In: Math
4. Suppose from our class (with 25 students), we find the
average GPA is 2.7, with standard deviation
0.4. Regard our class as a random sample from the whole university.
Based on the information from our
class, can we believe at significance level 0.05, that the average
GPA for all students can be at least 2.8?
The test of interest is
H0:
=2.8 vs H1:
>2.8
where
is the average GPA for all students in the university
H0 represents the hypothesis that the average GPA for all students of the university is at least 2.8
H1 represents the hypothesis that the average GPA for all students of the university is more than 2.8
Given
n = sample size = 25

= sample mean = 2.7
s = sample standard deviation = 0.4
= level of significance = 0.05
We see that population variance is unknown.
Therefore T = test statistic = (
-
)
/ s
t = observed value of T = (2.7 - 2.8)
/ 0.4 = -1.25
T ~ tn-1 distribution i.e T ~ t24 distribution
The test is a right-tailed test.
Therefore we reject H0 if t > t
,n-1
i.e if t > t0.05,24
From the t-distribution table we get the value that t0.05,24 = 1.711
Thus t < t0.05,24
Therefore we accept H0
Thus; based on the information from our class, we can believe that at significance level 0.05, the average GPA for all students of the university is at least 2.8.