (doubt). So you have to extract a new 1 cm^3 from this dilution, again fairly reliably with an uncertainty of 1:10^3, perhaps 1:10^4 but put together with the previous uncertainty you are now 1:10^2 - 1:10^3 and now your dilution is at 1:10^18 so you need a third stage.At this point your experimental uncertainty is really dubious.
>>1690193617-21=4The difference is only four. You probably have more fingers than that on one of your hands. Touch grass.
>>16903393Not how dilutions work
