You are on page 1of 7

Answer 1:

1.1) To determine if there a difference between living costs between Toronto and Mexico
City

Hypothesis:

Ho:Mean(Toronto)=Mean(mexico)

H1:Mean(Toronto)!=Mean(mexico)

Mean(Toronto)= 67336.36

SD(Toronto)= 2045.616

Mean(mexico)= 63481.82

SD(mexico) = 1594.251

Test statistics:

t=( 67336.36- 63481.82)/sqrt((2045.616^2/11)+(1594.251^2/11)) = 4.929293

Critical region:

T(0.01.11+11-2) =2.845

As 4.929293>2.845. SO, NULL Hypothesis is rejected.

Therefore, there is enough evidence that there is significant difference in the average
annual cost of supporting her family of four in the manner to which they are
accustomed between Toronto and Mexico City.

1.2)

The test statistic for two means is,

Here,

The degrees of freedom is given by, df = {n_1} + {n_2} - 2df=n1+n2−2


The formula for two sample t test equal variance is {S_P}SP be the pooled
variance obtained below,

Rejection rule for test statistic:


If p−value≤α , then the null hypothesis is rejected.
Answer 2)

X Y X-Y (X-Y)^2

20427 25163 -4736 22429696

27255 24625 2630 6916900

22115 12600 9515 90535225

23256 24588 -1332 1774224

21887 19267 2620 6864400

24255 20150 4105 16851025

19852 22500 -2648 7011904

23624 16667 6957 48399849

25885 26875 -990 980100

28999 35333 -6334 40119556

20836 16292 4544 20647936

14331 262530815

Confidence Interval

CI = d ± t a/2 * (Sd/ Sqrt(n))

Where,
d = ∑ di/n
Sd = Sqrt( ∑ di^2 – ( ∑ di )^2 / n ] / ( n-1 ) )
a = 1 - (Confidence Level/100)
ta/2 = t-table value

CI = Confidence Interval

d = ( ∑ di/n ) =14331/11=1302.818

Pooled Sd( Sd )= Sqrt [ 262530815- (14331^2/11 ] / 10 = 4938.22

Confidence Interval = [ 1302.818 ± t a/2 ( 2851.082/ Sqrt ( 11) ) ]

= [ 1302.818 - 3.169 * (1488.929) , 1302.818 + 3.169 * (1488.929) ]

= [ -3415.599, 6021.235 ]
Answer 3)

3.1)

Null Hypothesis H0:  P1 = P2

Alternative Hypothesis H1:  P1 > P2

Sample Proportion for machine 1, p1 = 38/191 = 0.1990

Sample Proportion for machine 2 , p2 = 21/202 =0.1040

Under H0, the test statistic is

The corresponding p value is 0.008, which is less than the value of  = 0.05

(P-value = 0.008) < ( = 0.05)

Therefore, by the rejection rule, it can be concluded that there is evidence to reject H0 at  = 0.05

Hence there is difference between the proportion of sheets drilled with defective holes between
machine 1 and machine 2

3.2)

The P-Value is 0.0042

Since p value is less than significance level. Reject H0.

Conclusion : the propotion of sheets drilled with defective holes is significantly larger for machine 1.

Answer 4)

Confidence = 90%
Margin of error E = 100
Bill range (600,2500)
Range R = 2500-600 = 1900 USD
The standard deviation is not given so an estimate must be used.
 = 1900/4 = 475 (approximately)

Z score for 90% confidence interval = 1.645

E = (sd*z)/sqrt(n)
or, 100 = (475*1.65)/sqrt(n)
or, n = (475*1.645)^2/100^2 = 61.02

So sample size of 61 restaurants should be taken

Answer 5)

For given data n=15


Mean=45.267
SD=1.75

For 99% confidence and


df = 15-1 = 14
2 (0.995,14) = 4.07468
2 (0.005,14) = 31.3193

(15-1)*(3.067)  2  (15-1)*(3.067)
31.3193 24.07468

1.37  2  10.54

The 99% confidence interval for variance is therefore (1.37,10.54)

Answer 6)

Let,
Proportion in sample study p' = p which is the package proportion of 24% Blue, 20% Orange, 16%
Green, 14% Yellow, 13% Red, 13% Brown.

To Test H0: p' = p

sample size=105+72+89+84+70+80=500

n=500

We have to use two-tailed test

a.) BLUE=> p=.24

p'=105/500=.21

z=(p'-p)/sqrt(p*(1-p)/n)=-1.5707

alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96
1.5707< 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

b.) BROWN=> p=.13

p'=72/500=.144

z=(p'-p)/sqrt(p*(1-p)/n)=0.9308

alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96

0.9308< 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

c.) GREEN=> p=.20

p'=89/500=.178

z=(p'-p)/sqrt(p*(1-p)/n)=-1.23

alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96

1.23 < 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

d.) Orange=> p=.16

p'=84/500=.178

z=(p'-p)/sqrt(p*(1-p)/n)=0.488

alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96

0.488 < 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

e.) Red=> p=.13

p'=70/500=.14

z=(p'-p)/sqrt(p*(1-p)/n)=0.6649
alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96

0.6649 < 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

f.) Yellow=> p=.14

p'=80/500=.16

z=(p'-p)/sqrt(p*(1-p)/n)=1.289

alpha=0.05, alpha/2=0.025

1-alpha/2=0.975, z(alpha/2)=1.96

1.289 < 1.96

Since |z|<z(alpha/2), the null hypothesis is true.

Since the hypothesis is true for all colours, there is enough evidence to say that the
proportion of different colour candies in the package is correct

Alternate Solution using Chi Square

Ho = Given proportions are tru


Ha = one of the proportion differs

A = 0.05 (level of significance

DoF= n-1=6-1=5

Category Observed Proportion Expected Contriution to


Chi Square
Blue 105 .24 120 1.875
Brown 72 .13 65 .75
Green 89 .20 100 1.21
Orance 84 .16 80 .20
Red 70 .13 65 .384
Yellow 80 .14 70 1.42
Sum = 5.852
2 (0.05,5) = 11.071
2 obs = 5.852

Since

2 obs < 2 (0.05,5)

So null is accepted at significance level of 0.05, there is enough evidence to say that the
proportion of different colour candies in the package is correct

You might also like