Hmm. A sample was 10 mL. At this point, the number of water molecules in the sample is already somewhere in the 1e23 range, right? So at at highest stated dilution the underdog actually detected, the remaining molecules of the eucalyptus hydrolat would be maybe in single digit to low double digit range, right? At this point I wouldn't expect even a single molecule of the scent to enter the nose of the dog in the first place. So it seems likely to me that either the effective dilution factor was lower due to some unknown effect, or it's not necessarily the eucalyptus the dog detected. My money is on the latter. The non-target samples were using the ultra-pure water directly, without also going through the same steps of the dilution process right? So given their wording, they didn't do "placebo steps" with the pipette also "diluting" the pure water control samples I assume? Maybe the dogs could detect the scent of the rubber part of the pipette itself or something like that? If some aspect of the dilution process itself left a scent, then it might remain at full strength no matter how often the steps were repeated.
It’s an extraordinary claim, so it’s wise to be skeptical. I agree that the most likely alternate explanation involves something in the preparation of the samples, and to their credit, the authors describe their dilution protocol in detail in the supplemental materials, so that’s the best place to look to spot the contaminant-hint. Whatever it is, it would have to either be present in the training sample that the underdog learned before the test, or the training had no effect and the underdog (and possibly the dogs who received 4 months of training in Study 3) were playing spot-the-difference rather than identifying the scent. If it’s a contaminant, the specific results of the study are unreliable, but even so, whatever contaminant was present must be incredibly subtle in its own right! Either way, this paper has definitely piqued my interest in dog olfaction studies!
10^23 how many eucalyptus molecules are even in the air at that concentration. must be single molecules per multiple breaths? it's crazy. My brain is suspecting flaws while handling the mixtures, so would have been nice to include that a bit more in the video. now I have to read it up 😅
Hmm. A sample was 10 mL.
At this point, the number of water molecules in the sample is already somewhere in the 1e23 range, right?
So at at highest stated dilution the underdog actually detected, the remaining molecules of the eucalyptus hydrolat would be maybe in single digit to low double digit range, right?
At this point I wouldn't expect even a single molecule of the scent to enter the nose of the dog in the first place.
So it seems likely to me that either the effective dilution factor was lower due to some unknown effect, or it's not necessarily the eucalyptus the dog detected.
My money is on the latter.
The non-target samples were using the ultra-pure water directly, without also going through the same steps of the dilution process right?
So given their wording, they didn't do "placebo steps" with the pipette also "diluting" the pure water control samples I assume?
Maybe the dogs could detect the scent of the rubber part of the pipette itself or something like that?
If some aspect of the dilution process itself left a scent, then it might remain at full strength no matter how often the steps were repeated.
It’s an extraordinary claim, so it’s wise to be skeptical. I agree that the most likely alternate explanation involves something in the preparation of the samples, and to their credit, the authors describe their dilution protocol in detail in the supplemental materials, so that’s the best place to look to spot the contaminant-hint. Whatever it is, it would have to either be present in the training sample that the underdog learned before the test, or the training had no effect and the underdog (and possibly the dogs who received 4 months of training in Study 3) were playing spot-the-difference rather than identifying the scent.
If it’s a contaminant, the specific results of the study are unreliable, but even so, whatever contaminant was present must be incredibly subtle in its own right! Either way, this paper has definitely piqued my interest in dog olfaction studies!
with 1 out of three there is also a high chance that the last few orders of magnitude are just random chance?
Olfaction for the olfaction throne!
Do you think the other dogs call the underdog “teacher’s pet”?
Is this homeopathy levels of dilution?!
10^23 how many eucalyptus molecules are even in the air at that concentration. must be single molecules per multiple breaths?
it's crazy.
My brain is suspecting flaws while handling the mixtures, so would have been nice to include that a bit more in the video.
now I have to read it up 😅