You are on page 1of 11

Tutorial 6

1. A civil engineer is analyzing the compressive strength of concrete. Compressive strength


is approximately normally distributed with variance, 2 = 1000psi2. A random sample of
12 specimens has a mean compressive strength of x =3525psi.

i. Test the hypothesis that mean compressive strength is 3500psi. Use =0.01;
ii. What is the smallest level of significance at which you would be willing to reject
the null hypothesis?;
iii. Construct a 95% two-sided CI on mean compressive strength; and
iv. Construct a 99% two-sided CI on mean compressive strength. Compare the
width of this confidence interval with the width of the one in part (iii). What is your
comment?

Solution:

(a) The parameter of interest is to test the true mean, μ.


(b) The hypothesis Testing:
H 0 :   3500
H 1 :   3500
(c) The significance level α = 0.01
(d) The test statistics is:

Computation

x  3525,  2  1000
3525  3500
 z0   2.74
1000 / 12

(e) Decision:
Reject H0 if z0 > z , where z0.005 = 2.58

(f) Result and conclusion:


Since 2.74 > 2.58, so we reject the null hypothesis and conclude the true mean is
significantly different from 3500 at α = 0.01.

ii. The P-value = 2[1- (2.74)] = 2[1-= 0.0062

iii. A 95% two-sided lower CI on mean is

x  3525,  2  1000, n  12, z 0.025  1.96


     
x  z / 2      x  z / 2  
 n  n
  
3525  (1.96) 1000 / 12    3525  (1.96) 1000 / 12 
3507    3543
iv. A 99% two-sided lower CI on mean is

x  3525,  2  1000, n  12, z 0.005  2.58


     
x  z / 2      x  z / 2  
 n  n
  
3525  (2.58) 1000 / 12    3525  (2.58) 1000 / 12 
3501    3549

oooOOOooo

Tutorial 6/RR/Sept. 2013


2. One of the Cigarette Company claims that their cigarettes contain an average of only
10mg. of tar. A random sample of 25 cigarettes shows the average tar content to be
12.5mg with standard deviation of 4.5mg.

i. Construct a hypothesis test to determine whether the average tar content of


cigarettes exceeds 10mg. using the P-value approach;
ii. Construct a 95% two-sided CI on the average tar content of cigarettes.

Solution:

i.
(a) The parameter of interest is to test the true average tar content, μ.
(b) The hypothesis testing:

H 0 :   10mg vs H1 :   10mg

(c) The test statistics is:

x 12.5  10
T   2.78
s/ n 4.5 / 25

(d) Decision:

From a t-distribution table, for a t – distribution with 24 degree of freedom, that T =2.78
is falls between two values: 2.797 for which =0.005 and 3.091 for which =0.0025. So
the P-value is: 0.0025 < P < 0.005. Since P < 0.05, thus we reject H0 and conclude that
the true mean tar content is significantly exceeds 10mg. at α = 0.05

ii. A 95% two-sided CI on mean is


x  12.5, s  4.5, n  25, t 0.025, 24  2.064
 s   s 
x  t / 2.,24      x  t / 2, 24  
 n  n
 
12.5  (2.064) 4.5 / 25    12.5  (2.064) 4.5 / 25  
10.64    14.36

oooOOOooo

Tutorial 6/RR/Sept. 2013


3. A survey done one year ago showed that 45% of the population participated in recycling
programs. In a recent poll a random sample of 1250 people showed that 588 participate
in recycling programs.

i. Test the hypothesis that the proportion of the population who participate in
recycling programs is greater than 45% was one year ago. Use a 5% significance
level.
ii. Construct a 95% two-sided CI on the proportion.
Solution:
i.
(a) The parameter of interest is test about proportion, p.
(b) The hypothesis testing:
H 0 : p  0.45
vs
H 1 : p  0.45
(c) Test statistics is:
pˆ  X / n  p0
z0 
p0 (1  p0 ) / n
(d) Computation:

588 / 1250  0.45


z0   1.45
(0.45)(0.55) / 1250
(e) Decision:
Reject H0 if z0 > z , where z0.05 = 1.65
(f) Result and conclusion:
Since Zo < 1.65, thus we cannot reject the H0 and conclude that more than 45% of the
population participated in recycling programs as claim is not true

iii. A 95% two sided CI for the proportion is

Given, n  1250, x  588, , pˆ  0.47, Z  / 2  1.96

pˆ  z pˆ (1 pˆ )  p  pˆ  z pˆ (1 pˆ )
 /2 n  /2 n
 0.47(0.53)   0.47(0.53) 
0.47  (1.96)   pˆ  0.47  (1.96)
 

 1250   1250 
0.44  pˆ  0.497

oooOOOooo

Tutorial 6/RR/Sept. 2013


4. An Aerospace Engineers claim that the standard deviation of the percentage in an alloy
used in aerospace casting is greater than 0.3. 51 parts were randomly selected and the
sample standard deviation of the percentage in an alloy used in aerospace casting is
s =0.37.

i. At α = 0.05, do these data support the claim of the engineers?


ii. What is the P-value for this test?
iii. Construct a 95% two-sided CI for . What is conclusion?

Solution:
i.
(a) The parameter of interest is to test the variance of sugar content, .
(b) The hypothesis testing:
H 0 :  2  0.09 vs H1 :  2  0.09
(c) The test statistics is:
(n  1) s 2
 o2 
2
(d) Critical value, = 67.505

(e) The critical region is reject H0 if 67.5


(f) Computation

n  51,  2  0.09 , s 2  0.1369


50(0.1369)
 o2   76.06
0.09
(g) Result and conclusion:
Since  > 67.5, so we reject H0 and conclude that an Aerospace Engineers claim that
the standard deviation of the percentage in an alloy used in aerospace casting is greater
than 0.3 is true.

ii. The P-value for this test is 0.01<p<0.025. Since P-value < 0.05, then we reject H0. There
is enough evidence to support the claim that the standard deviation of the percentage
in an alloy used in aerospace casting is greater than 0.3.

iii. A 95% two-sided CI on the variance is:


(n  1) s 2 (n  1) s 2
 2 
  / 2, n 1
2
 12 / 2, n 1
50(0.1369) 50(0.1369)
 2 
71.42 32.36
0.096    0.21
2

From the 95% CI, since  = 09 is falls outside the interval, then we reject Ho. So the
claim that the standard deviation of the percentage in an alloy used in aerospace casting
is greater than 0.3.
oooOOOooo

Tutorial 6/RR/Sept. 2013


5. Two machines are used to fill plastic bottles with dishwashing detergent. The standard
deviations of fill volume are known to be 10.01 and = 0.15 fluid ounce for two
machines, respectively. Two random samples of n1 = 12 bottles from machine 1 and
n2=10 bottles from machine 2 are selected, and the sample mean fill volumes are
x 1 =30.61 and x 2 =30.24 fluid ounces. Assume normality.

i. Test the hypothesis that both machines fill to the same mean volume. Use the P-
value approach;
ii. Construct a 90% two-sided CI on the mean difference in fill volume; and
iii. Construct a 95% two-sided CI on the mean difference in fills volume. Compare
and comment on the width of this interval to the width of the interval in part (ii).
Solution:
i.
(a) Hypothesis testing between two means, and variances are unknown but
large samples size.
(b) The hypothesis testing:

H 0 : 1   2  0 vs H 1 : 1   2  0
(c) The test statistics is:

( x1  x 2 )  ( 1   2 )
z0 
 12  22

n1 n2
30.61  30.24
  7.79
0.012 0.15 2

12 10
(d) P-value = 2[1 - (7.79)] = 2[1 – 1] = 0. Since P-value < 0.05, then we have
to reject H0 and we conclude that there are difference in means at α = 0.05.

ii. A 90% CI for the difference in mean and is

x 1  30.61, x 2  30.24, s1  0.01, s 2  0.15, n1  12, n 2  10


 12  22  12  22
( x 1  x 2 )  z / 2   1   2  ( x 1  x 2 )  z / 2 
n1 n2 n1 n2
 0.012 0.15 2   2 2 
(30.61  30.24)  1.65        (30.61  30.24)  1.65 0.01  0.15 
 12 10  1 2
 12 10 
   
0.29  1   2  0.45
iii. A 95% CI for the difference in mean and is

x 1  30.61, x 2  30.24, s1  0.01, s 2  0.15, n1  12, n 2  10


 12  22  12  22
( x1  x 2 )  z / 2   1   2  ( x 1  x 2 )  z  / 2 
n1 n2 n1 n2
 0.012 0.15 2   2 2 
(30.61  30.24)  1.96        (30.61  30.24)  1.96 0.01  0.15 
 12 10  1 2
 12 10 
   
0.28  1   2  0.46

From CI, since     0 is not in the interval, so we reject H0. Strong evidence to
1 2

support that and 2 is difference


6. Professor A claims that a probability and statistics student can increase his or her score
on tests if the person is provided with a pre-test the week before the exam. To test her
theory she selected 16 probability and statistics students at random and gave these
students a pre-test the week before an exam. She also selected an independent random
sample of 12 students who were given the same exam but did not have access to the
pre-test. The first group had a mean score of 79.4 with standard deviation 8.8. The
second group had sample mean score 71.2 with standard deviation 7.9.

i. Do the data support Professor A claims that the mean score of students who get
a pre-test are different from the mean score of those who do not get a pre test
before an exam. Use the P-value approach and assume that their variances are
not equal.
ii. Construct a 95% two-sided CI for the difference in mean score of students who
get a pre-test and those who do not get a pre-test before an exam. Interpret this
interval.

Solution:
i.
(a) The parameter of interest is to test the different between the two means, 1and
2variance unknown but not equal.
 (b) Hypothesis testing:
H 0 : 1   2  0 vs H 1 : 1   2  0
(c) Test statistics is:
( x1  x 2 )  ( 1   2 )
T
s12 s22

n1 n2
(d) Degree of freedom is:
2 2
 s12 s 22   8.8 2 7.9 2 
     
 n1 n2   16 12 
v   25
s12 / n2 2  s s2 / nw 2  8.8 2  2 / 15   7.9 2 
2

 / 11
n1  1 n2  1  16   12
   

(e) Computation:
x1  79.4, x 2  71.2, s1  8.8, s 2  7.9, n1  16, n 2  12
So
(79.4  71.2)
T  2.59
8.8 2 7.9 2

16 12
From a t-distribution table, for a t – distribution with 25 degree of freedom, that T =2.59
falls between two values: 2.484 for which =0.01 and 2.612 for which =0.0075. So the
P-value is : 2( 0.0075 < P < 0.01) = 0.005<P<0.02.

(f) Result and conclusion:


Since P < 0.05 thus we reject H0 and conclude that the means are difference at  =0.05.
ii. A 95% two-sided CI for the difference in mean and 2is

x1  79.4, x 2  71.2, s1  8.8, s 2  7.9, n1  16, n2  12


s12 s 22 s2 s2
( x 1  x 2 )  t / 2 ,v   1   2  ( x1  x 2 )  t / 2,v 1  2
n1 n2 n1 n2
 8.8 2 7.9 2   2 2 
(79.4  71.2)  2.06    1   2  (79.4  71.2)  2.06 8.8  7.9 
 16 12    12 
  16
1.67  1   2  14.73

From CI, since 1  2  0 is not in the interval, so we reject H0.

oooOOOooo

Tutorial 6/RR/Sept. 2013


7. The rollover rate of sport utility vehicles is a transportation safety issue. Safety
advocates claim that the manufacturer A’s vehicle has a higher rollover rate than that of
manufacturer B. One hundred crashes for each of this vehicle were examined. The
rollover rates were pA=0.35 and pB=0.25.

i. By using the P-value approach, does manufacturer A’s vehicle has a higher
rollover rate than manufacturer B’s?
ii. Construct a 95% two sided CI on the difference in the two rollover rates of the
vehicle. Interpret this interval.

Solution:
i.
(a) The parameters of interest are to test the proportion of the vehicle rollover rate
manufactured by manufacturer A, pA and manufacturer, pB
 (b) Hypothesis testing:
H 0 : p A  pB vs H1 : p A  pB
(c) Test statistics is:

( pˆ A  pˆ B )  ( p A  p B ) x A  xB
z0  , where pˆ 
 1 1  n A  nB
pˆ (1  pˆ )  
 n A nB 

(d) Computation:
nA  100, nB  100, x A  35, xB  25, pˆ A  0.35, pˆ B  0.25
0.35  0.25 35  25
 z0   1.54 , where pˆ   0.3
 1 1  100  100
0.3(1  0.3)  
 100 100 

P-value=1-(1.54)=1-0.9382=0.062

(e) Result and conclusions:


Since the P-value > 0.05, thus we cannot reject the null hypothesis. There is not enough
evidence to claim that the proportion of the vehicle rollover rate manufactured by
manufacturer A, pA is greater than manufacturer, pB
ii. A 95% CI on the difference in the two proportions of the vehicle rollover rate
manufactured by manufacturer A, pA and manufacturer, pB
n A  100, nB  100, x A  35, xB  25, pˆ A  0.35, pˆ B  0.25
pˆ A (1  pˆ A ) pˆ B (1  pˆ B ) pˆ A (1  pˆ A ) pˆ B (1  pˆ B )
( pˆ A  pˆ B )  z / 2   p A  p B  ( pˆ A  pˆ B )  z / 2 
nA nB nA nB
 (0.35)(0.65) (0.25)(0.75) 
(0.35  0.25)  (1.96)    p A  pB 

 100 100 
 (0.35)(0.65) (0.25)(0.75) 
(0.35  0.25)  (1.96)  

 100 100 
- 0.026  p A  p B  0.226

From CI, since p  p  0 is in the interval, so we cannot reject H0.


A B

oooOOOooo

Tutorial 6/RR/Sept. 2013


8. A professor wanted to study the effect of using a computer based learning (CBL) to
teach statistics. In his first class of 43 students, he used the CBL as an instructional aid
in class. In his second class of 35 students, he taught without CBL, On the common final
exam, the first class scored a mean of 78.32, with standard deviation of 8.07; while the
second class scored a mean of 80.41, with standard deviation of 8.53.

i. Test if the use of the CBL makes any difference in mean scores. Use the P-value
approach.

iii. Construct a 95% CI for the difference in mean score of students when he used
the CBL and when he did not used the CBL.

Solution:
i.
(a) Hypothesis testing between two means, mean score with CBL and mean score 
without CBL:
H 0 :  A   B  0 vs H1 :  A   B  0
(b) Since the samples size are large, so the test statistics is:

( x  x )  (   )
z 
A B
A B
0 2 2
s s
 A B

n n A B

78.3  80.41 2.11


   1.11
8.07 2
8.53 2
1.89

43 35
P-value = 2[1 - (1.11)] = 2[1 – 0.8643] = 0.27.

(c) Result and conclusions:


Since P-value > 0.05, then we fail to reject H0 and we conclude that there are no
difference in means.
ii. A 95% two-sided CI for the difference in mean and 2is

x A  78.3, x B  80.41, s A  8.07, s B  8.53, n A  43, n B  35


s12 s 22 s2 s2
( x A  x B )  z / 2    A   B  ( x A  x B )  z / 2 A  B
n1 n2 n A nB
 8.07 2 8.53 2   2 2 
(78.3  80.41)  1.96     A   B  (78.3  80.41)  1.96 8.07  8.53 
 43 35    35 
  43
- 5.83  1   2  3.92

From CI, since 1  2  0 is in the interval, so we cannot reject H0.

oooOOOooo

Tutorial 6/RR/Sept. 2013