3 ms·

An interesting analysis with a fantastic level of detail. Much of the gwern's stuff is similarly well researched - I'm a big fan! However, I can't help feeling

by leoedin 13y ago

An interesting analysis with a fantastic level of detail. Much of the gwern's stuff is similarly well researched - I'm a big fan! However, I can't help feeling that it's fundamentally flawed. I understand that much of the statistical analysis done is to provide as rigorous a study as is practical, however I think in some cases it misses the point slightly. The biggest issue is that the variables that drive all the analysis are ultimately almost all perception based. I'd imagine that gwern is far more in tune with his perception of things than the average person will be (a by product of doing so many double blind tests), but ultimately he's still just a human. For me, many of the variables in the study - how well I slept, my mood, my willingness to work - vary considerably day to day anyway. Ultimately, the study found no difference. Given that we know that LSD is very active at the right doses, the only thing that you can really take away from this is that the dose wasn't high enough. Until a study has been done which yields perceptable differences, positive or negative, you can't really draw any conclusions. It's probably also fair to assume that the dose used was likely to be smaller than stated (a by product of the black market), and so the microdoses used were correspondingly reduced.

gwern 13y ago
> For me, many of the variables in the study - how well I slept, my mood, my willingness to work - vary considerably day to day anyway. Certainly. That's why it took closer to 6 months rather than 6 days. The more variation, the more data you need. > Ultimately, the study found no difference. Given that we know that LSD is very active at the right doses, the only thing that you can really take away from this is that the dose wasn't high enough. No. What you can take away from this is 'while no one doubts that at 100μg and higher one gets dramatic psychedelic effects, the suggestions that at tiny doses like 10μg there are non-psychedelic genericly positive effects, may not be true'. Don't confuse your hypotheses here.
- munchhausen 13y ago
  > No. What you can take away from this is 'while no one doubts that at 100μg and higher one gets dramatic psychedelic effects, the suggestions that at tiny doses like 10μg there are non-psychedelic genericly positive effects, may not be true'. Don't confuse your hypotheses here. What?? No, you can't take that away. In fact, what GP said still makes a lot more sense. 10ug didn't work for you, thus the dose was too low. Have you tried 15? 20? 30? It would almost certainly be sub-threshold when it comes to the psychedelia (unless you are a true freak of nature), but, who knows, maybe you would get an amazing nootropic effect! Why did you decide to abort your experiment after trying only an arbitrary dose of 10ug?
  - gwern 13y ago
    > Have you tried 15? 20? 30? It would almost certainly be sub-threshold when it comes to the psychedelia (unless you are a true freak of nature), but, who knows, maybe you would get an amazing nootropic effect! What? Dose-response curves do not work that way. They do not look like binary circuits where at 10mcg they do absolutely nothing and then at 15mcg go to the moon. They look like curves. If 15 mcg was better than 12, then I should have observed something like a medium effect. If 20mcg was better than 12, I should have observed something like a small-medium effect. If 30mcg was best, I should have seen signs of a small effect, possibly not statistically-significant, but clear point-value shifts. I observed none of that. > Why did you decide to abort your experiment after trying only an arbitrary dose of 10ug? It was closer to 12mcg, I did not 'abort' the experiment but carried it out to completion exactly as I had planned, and it was not arbitrary but close to the middle of the reasonable range of 5-30mcg.

by leoedin 13y ago

gwern 13y ago

> For me, many of the variables in the study - how well I slept, my mood, my willingness to work - vary considerably day to day anyway. Certainly. That's why it took closer to 6 months rather than 6 days. The more variation, the more data you need. > Ultimately, the study found no difference. Given that we know that LSD is very active at the right doses, the only thing that you can really take away from this is that the dose wasn't high enough. No. What you can take away from this is 'while no one doubts that at 100μg and higher one gets dramatic psychedelic effects, the suggestions that at tiny doses like 10μg there are non-psychedelic genericly positive effects, may not be true'. Don't confuse your hypotheses here.

munchhausen 13y ago
> No. What you can take away from this is 'while no one doubts that at 100μg and higher one gets dramatic psychedelic effects, the suggestions that at tiny doses like 10μg there are non-psychedelic genericly positive effects, may not be true'. Don't confuse your hypotheses here. What?? No, you can't take that away. In fact, what GP said still makes a lot more sense. 10ug didn't work for you, thus the dose was too low. Have you tried 15? 20? 30? It would almost certainly be sub-threshold when it comes to the psychedelia (unless you are a true freak of nature), but, who knows, maybe you would get an amazing nootropic effect! Why did you decide to abort your experiment after trying only an arbitrary dose of 10ug?
- gwern 13y ago
  > Have you tried 15? 20? 30? It would almost certainly be sub-threshold when it comes to the psychedelia (unless you are a true freak of nature), but, who knows, maybe you would get an amazing nootropic effect! What? Dose-response curves do not work that way. They do not look like binary circuits where at 10mcg they do absolutely nothing and then at 15mcg go to the moon. They look like curves. If 15 mcg was better than 12, then I should have observed something like a medium effect. If 20mcg was better than 12, I should have observed something like a small-medium effect. If 30mcg was best, I should have seen signs of a small effect, possibly not statistically-significant, but clear point-value shifts. I observed none of that. > Why did you decide to abort your experiment after trying only an arbitrary dose of 10ug? It was closer to 12mcg, I did not 'abort' the experiment but carried it out to completion exactly as I had planned, and it was not arbitrary but close to the middle of the reasonable range of 5-30mcg.