In: Statistics and Probability
In 2010 the Maricopa Community College District's enrollment data showed the following breakdown of students by ethnicity: 54.9% White; 21.1% Hispanic; 7.9% Black; 4.5% Asian/Pacific Islander; 2.9% Native American; 8.8% Other. Information was collected from a random sample of 300 students in 2017 to determine whether or not the data has changed significantly. The sample data is given in the table below. At the alph=0.05 level of significance, test the claim that the ethnic breakdown of students at MCCCD has not changed significantly since 2010.
Which would be the correct hypothesis for this test?
H0: u1 = u2; H1: u1 ≠ u2
H0: p1 = p2; H1: p1 ≠ p2
H0: The breakdown of students by ethnicity has not changed significantly since 2010 (i.e. the given distribution still fits); H1: The breakdown of students by ethnicity has changed significantly since 2010 (i.e. the given distribution no longer fits)
H0: The breakdown of students by ethnicity has changed significantly since 2010 (i.e. the given distribution no longer fits); H1: The breakdown of students by ethnicity has not changed significantly since 2010 (i.e. the given distribution still fits)
Ethnicity of students in sample:
White - 160
Hispanic - 89
Black - 25
Asian/Pacific Islander - 11
Native American - 15
Other - 0
Test Statistic:
______________
Give the P-value
_____________
Solution:
Given: In 2010 the Maricopa Community College District's enrollment data showed the following breakdown of students by ethnicity: 54.9% White; 21.1% Hispanic; 7.9% Black; 4.5% Asian/Pacific Islander; 2.9% Native American; 8.8% Other.
We have to test the claim that the ethnic breakdown of students at MCCCD has not changed significantly since 2010.
Level of significance =
Part a) Which would be the correct hypothesis for this test?
H0: The breakdown of students by ethnicity has not changed significantly since 2010 (i.e. the given distribution still fits);
H1: The breakdown of students by ethnicity has changed significantly since 2010 (i.e. the given distribution no longer fits)
Part b)
Given:
Ethnicity of students in sample:
White - 160
Hispanic - 89
Black - 25
Asian/Pacific Islander - 11
Native American - 15
Other - 0
We have to find test statistic:
where Oi = Observed frequencies and Ei = Expected frequencies.
To get Ei, we multiply given % values by N = 300
Thus we need to make following table:
Ethnicity of students | Expected % | Ei | Oi | Oi^2 /Ei |
White | 54.90% | 164.7 | 160 | 155.434 |
Hispanic | 21.10% | 63.3 | 89 | 125.134 |
Black | 7.90% | 23.7 | 25 | 26.371 |
Asian/Pacific | 4.50% | 13.5 | 11 | 8.963 |
Naive American | 2.90% | 8.7 | 15 | 25.862 |
Other | 8.80% | 26.4 | 0 | 0.000 |
N=300 |
Thus Chi-square test statistic is:
Test statistic =
Part c) P-value:
df = k -1 = 6 - 1 = 5
Use excel command:
=CHISQ.DIST.RT( x , df)
where x = and df = 5
Thus
=CHISQ.DIST.RT( 41.765 , 5 )
= 0.0000
Thus P-value= 0.0000
Conclusion: Since P-value = 0.0000 < 0.05 level of significance, hence we reject H0 and conclude that: The breakdown of students by ethnicity has changed significantly since 2010 (i.e. the given distribution no longer fits)