In: Statistics and Probability
a die is tested for fairness using Ha : p does not equal 1/6 where p is the proportions of the sixes that occur . Would ubserving 12 sixes in 114 rolls of the die be statistically significant evidence that the true probability of a six is different from 1/6? use 10% significance level.
Solution:
Here, we have to use one sample z test for the population proportion.
The null and alternative hypotheses for this test are given as below:
Null hypothesis: H0: A die is fair.
Alternative hypothesis: Ha: A die is not fair.
H0: p = 1/6 = 0.1667 versus Ha: p ≠ 0.1667
This is a two tailed test.
We are given
Level of significance = α = 0.10
Test statistic formula for this test is given as below:
Z = (p̂ - p)/sqrt(pq/n)
Where, p̂ = Sample proportion, p is population proportion, q = 1 - p, and n is sample size
x = number of items of interest = 12
n = sample size = 114
p̂ = x/n = 12/114 = 0.105263158
p = 0.1667
q = 1 - p = 0.8333
Z = (p̂ - p)/sqrt(pq/n)
Z = (0.105263158 - 0.1667)/sqrt(0.1667*0.8333/114)
Z = -1.7600
Test statistic = -1.76
P-value = 0.0784
(by using z-table)
P-value < α = 0.10
So, we reject the null hypothesis
There is not sufficient evidence to conclude that the die is fair.