Acetaminophen (Tylenol) reduces empathy?

[Pup]

I don't know why people always think bigger is better for sample size. Larger samples contribute to higher power but they also make it easier to find results that aren't practically significant.

If researchers blatantly data-mined, I think it would be obvious. As long as they reported the overall sample size, they couldn't really get away with "but on red headed women age 40-60, we found a significant result."

True unethicality would be starting with a sample size of 1,000, then writing a paper on the significant effect of Drug X on 50 middle-aged women with red hair without mentioning the other 950 people.

Those kinds of researchers might get away with it once or twice, but wouldn't last long, I hope.

 

[Elaedith]

I don't know why people always think bigger is better for sample size. Larger samples contribute to higher power but they also make it easier to find results that aren't practically significant.

Yes, the main problem with small sample size is failure to find a statistically significant effect even if the effect is present. However, that issue is irrelevant if the study did in fact find a statistically significant effect, since the statistical test takes sample size into account. If an effect is statisically significant despite a small sample size it indicates the effect is large and/or error variance is very low.

In contrast if you use a large enough sample you will end up finding statistically significant effects which are so small as to be meaningless.

However, because a conventional statistical test doesn't tell you the probability that the null hypothesis is false, but rather the probability of getting the outcome if the null is false, there is still a higher probability that an effect obtained with a small sample is not genuine (i.e. that it is the one in 20 that occurs by chance). The main issue with small samples is whether or not the effect consistently replicates. If it is a very robust effect it may replicate consistently even with very small samples.

 

[jaydeehess]

When I had Tylenol3s after surgery i was numb, literally numb. Was it the acetaminophen or the codeine?

Hard to say.

However, while being prepped for surgery I was given fentanyl. Surgerical procedure was a simple vascular repair in my left arm. They could have told me that due to a complication I would have to have my testicles removed and in the state that stuff puts one, I'd have said "whatever you think is best doc."

I doubt that i would have felt much different about others either.

For that matter it's pretty obvious that alcohol has this effect on empathy. Bar fights would seem to prove the point.