In: Statistics and Probability
Do women feel differently from men when it comes to tax rates? One question on a survey of randomly selected adults asked, "What percent of income do you believe individuals should pay in income tax?" Complete parts a) through c).
Gender Tax Rate Gender Tax
Rate
Female 10 Male 15
Female 10 Male 20
Female 4 Male 10
Female 21 Male 10
Female 20 Male 17
Female 15 Male 2
Female 8 Male 1
Female 15 Male 15
Female 3 Male 4
Female 5 Male 8
Female 23 Male 15
Female 10 Male 5
Female 0 Male 16
Female 19 Male 33
Female 10 Male 10
Female 0 Male 15
Female 7 Male 2
Female 20 Male 0
Female 14 Male 4
Female 13 Male 7
Female 20 Male 15
Female 15 Male 6
Female 9 Male 10
Female 14 Male 6
Female 5 Male 10
Female 13 Male 23
Female 15 Male 15
Female 5 Male 15
Female 17 Male 8
Female 23 Male 10
c) Test whether the mean tax rate for females differs from that of males at the alphaαequals=0.01 level of significance.
Determine the null and alternative hypotheses for this test. Let μM represent the mean income tax rate for males and let μF represent the mean income tax rate for females.
d) Find t0, the test statistic for this hypothesis test.
Ho :
Ha :
Level of significance (l.o.s.) : = 0.01
Test Statistics : t-Test: Two-Sample Assuming Equal Variances ( Student's t distribution is used since population variances are unknown )
Decision Criteria : Reject Ho at 1% l.o.s. if |
to | > t tab,
where t tab = t(/2,
n-1) = t( 0.005, 58) = 2.6633
Calculations : The results for this test is obtained using MS-Excel and for this, following steps are needed to be followed :
Step 1 : Enter the data in excel.
Step 2 : Select the "t-Test: Two-Sample Assuming
Equal Variances" function available in the data analysis package
under the data tab in MS-Excel.
Step 3 : Enter the range of the female tax rates
in the first cell and that of male tax rates in the second
cell.
Step 4 : Change the default l.o.s. from 0.05 to
0.01 and then press ok. Following results would be obtained :
t-Test: Two-Sample Assuming Equal Variances | ||
Males | Females | |
Mean | 10.9 | 12.1 |
Variance | 51.33448276 | 44.16206897 |
Observations | 30 | 30 |
Pooled Variance | 47.74827586 | |
Hypothesized Mean Difference | 0 | |
df | 58 | |
t Stat | -0.672586318 | |
P(T<=t) one-tail | 0.25194069 | |
t Critical one-tail | 2.392377475 | |
P(T<=t) two-tail | 0.50388138 | |
t Critical two-tail | 2.663286954 |
Thus, to = t Stat = -0.672586318
Conclusion : SInce | to | < t tab, we do not reject Ho at 1% l.o.s. and thus, conclude that the mean tax rate for females does not differs significantly from that of males.
Hope this answers your query!