In: Statistics and Probability
To test ?0: p = 0.4; ?1: p ≠ 0.4, a simple random sample of size n = 1000 is obtained from a population such that ? ≤ 0.05?. (a) If x = 420 and n = 1000, compute the test statistic 0 z . (b) Test the hypothesis using (i) the classical approach and (ii) the P-value approach. Assume an ? = 0.01 level of significance. (c) What is the conclusion of the hypothesis test?
Solution:
Given:
?0: p = 0.4; ?1: p ≠ 0.4
n = 1000
x = 420
Part a) Compute the test statistic z
where
thud
Part b) Test the hypothesis using
(i) the classical approach:
level of significance = ? = 0.01
Since this is two tailed test , find
Look in z table for Area = 0.0050 and find corresponding z value.
Area 0.0050 is in between 0.0049 and 0.0051, and both the area are at same distance from 0.005
thus we look for both area and find both z values.
Area 0.0049 corresponds to -2.5 and 0.08 , thus z= -2.58
Area 0.0051 corresponds to -2.5 and 0.07 , thus z= -2.57
Thus average of both z values is = ( -2.57 + -2.58 ) / 2 = -2.575
Thus critical z value is = -2.575
Since this is two tailed test , there are two z critical values = ( -2.575 , 2.575 )
Decision Rule:
Reject null hypothesis ,if z test statistic value < z critical
value=-2.575 or z test statistic value > z critical value=2.575
, otherwise we fail to reject H0.
Since z test statistic value = is neither < -2.575 , nor > 2.575, that is it does not fall in rejection region, we fail to reject H0.
ii) the P-value approach.
i) For right tailed test , p-value is:
p-value = P(Z > z test statistic)
ii) For left tailed test , p-value is:
p-value = P(Z < z test statistic)
iii) For two tailed test , p-value is:
p-value = 2* P(Z > z test statistic) if z is positive
p-value = 2* P(Z < z test statistic) if z is negative
Since this is two tailed test and z is positive, we use:
p-value = 2* P(Z > z test statistic)
p-value = 2* P(Z > 1.29)
p-value = 2* [ 1 - P(Z < 1.29) ]
Look in z table for z = 1.2 and 0.09 and find corresponding area.
P( Z< 1.29) = 0.9015
thus
p-value = 2* [ 1 - P(Z < 1.29) ]
p-value = 2* [ 1 - 0.9015 ]
p-value = 2* 0.0985
p-value = 0.1970
Decision Rule:
Reject null hypothesis H0, if P-value < 0.01 level of
significance, otherwise we fail to reject H0
Since p-value = 0.1970 > 0.01 level of significance, we fail to reject H0
Part c) What is the conclusion of the hypothesis test?
At 0.01 level of significance, we do not have sufficient evidence to reject the null hypothesis that p =0.4
that is : population proportion is not different from 0.4