In: Statistics and Probability
In 2008, Devry “University” infamously collected data comparing the starting salaries of graduating students with surnames beginning with the letters A through M with those whose surnames begin with N through Z. The study was famously panned. For a sample of 30 students in the A-M category, the avg starting salary was $37,233.33, with a standard deviation of $3475.54. For a sample of 36 students with surnames beginning with N-Z, the average starting salary was $35855.81, with a standard deviation of $2580.02. Test to see in the populations are actually equal using a 2% significance level.
To test to see in the populations are actually equal using a 2% significance level,
we conduct a hypothesis test to see if the population means are equal or not
n | Mean (M) | Variance (SS) | Standard Deviation (SD) | |
X | n1 = 30 | M1 = 37233.33 | SS1 = 12079378.2916 | s1 = 3475.54 |
Y | n2 = 36 | M2 = 35855.81 | SS2 = 6656503.2004 | s2 = 2580.02 |
The null and alternative hypotheses are
Ho : μ1 =
μ2
(Populations are equal)
Ha : μ1 ≠
μ2
(Populations are not equal)
where μ1, μ2 are population means for starting salaries of
the two categories (surnames A-M and N-Z)
respectively.
α =
0.02
2% level of significance.
We use independent Samples T test since population standard
deviation is unknown.
Using the given formulae, we get
Degrees of
Freedom
df1 = n1 - 1 , df2 = n2 - 1 , df = n1 + n2 - 2 = 30 + 36 - 2 =
64
df = 64
Pooled
Variance
Sp2 = 9113743.4761
Mean Squared Error
Sm1-m2
Sm1-m2 = 746.2915
t-statistic
t-statistic = t = 1.8458
P-value
For t = 1.8458, df = 64 we find the two tailed p-value using t
tables or Excel function t.dist.2t
p-value = t.dist.2t(abs(1.8458), 64)
p-value = 0.0695
Decision
0.0695 > 0.02
that is p-value is greater than α.
Hence we DO NOT Reject Ho.
Conclusion
There does not exist enough statistical evidence at α = 0.02 to
show that the samples are from different populations.
In other words,
The populations are actually equal.