61–70 of 204 posts
Not sure how to reconcile this with the fact that suicide rates in rural areas of the US far exceed suicide rates in urban areas. Do the availability of guns in rural areas really offset this effect that much?
> The results show strong correlations, but research that would prove a causal link is difficult because ethical experiments cannot deliberately expose people to harm. This is kind of my hobby horse: you don't need to subject people to harm to test causuality. You can do the opposite: Remove harm and see what happens. Maybe give people free HVAC upgrades to filter more particles.
Even when you lack this level of control it may still be possible to infer causal effects. There's some interesting work that I've seen on this, good write up here: http://www.michaelnielsen.org/ddi/if-correlation-doesnt-impl...
Earlier quoted context omitted.
A treatment being a placebo implies nothing about the effect size!
Most placebo effects are modest and the ones that aren't are usually mixed in with regression to the mean effects(which are usually pretty modest). So OP is implying that if an effect is especially large it's less likely to be placebo.
Earlier quoted context omitted.
You do not deserve the downvotes. Just because the purifier takes particulates out of the air doesn’t mean the effects on humans isn’t placebo
This seems ripe for placebo. Installing purifier and indeed measuring lower PM, means naturally you feel less worried and more relaxed!
Earlier quoted context omitted.
Making the air smell like ozone means it’s probably doing something unhelpful/actively harmful in addition to filtering. Did it have an ionizer or similar?
No, it’s a Blue Air so doesn’t include an ionizer, it just when HEPA filter.
> The results show strong correlations, but research that would prove a causal link is difficult because ethical experiments cannot deliberately expose people to harm. This is kind of my hobby horse: you don't need to subject people to harm to test causuality. You can do the opposite: Remove harm and see what happens. Maybe give people free HVAC upgrades to filter more particles.
Plus unfiltered air is elsewhere, as a confounder.
Earlier quoted context omitted.
You do not deserve the downvotes. Just because the purifier takes particulates out of the air doesn’t mean the effects on humans isn’t placebo
This seems ripe for placebo. Installing purifier and indeed measuring lower PM, means naturally you feel less worried and more relaxed!
Earlier quoted context omitted.
I think he means based on the size of the effect it's unlikely to be a placebo.
Well the reasoning was pointing towards high chance of placebo though. He said there was lots of grime and also something he could measure was different, which means psychologically a lot greater chance and effect for placebo.
> The results show strong correlations, but research that would prove a causal link is difficult because ethical experiments cannot deliberately expose people to harm. This is kind of my hobby horse: you don't need to subject people to harm to test causuality. You can do the opposite: Remove harm and see what happens. Maybe give people free HVAC upgrades to filter more particles.
Problem being of estimating NNT to see the effect. If the effect size is small you will need huge numbers to detect it Plus unfiltered air is elsewhere, as a confounder.
Earlier quoted context omitted.
Even when you lack this level of control it may still be possible to infer causal effects. There's some interesting work that I've seen on this, good write up here: http://www.michaelnielsen.org/ddi/if-correlation-doesnt-impl...
Can recommend Judea Pearl's recent book on the subject. The Book of Why