9 - further hyp tests + confidence intervals Flashcards
1
Q
hyp test for mean of a large sample
A
H0 μ =
H1 μ !=/</>
sample mean ~N(μ, σ²)
find p
if p<sig level reject H0
2
Q
hype test for mean of a large sample assumptions
A
- data is draw from normal
- population variance is known
(if n>25 you can use CLT or if you find the unbiased variance you can use that to bypass these assumptions)
3
Q
confidence interval
A
an interval that has a specified prop (eg 90%) of including the true pop value in it
4
Q
confidence interval method
A
(SM - Zσ/√n, SM + zσ/√n)
where SM = sample mean
find the Z value that gives 90% then do the conversion from Z world to normal
5
Q
width of confidence interval
A
2zσ/√n
midpoint is always the mean