In: Statistics and Probability
1) Suppose there is a random sample of n observations, divided into four groups. The table below summarizes the count of observations that were seen in each group.
Group 1 |
Group 2 |
Group 3 |
Group 4 |
16 |
46 |
51 |
37 |
We are interested in testing the null hypothesis H0:p1=p2=p3=p4=0.25, against the alternative hypothesis HA: At least one proportion is incorrect
a) What is the expected count for each of the groups?
Expected: (number)
b) What is the value of the test statistic? Round your response to at least 2 decimal places.
c) What are the appropriate degrees of freedom?
d) What is the P-value? Round to at least 4 decimal places
e) What conclusion can be made at the 5% level of significance?
A |
There is evidence that at least one proportion is not equal to 0.25. |
|
B |
There is no significant evidence that any of the proportions is not equal to 0.25. |
2) Suppose there is a random sample of 1,159 observations, divided into four groups. The table below summarizes the observations that were seen in each group.
Group 1 |
Group 2 |
Group 3 |
Group 4 |
556 |
197 |
127 |
279 |
We are interested in testing the Null hypothesis Observed=Expected, under the assumption that the expected proportions are .50, .20, .10, and .20 for the 4 groups, respectively.
a) What are the expected values?
Group 1 |
Group 2 |
Group 3 |
Group 4 |
b) What is the value of the test statistic? Round your response to at least 3 decimal places.
c) What is the P-value for the test? Round your response to at least 4 decimal places.
d) What conclusion can be made at the 5% level of significance?
A |
There is evidence against the null hypothesis, and therefore at least one of the observed proportions is not the same as the expected proportions. |
|
B |
There is no significant evidence against the null hypothesis, and therefore we do not have evidence the observed and expected proportions are different. |
Question 1
a) What is the expected count for each of the groups?
Group |
Observed |
Expected |
1 |
16 |
150*0.25 = 37.5 |
2 |
46 |
150*0.25 = 37.5 |
3 |
51 |
150*0.25 = 37.5 |
4 |
37 |
150*0.25 = 37.5 |
Total |
150 |
150 |
b) What is the value of the test statistic?
Test statistic formula is given as below:
Chi square = ∑[(O – E)^2/E]
Where, O is observed frequencies and E is expected frequencies.
Group |
O |
E |
(O - E)^2 |
(O - E)^2/E |
1 |
16 |
37.5 |
462.25 |
12.32666667 |
2 |
46 |
37.5 |
72.25 |
1.926666667 |
3 |
51 |
37.5 |
182.25 |
4.86 |
4 |
37 |
37.5 |
0.25 |
0.006666667 |
Total |
150 |
150 |
19.12 |
Chi square = ∑[(O – E)^2/E]
Test statistic = Chi square = 19.12
c) What are the appropriate degrees of freedom?
We are given N = 4
Degrees of freedom = df = N – 1 = 4 – 1 = 3
d) What is the P-value?
P-value = 0.000258215
(by using Chi square table)
e) What conclusion can be made at the 5% level of significance?
P-value < α = 0.05
So, we reject the null hypothesis
There is evidence that at least one proportion is not equal to 0.25.
Question 2
a) What are the expected values?
Group |
O |
Exp. Prop. |
Expected |
1 |
556 |
0.5 |
1159*0.5 = 579.5 |
2 |
197 |
0.2 |
1159*0.2 = 231.8 |
3 |
127 |
0.1 |
1159*0.1 = 115.9 |
4 |
279 |
0.2 |
1159*0.2 = 231.8 |
Total |
1159 |
1 |
1159 |
b) What is the value of the test statistic?
Test statistic formula is given as below:
Chi square = ∑[(O – E)^2/E]
Where, O is observed frequencies and E is expected frequencies.
Group |
O |
Exp. Prop. |
E |
(O - E)^2 |
(O - E)^2/E |
1 |
556 |
0.5 |
579.5 |
552.25 |
0.952976704 |
2 |
197 |
0.2 |
231.8 |
1211.04 |
5.224503883 |
3 |
127 |
0.1 |
115.9 |
123.21 |
1.063071613 |
4 |
279 |
0.2 |
231.8 |
2227.84 |
9.611044003 |
Total |
1159 |
1 |
1159 |
16.8515962 |
Chi square = ∑[(O – E)^2/E]
Test statistic = Chi square = 16.8515962
c) What is the P-value for the test?
We have N = 4, df = N – 1 = 3
P-value = 0.000758185
(By using Chi square table)
d) What conclusion can be made at the 5% level of significance?
P-value < α = 0.05
So, we reject the null hypothesis
There is evidence against the null hypothesis, and therefore at least one of the observed proportions is not the same as the expected proportions.