In: Statistics and Probability
Is the proportion of deaths from heart disease less in the Medication group than in the Placebo group? Test at the 5% significance level. Heart disease and Medication data:
Group Died of Heart disease Subjects
Medication 118 2900
Placebo 100 2100
Ho:
Ha:
Test statistic:
p-val:
Statistical Conclusion:
Conclusion in plain English:
Answer)
Null hypothesis Ho : p1 = p2
Alternate hypothesis Ha : p1 < p2
N1 = 2900, P1 = 118/2900
N2 = 2100, P2 = 100/2100
First we need to check the conditions of normality that is if n1p1 and n1*(1-p1) and n2*p2 and n2*(1-p2) all are greater than equal to 5 or not
N1*p1 = 118
N1*(1-p1) = 2782
N2*p2 = 100
N2*(1-p2) = 2000
All the conditions are met so we can use standard normal z table to conduct the test
Test statistics z = (P1-P2)/standard error
Standard error = √{p*(1-p)}*√{(1/n1)+(1/n2)}
P = pooled proportion = [(p1*n1)+(p2*n2)]/[n1+n2]
After substitution
Test statistics z = -1.18
From z table, P(Z<-1.18) = 0.119
So, P-Value = 0.119
As the obtained P-Value is greater than the given significance 0.05.
We fail to reject the null hypothesis Ho.
We do not have enough evidence to conclude that proportion of deaths from heart disease less in the Medication group than in the Placebo group.