Sepplord wrote: ↑Thu 16 May 2019 11:45 AM
Hellzakk wrote: ↑Thu 16 May 2019 11:05 AM
i expect that someone who would do some irony about the "gaussian distribution" can understand the word -example-
uhhh...example doesn't mean you can just take a smallsample and derive statistics from it
an example still needs a sufficient amount of datapoints.
Everything you could deliver would be "just an example" since you do not have acces to the data of every single craft that was ever made within a timeframe
Why is admitting small errors such a huge problem?
Doubling down with even more ridiculous statements does the OPPOSITE of saving face
And you have determined what is a sufficient sample size? How exactly? Differences in the spread of the data between two groups can influence what sample size is adequate. What's your reasoning for saying that, with the differences reported, this sample size is insufficient? If your answer is just your own "feeling" or "lol it's only 10" you're going to sound even more ridiculous than you think he sounds. For the record, I'm not stating my own opinion on whether his numbers are 'adequate' because (and this is the most important part):
You're misinterpreting what he means by 'example' and I think you're doing it intentionally. You say "everything is just an example" because he doesn't have the entire data set of all crafts ever in the timeframe... as if that's an intelligent argument. This almost never happens, that's why we take
samples of populations. Sample and example are two different words, because they mean different things. By your logic we can never draw any conclusions about anything because we will never have access to all the data relevant to any question. He hasn't made any claims this is a perfect, adequate, or random sample, which is probably why he didn't even use the word sample. He also didn't try to "derive statistics".