Hacker News new | ask | show | jobs
by FerociousTimes 1525 days ago
What about the month of March? Symptom-free too?

Not to take away from your great efforts to solve this problem but you controlled for a few key variables in your experiment but not for the placebo effect which is one of the most important factors to rule out in such experiments.
2 comments
jsherer 1525 days ago
Good question. Indeed, I prioritized improving my vitamin levels for health reasons, including the possibility that it may reduce my headaches, rather than running a purely scientific experiment. If the migraines come back while I maintain normal vitamin levels, I'll continue my experimentation!

March: 0

April: 0

link
anikan_vader 1525 days ago
Why would you want to rule out the placebo effect? If the placebo effect eliminates your pain, that's fantastic!
link
FerociousTimes 1525 days ago
What I meant is that you can't say with absolute confidence that supplementing for vitamins cured his migraine without accounting for the placebo effect in that experiment.
link
jfengel 1525 days ago
It's not a valid experiment regardless. There are too many confounding factors and too small a group.

It doesn't matter. Medicine is a lot less about proper science than we'd like to believe. It's about helping people feel better. Science is a crucial tool for that, but it doesn't always work, while random guessing sometimes does. Even the science is often supplemented with random guessing.

Eventually, you take enough random-guesses-that-worked and apply science to it. That gives you a lot more confidence to recommend it to other people. But often, confidence is all you get, rarely certainty.

link
FerociousTimes 1525 days ago
Yeah, you're definitely right; I should have added in the end "among many other factors".

While I admit that n is pretty small indeed and thus doesn't qualify for a proper and rigorous clinical experiment, the reason I mentioned the placebo effect in particular is because of the fact that some symptoms or diseases have some psychosomatic streak to them, if you will, and therefore the role of the placebo effect in these cases becomes more pronounced to research and investigate but if the OP felt way better after the treatment, that's what it all counts like you said.

It was no nitpicking or mischief on my part raising this issue :)

link