sig test pt.2 with proof Flashcards
(20 cards)
one sample z/t TEST!!!!!! NOT INTERVAL!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Po= null value, it’s like the P for population proportion
Random?
SRS/Random sample/randomized experiment of ________ SELECTED
Normal?
n(P) n(1-P) –> both counts are greater or larger than 10 (proportions)
n is greater than or equal to 30 OR the dot plot does not show strong skewness or outliers
we want the data APPROXIMATELY normal!!!!
10%
It is REASONABLE to assume that there are AT LEAST n(10) in __________ (if NOT stated, say our population of interest)
BECAUSE I want to make an accurate generalization (from a random sample)
if population not stated, say our population of interest
z= phat-P/sqr.P(1-P)/n
t= xbar-u/s/sqr.n
P/U=
true population/mean of ______
type 1 error
we could have made a type 1 error. we (incorrectly) reject Ho and conclude that there is convincing evidence that _______ but in reality _______
type 2 error
We could have made a type 2 error. we (incorrectly) fail to reject Ho and do not conclude that there is convincing evidence that_____ but in reality there is convincing evidence that ______
If the Pvalue is double sided (not greater than or less than) you have to multiply the ncdf or tcdf by 2
LABEL!!!!!!!!!!!!!!!!! INCLUDING ON THE DOT PLOT!!!!!!!!!!!!!!!!!!!!!!!!***
state, plan, do
exceptions to 10%
randomized experiment or infinite population
when doubling p value, z or t is always LESS than the value otherwise the result will be more than 1
DOUBLE!!!
if ha is not equal, the interpreting is that because p= ___ < alpha = 0.05, we reject Ho and conclude Ha, there is CE that the true prop/mean of ______ is different than ______ or we don’t reject ho and don’t conclude ha, so there isn’t ce that the true prop/mean of _____ is different than ________
different, change, etc.
UNLESS EXPLICITLY STATED with greater than less than, “different” “effective” “change” ARE NOT EQUAL!!!!!!!!!!!!!!!!!
the creek question in particular- it needs to be under 5 mg to be safe, so do we have ce that the animals are at risk–> we want to check if it’s under 5 mg or not NOT whether it’s above 5g!!!!!!!!!!!!!
we want to see if it’s below 5mg first because that’s the info provided to us to then be able to tell whether the animals are safe or not
DO NOT FORGET DF
with nonequal, use negative t or z value
