1 sample ci quiz Flashcards
why is it necessary that when testing normality for proportions that both numbers are greater than or equal to 10?
the condition is necessary because the formula for the CONFIDENCE INTERVAL RELIES ON THE FACT that the BINOMIAL DISTRIBUTION CAN BE APPROXIMATED BY A NORMAL DISTRUBUTION
which then results in the SAMPLING distribution of the sample proportion being approximately normal
+ approximately won’t work well unless both n phat & n(1-phat) (**aka both counts of success and failure) are greater than or equal to 10
“is the two sample z interval for a difference between proportions an appropriate procedure to investigate the difference?”
the procedure isn’t appropriate because one of the requirements for using a 2 sample z interval for a difference between proportions is that the two proportions are based on 2 INDEPENDENT samples
but in this situation the 2 proportions come from a single sample so they cannot be independent
DO NOT FORGET UNITS FOR MEAN (t interval, sample mean, sample st.d)
ppm, etc.
when testing normality for means and n does not meet 30 but is a stemplot
n= 23 which is too small because it is not greater than or equal to 30 BUT the stemplot in the problem does not show STRONG skewness or outliers
It is plausible that the average delivery time is less than 30 min because the interval 26 min-40 min captures 30 and other times
There is convincing evidence that the average delivery time to your house is more than 20 minutes because all of the values in the interval 26 min-40 min are greater than 20
when stating plan:
- what type of sample (1 or 2)
- z or t interval
state: we want to estimate the true prop/mean of….(abbreviate!)
for 2 sample ci, we want to estimate (PARAMETER NOT SAMPLE) must specify for both units
interpret: we can be 95% confident that the interval (0.449, 0.511) capture the true proportion of american …(abbreviate)
when stating, must define each value ex. U1-U2 individually
ex. P1= the prop of all parts produced in a wing…
also use parameters!! not staty stuff
random??
If no experiment mentioned- Random INDY samples of 200 batteries or whatever selected from both shifts
If experiment mentioned- 2 groups of batteries of parts selected in a RANDOMIZED EXPERIMENT
normal?
if it’s prop, do n1(p1), n1(1-p1), n2(p1), and n2(1-p2)
if it’s mean, n greater than or equal to 30, if this doesn’t happen there should be a dot/stem plot in which you can then say the dot plot or whatevs doesn’t indicate strong skewness or outliers
MUST MENTION DF
use the smaller df (thus smaller n) when doing diff of means or props
- we can be 96% confident that the interval (-0.0156-0.0956) captures the true diff in prop of parts produced within specifics from the 2 shifts.
- the interval suggests that the proportions of all parrts produced specifically by the day shift is between 0.0156 less and 0.0956 more than the prop of the night shift ON AVERAGE
because zero is captured in the 96% ci of (-0.0156, 0.0956) zero is a plausible value for Pd-Pn and thus we don’t have convincing evidence to support the belief that there’s a significant difference from 0 in the diff of prop of parts produced between the 2 shifts
