AP STATISTICS NAME: ____________________________

INFERENCE (sample means and proportions)

MULTIPLE CHOICE: Choose the best answer from the answer choices.
1. You have sampled 25 students to find the mean SAT scores at orris !nolls "i#h School. A $5% confidence interval for the
mean SAT score is $&& to 11&&. 'hich of the followin# statements #ives a valid interpretation of this interval(
(A) $5% of the 25 students scored between $&& and 11&&.
()) $5% of the population of all students at orris !nolls "i#h School have a score between $&& and 11&&.
(C) *f this procedure were repeated man+ times, $5% of the resultin# confidence intervals would contain the true mean of
orris !nolls "i#h School.
(-) *f this procedure were repeated man+ times, $5% of the resultin# confidence intervals would be between $&& and 11&&.
(.) *f 1&& samples were ta/en and a $5% confidence interval was computed, 5 of them would be in the interval $&& to 11&&.
2. A candidate for ma+or hires a statistician to determine the amount of support he has for the up comin# election. The
statistician tests the null h+pothesis that the population proportion e0uals 5&% a#ainst the alternative h+pothesis that the
population proportion is #reater than 5&%. The results of a simple random sample of $1 re#istered voters can be found in the
printout below.
Test of p 1 &.5 vs p 2 &.5
Success 1 for candidate
3ariable 4 5 Sample 6 .7act 6 value
8or candidate 59 $1 &.:;<;:; &.&&==
The conclusion which can be reached is>
*. The candidate can be 0uite confident that he has more than 5&% of the vote.
**. The p?value of &.&&== tells us that we cannot re@ect the null h+pothesis and that the candidate has 5&% or less of
the vote.
***. The p?value of &.&&== indicates that it is not ver+ li/el+ to #et an observed value of &.:;< if the null
h+pothesis is true.
(A) * onl+
()) ** onl+
(C) *** onl+
(-) * and **
(.) * and ***
;. *n each of the followin#, the p?value and the si#nificance level A are #iven for a test of h+pothesis. 'hich pair of values
warrants a re@ection of the null h+pothesis(
(A) p?value 1 &.&;12, A 1 &.&1
()) p?value 1 &.&=11, A 1 &.&5
(C) p?value 1 &.&52$, A 1 &.&5
(-) p?value 1 &.&:<=, A 1 &.&5
(.) p?value 1 &.1;29, A 1 &.1&
=. *f a $5% confidence interval is #iven b+ (9:.52, 9$.=9), which of the followin# could be a $$% confidence interval for the
same data(
*. ( 9:.$9, 9$.&2)
**. (9:.;<,9$.:;)
***. (9<,9$)
(A) * onl+
()) ** onl+
(C) *** onl+
(-) * and **
(.) ** and ***
5. 'hich of the followin# are the characteristics of a t?distribution curve(
*. The #raph of a t?distribution e7tends infinitel+ to the left and to the ri#ht and approaches the x?a7is as+mptoticall+
as x increases in absolute value.
**. There are an infinite number of different t?curves, each determined b+ a parameter called the de#rees of freedom.
***. The ma7imum point on the #raph of a t?distribution occurs at its mean which is alwa+s &.
(A) * onl+
()) ** onl+
(C) *** onl+
(-) * and *** onl+
(.) *, **, and ***
:. *n conductin# a h+pothesis test, the p?value is the
(A) 6robabilit+ of obtainin# a result as e7treme or more e7treme than the one obtained than the one obtained if the null
h+pothesis is true.
()) Si#nificance level of the test
(C) 6robabilit+ of ma/in# a T+pe * error
(-) 6robabilit+ of ma/in# a T+pe ** error
(.) 6robabilit+ that the null h+pothesis is true
<. A tire manufacturer claims that it has developed a new all?season radial tire for passen#er cars (e7cludin# SB3s) which has a
shorter s/id distance than the /nown mean s/id distance, 1=& feet, for all tires currentl+ available. A consumer #roup wishes
to test this claim. *f represents the true mean s/id distance for this new tire, which of the followin# states the null
h+pothesis and the alternative h+pothesis that the consumer #roup should test(
(A) "o> 2 1=& ft. "a> C 1=& ft.
()) "o> D 1=& ft. "a> E 1=& ft.
(C) "o> 1 1=& ft. "a> 2 1=& ft.
(-) "o> 1 1=& ft. "a> 1=& ft.
(.) "o> 1 1=& ft. "a> 1=& ft.
9. *n a sample surve+ of =5& residents of a #iven communit+, 19& of them indicated that the+ shop at the local mall at least once
per month. Construct a $5% confidence interval to estimate the true percenta#e of the residents who shop monthl+ at the local
mall.
(A) (&.;55, &.=55)
()) (&.;::, &.=;=)
(C) (&.;<<, &.=2;)
(-) (&.;9&, &.=2&)
(.) *t cannot be determined from the information #iven.
$. A h+pothesis test tests the null h+pothesis "o> 1 2& a#ainst the alternative h+pothesis "a> 2& at a si#nificance level of
alpha. *f the same procedure is used to test "o> 1 2& a#ainst "a> > 2& with the same data then the p?value will
(A) *ncrease
()) -ecrease
(C) Femain the same
(-) )e doubled
(.) )e decreased b+ one
1&. A radio tal/ show host with a lar#e listenin# audience is interested in the proportion p of adults in his listenin# area that thin/
the Bnited States should not be at war in Af#hanistan. "e as/s the 0uestion to his listeners, G-o +ou thin/ the Bnited States
should be at war in Af#hanistan(H "e as/s his listeners to phone in and vote G+esH if the+ a#ree that the Bnited States should
be at war in Af#hanistan and no if the+ disa#ree. If the 1&& people who phoned in 95% answered G+esH. 'hich of the
followin# assumptions for inference about a proportion usin# a confidence interval are violated(
(A) The data are a simple random sample from the population of interest.
()) The population is at least 1& times as lar#e as the sample.
(C) The sample siJe n is so lar#e that both the count of successes np and the count of failures n(1?p) are 1& or more.
(-) 'e are interested in inference about a proportion.
(.) There appear to be no violations.
11. 8ive tablets of headache medicine T too/ an avera#e of :5 seconds to dissolve with a standard deviation of 1.; seconds. 8ive
tablets of headache medicine A too/ an avera#e of :9 seconds to dissolve with a standard deviation of 1.2 seconds. *f we
assume that the dissolvin# times are normall+ distributed and the population standard deviation is the same for both medicine
A and medicine T, compute a $5% confidence interval for the difference in dissolvin# time (in seconds) between brands A
and T. Calculate usin# ei#ht de#rees of freedom.
(A) ; K 1.=<1
()) ; K 1.55&
(C) ; K 1.:9:
(-) ; K 1.925
(.) ; K 2.1$:
12. A bone marrow transplant from a close relative is a commonl+ used techni0ue in fi#htin# certain t+pes of leu/emia. A
hospital finds that such a procedure results in recover+ <&% of the time. The+ are e7perimentin# with a new procedure that
uses bone marrow produced in a laborator+. So far the+ have used the artificial bone marrow in a random sample of <5
patients and :& have recovered. *s there sufficient evidence to @ustif+ the hospitalLs claims that the e7perimental procedure
has a hi#her recover+ rate than the traditional procedure(
(A) There is sufficient evidence at A 1 &.&1
()) There is sufficient evidence at A 1 &.&5
(C) There is not sufficient evidence at A 1 &.&5
(-) There is not sufficient evidence at A 1 &.1&
(.) The sample siJe is not lar#e enou#h to answer this 0uestion.
1;. *n testin# the desi#n of a new tennis rac/et, a compan+ randoml+ selects 12 professional tennis pla+ers and counted the
number of aces each pla+er served when usin# each rac/et in a 2?minute period. The results are as follows>
Player Racket #1 Racket #2
1 1& 15
2 9 9
; 12 1&
= 1= 12
5 < $
: $ 11
< 15 15
9 9 =
$ 12 $
1& < 11
11 9 1&
12 1; 12
'hat is the number of de#rees of freedom associated with a t?test to determine if there is a difference between the number of
aces served usin# the two rac/ets(
(A) :
()) 1&
(C) 11
(-) 12
(.) 2=
1=. 8ind the value of the appro7imate test statistic for the #iven h+pothesis test.
"o> 1 59.2 x 1 5:.95
"a> 59.2 s 1 =.9&
A 1 &.&1 n 1 1:
(A) ?=.5&
()) ?1.125
(C) ?&.291
(-) 1.125
(.) =.5&
FREE RESPONSE: Show all +our wor/. *ndicate clearl+ the methods +ou use because +ou will be #raded on the
correctness of +our methods as well on the accurac+ and the completeness of +our results and e7planations.