In: Statistics and Probability
The accuracy of a census report on a city in southern California was questioned by some government officials. A random sample of 1215 people living in the city was used to check the report, and the results are shown below.
Ethnic Origin | Census Percent | Sample Result |
Black | 10% | 121 |
Asian | 3% | 47 |
Anglo | 38% | 471 |
Latino/Latina | 41% | 504 |
Native American | 6% | 59 |
All others | 2% | 13 |
Using a 1% level of significance, test the claim that the census distribution and the sample distribution agree.
(a) What is the level of significance?
State the null and alternate hypotheses.
H0: The distributions are the same.
H1: The distributions are the
same.H0: The distributions are different.
H1: The distributions are
different. H0: The
distributions are the same.
H1: The distributions are
different.H0: The distributions are
different.
H1: The distributions are the same.
(b) Find the value of the chi-square statistic for the sample.
(Round the expected frequencies to at least three decimal places.
Round the test statistic to three decimal places.)
Are all the expected frequencies greater than 5?
YesNo
What sampling distribution will you use?
Student's tnormal uniformbinomialchi-square
What are the degrees of freedom?
(c) Estimate the P-value of the sample test statistic.
P-value > 0.1000.050 < P-value < 0.100 0.025 < P-value < 0.0500.010 < P-value < 0.0250.005 < P-value < 0.010P-value < 0.005
(d) Based on your answers in parts (a) to (c), will you reject or
fail to reject the null hypothesis that the population fits the
specified distribution of categories?
Since the P-value > α, we fail to reject the null hypothesis.Since the P-value > α, we reject the null hypothesis. Since the P-value ≤ α, we reject the null hypothesis.Since the P-value ≤ α, we fail to reject the null hypothesis.
(e) Interpret your conclusion in the context of the
application.
At the 1% level of significance, the evidence is sufficient to conclude that census distribution and the ethnic origin distribution of city residents are different.At the 1% level of significance, the evidence is insufficient to conclude that census distribution and the ethnic origin distribution of city residents are different.
Solution:
We have to test the claim that the census distribution and the sample distribution agree.
Part a) What is the level of significance?
The level of significance = 1% = 0.01
State the null and alternate hypotheses.
H0: The distributions are the same.
Vs
H1: The distributions are different
Part b) Find the value of the chi-square statistic for the sample.
where
Oi = Observed Counts
Ei =Expected Counts = N * Census Percent
Thus we need to make following table:
Ethnic Origin | Census Percent | Oi : Sample Result | Ei : Expected Frequencies | Oi2 / Ei |
Black | 10% | 121 | 121.500 | 120.50206 |
Asian | 3% | 47 | 36.450 | 60.60357 |
Anglo | 38% | 471 | 461.700 | 480.48733 |
Latino/Latina | 41% | 504 | 498.150 | 509.91870 |
Native American | 6% | 59 | 72.900 | 47.75034 |
All others | 2% | 13 | 24.300 | 6.95473 |
N = 1215 |
Thus
Are all the expected frequencies greater than 5?
Yes
What sampling distribution will you use?
Chi-square
What are the degrees of freedom?
df = k - 1
We have k = 6 Ethnic Origin, thus
df = 6 - 1
df = 5
Part c) Estimate the P-value of the sample test statistic
Look in Chi-square table for df = 5 row and find the interval in which Chi-square test statistic value = fall.
Then find corresponding right tail area, which would be P-value interval.
fall between 11.070 and 12.833 and these values corresponds to area 0.050 and 0.025 respectively
Thus range of P-values is ( 0.025 , 0.050)
thus we get: 0.025 < P-value < 0.050
Part d) Based on your answers in parts (a) to (c), will you reject or fail to reject the null hypothesis that the population fits the specified distribution of categories?
Since 0.025 < P-value < 0.050, which implies P-value > 0.01 level of significance, thus we fail to reject H0.
Thus correct option is:
Since the P-value > α, we fail to reject the null hypothesis
Part e) Interpret your conclusion in the context of the application.
At the 1% level of significance, the evidence is insufficient to conclude that census distribution and the ethnic origin distribution of city residents are different.