CI, MOE

Question 1

Confidence interval = INTERVAL ESTIMATE

Answer 1

interval estimate, estimates pop parameter, NOT designed to estimate a sample stat

Question 2

sample stat (survey or sample taken mentioned)

Answer 2

x bar +- t* s/sqrn

Question 3

pop prop z score (if you don’t have sigma aka pop std use t)

Answer 3

phat +- zsqrphat(1-phat)/n OR z sigma/n

Question 4

CI does NOT equal probability or pt estimate

Answer 4

point estimate is like the median between the two points given and MOE is the respective distance

Question 5

Interpret CI

Answer 5

we are 90% confident that the interval from (200 to 250) WILL CAPTURE the TRUE NUMBER/MEAN/POP PROP of steps in an hour

Question 6

Interpret Confidence Level

Answer 6

If we took ALL the samples of the same size from N (the pizza delivery times to your house) and made a 95%, CI, 95% of these intervals will capture the true pizza delivery time to your house

Question 7

it is plausible that the average delivery is less than 30 minutes because the interval 26-40min captures delivery time of 30 and other times

Answer 7

**the true delivery time is not necessarily 30 min so do not say that

Question 8

there is convincing evidence that the average delivery time to your house is more than 20 min because ALL of the values in the interval 26min-40min are more than 20 min

Answer 8

there is convincing evidence that the average delivery time to your house is less than 50 minutes because all of the vales in the interval 26-40 min are below 50

Question 9

to find confidence level- normcdf
to find critical value- invnorm/invt
to find moe- cv(sd/se)

Question 10

standard error difference of means/proportions

SEphat1-phat2= sqr phat1(1-phat1)/n + phat2(1-phat)/n

SExbar1-xbars= sqr S1 SQUARED/n +S2 SQUARED/n

SEsigma1-sigma2= sqr sigma1SQUARED/n + sigma2SQUARED/n

Answer 10

SE= sqr phat(1-phat)/n

Question 11

check random
LITERALLY repeat question, SRS of 251 marbles from her desk, 100 mailboxes from the county, etc

Answer 11

check Large Count
prop: both counts of success and failure (nphat), (n)(1-phat) are greater than/equal to 10

mean: n= 30, 40, 35, etc which is greater than/equal to 0

Question 12

How to change a sample size to reduce MOE, for example, If sample size goes from 10 to 40, the MOE will be reduce by ½ because you multiplied the SE by sqrt(1/4) which is ½

Answer 12

if sample size n is divided by 4, the MOE will double
std/sqr100 vs std/sqr25

Question 13

The general structure/recipe for a CI is Pt estimate +/- MOE
–The Pt est can be sample median, sample mean, sample prop etc!!!!
–MOE is (CV)(SE