4 ms·

26 participants? That sound like a really solid statistical analysis. /s

• tasty_freeze 3y ago
  It sounds like they aren't reporting on what % of programmers use what substances. They interviewed 26 people for an hour each to find what they take, why, and what for.
• fsckboy 3y ago
  statistical significance is a very subtle topic that I'm not at all equipped to explain, but we can start with, "significance comes from sample size, and is independent of population size." If you flip a coin 26 times, you will establish its bias to a high degree of certainty, results that will hold if you flip the same coin a million more times. if you are presented with a bag of a hundred coins, and another bag of a million coins, and you sample 26 of them from each bag and flip them, you will establish the same level of certainty about the bias of the coins in each bag. or something like that? I'm sure there are people here who can explain it better than I can, my point is to say "don't question a sample size of 26 unless you really know what you are doing"
• travisjungroth 3y ago
  26 can certainly be solid if the sample is representative, the effect size is large and the variance is tight. If you heard that 26 people randomly selected from America took a pill and died within a year, you probably wouldn't want to take that pill. None of that even matters much because there's barely any stats in the paper. I didn't even see anything besides descriptive statistics. It is, by their own description, a qualitative study.