I have been really getting into stats recently instead of economics so my real life trolling has changed too, i just find macro econ really boring and my degree is in stats so I do kind of need to understand it. Just did some sickening trolling on the fat jewish kid i lift with occasionally. This is from fb chat. He wanted to train and 2 and I was feeling 3.

me:just wait till 3 man

fat jewish brah:why

me: so we can train together <3

fat jewish brah: its gonna get crowded then

me: we need to hypthosis test that

does the number of people follow a normal distribution?

do we know the variance

(obviously not)

fat jewish brah: im not doing this now

me: if it doesn't follow a normal distribution should we do a wilconcox test or a sign test?

fat jewish brah: just go eat and we will go

Me: is the data pairwise?

if not should we assume the sample error is an accurate representation of the variance

shall we assume the variance of both times is the same?

H0: the mean of both times is the same

BUT

is this a 1 tailed test or a 2 tailed test

are we test is the mean at 3 > then the mean at 2 or just different

fat jewish brah: stop

me: H1: u3>u2

u is obv mew, but i dont have a latin keyboard

fat jewish brah: im gonna go at like 2:30 and do a bit of cardio beforehand

me: what significance level are we testing this hypothesis at by the way?

fat jewish brah: i told u im not doing this

me: im going to assume the date is pairwise so we'll just use Uz and Sz because its the most apprpriate

fat jewish brah: im headed to the gym now, bored

cya later

me: xox

I reckon i'd use the wilconcox or whatever its called method because I wouldn't want to assume the data was normally distributed. Assuming it was normally distributed i'd make the assumption the data was pairwise because i'd measure the mean number of people at the gym at 2 and 3 on the same day.

Edited by fuji - 3/6/13 at 5:55am

me:just wait till 3 man

fat jewish brah:why

me: so we can train together <3

fat jewish brah: its gonna get crowded then

me: we need to hypthosis test that

does the number of people follow a normal distribution?

do we know the variance

(obviously not)

fat jewish brah: im not doing this now

me: if it doesn't follow a normal distribution should we do a wilconcox test or a sign test?

fat jewish brah: just go eat and we will go

Me: is the data pairwise?

if not should we assume the sample error is an accurate representation of the variance

shall we assume the variance of both times is the same?

H0: the mean of both times is the same

BUT

is this a 1 tailed test or a 2 tailed test

are we test is the mean at 3 > then the mean at 2 or just different

fat jewish brah: stop

me: H1: u3>u2

u is obv mew, but i dont have a latin keyboard

fat jewish brah: im gonna go at like 2:30 and do a bit of cardio beforehand

me: what significance level are we testing this hypothesis at by the way?

fat jewish brah: i told u im not doing this

me: im going to assume the date is pairwise so we'll just use Uz and Sz because its the most apprpriate

fat jewish brah: im headed to the gym now, bored

cya later

me: xox

I reckon i'd use the wilconcox or whatever its called method because I wouldn't want to assume the data was normally distributed. Assuming it was normally distributed i'd make the assumption the data was pairwise because i'd measure the mean number of people at the gym at 2 and 3 on the same day.

Edited by fuji - 3/6/13 at 5:55am