You are on page 1of 10

1

Your Name
STAT 250-0xx (your correct section)

Data Analysis Assignment 4

Problem 1

I. Null Hypothesis, H0:p 0.5

Alternative Hypothesis, Ha: p>0.5

ii.The significance level for this problem is alpha =0.05.

iii. Assumption 1: The sample was independent and random.

Assumption 2:Large Sample size,np =0.5*256 = 128,and 12810

n(1-p)=256*(1-0.5) = 128 10

Assumption 3: We can assume that there are atleast 256*10 =2560 games were played during

the 2016 NFL regular season.

iv.A z-test for proportion will be used.

n=256

p_hat = 147/256 = 0.5742

0.57420.5
z= = = 2.38
(1) 0.5(10.5)

256 256

v.Using the standard normal table, the p-value obtained was 0.0088

vi.We will reject the null as the p-value is less than the significance level of 0.05.
2

vii.We can conclude that the home team wins more than 50% of games in the National

Football League.

viii.Using Stat Crunch

The result is shown below

b)
3

c)

I am surprised with the result as the after simulating with 1 test; the result showed not to

reject the null which is contradictory to our result being obtained in part a.

d)

The result after running 1000 tests showed that the null hypothesis is rejected in only 44 of

the test when the sample size of 256 is considered.The z-statistic histogram plot shows

different z-value after running 1000 tests.Similarly,the histogram plot of p-value shows

different p-value obtained after running 1000 tests.

e) The first red number is 30th run which shows that the null hypothesis is rejected in the 30th

test.
4

f)Yes, I am surprised because the null hypothesis is rejected in only 44 of the samples out of

1000 runs.Thus,similating the test 1000 times that home team does not wins more than 50%

of the game in the NFL.

g) Running 1000 tests when p=0.6


5

When changing the sample came from the true proportion of 0.6, running 1000 tests lead to

the rejection of null hypothesis as opposed to running 1000 tests when we assumed sample

came from a true population proportion is 0.5.The difference is caused due to change in

proportion characteristics.

Problem 2

a)Sample proportion, p_hat =36/55 = 0.6545

_(1 ) 0.6545(10.6545)
Standard error of the sample,SE = = = 0.0641
55

Critical z-value of 95% confidence interval =1.96

Margin of error=Critical z-value*SE =0.0641*1.96 = 0.1257

95% Confidence interval = p_hat Margin of error = 0.6545 0.1257 = (0.5289, 0.7802)

Stat Crunch Output


6

The calculated result is in agreement with the Stat Crunch Output.

b)

i.Null Hypothesis,H0:p=0.588

Alternative Hypothesis,Ha:p0.588

ii.Significance level is assumed at 0.05(alpha).

iii. Assumption 1: The sample was independent and random.

Assumption 2: Large Sample size,np =0.588*55 = 32.34,and 32.3410

n(1-p)=55*(1-0.588) = 22.66 10

Assumption 3:We can assume that there are at least 55*10 =550 males between the ages of

20 and 39..

iv. A z-test for proportion will be used.

n=55

p_hat = 36/55 = 0.6545

0.65450.588
z= = = 1.0027
(1) 0.588(10.588)

256 55

v.Using the standard normal table,the p-value obtained was 0.3160


7

vi.We will not reject the null as the p-value is greater than the significance level of 0.05.

vii.We can conclude that the percentage of a male between the ages of 20 and 39 who

consume the recommended daily allowance of calcium has not changed.

viii.Using Stat Crunch

c)Both the result confirmed the fact that the percentage of male between the ages of 20 and

39 who consume the recommended daily allowance of calcium has not changed. The

confidence interval contained the true proportion of 0.588 in its interval indicated that the

proportion is not significantly different from 0.588 which is similar to result obtained using z-

test for proportion.


8

Problem 3

a) Using Stat Crunch sample proportion is 0.14107143.

b)

i.Null Hypothesis, H0:p = 0.14

Alternative Hypothesis,H0,p< 0.14

Where p is the proportion that the new newspaper will capture in order to be financially

viable.

ii.The significance level is denoted by alpha which is 0.02 as stated in the problem.

iii. Assumption 1: The sample was independent and random.

Assumption 2:Large Sample size,np =0.14*560 = 78.4,and 78.410

n(1-p)=560*(1-0.14) = 481.6 10

Assumption 3:We can assume that there are atleast 560*10 =5600 Toronto residents.

iv.Test statistic(z-test for proportion)

p_hat =0.14107143

p=0.14

n=560

(1) 0.14(10.14)
Standard error of the sample ,SE = = =0.0147
560

z-value = (p_hat-p)/SE = (0.14107143-0.14)/0.0147 = 0.0731

v.p-value using standard normal table is 0.53


9

vi.We will not reject the null as the p-value is greater than the assumed significance level of

0.02.

vii.The claim that new newspaper would have to capture at least 14$ of the Toronto market in

order to be financially viable is current.

viii.Stat Crunch Output is shown below

Problem 4

a)Histogram plot
10

The above histogram plot shows that the Fairfax home closing prices are skewed to the right.

b)In this scenarios the condition of large sample size is not met as the sample size is 10 which

is less than the central limit condition of large sample size of 25.Also,the distribution is

skewed.

c) In this case the central limit theorem of large sample size is met as sample size is 36 which

is greater than 25.

$700,000$510,000
P(X>$700,000) = ( > $145,000 ) = ( > 7.86)
36

Using normal distribution table

P(z>7.86) =1-P(z<7.86) =1-0.999999 =0.0000

You might also like