In: Statistics and Probability
let p1 and p2 be proportion of people who survived after cardiac arrest in day and night time resp.
Ho: p1 - p2 = 0
Ha: p1 - p2 > 0
first sample size, n1=
58593
number of successes, sample 1 = x1=
11604
proportion success of sample 1 , p̂1=
x1/n1= 0.198044135
second sample size, n2 =
28155
number of successes, sample 2 = x2 =
4139
proportion success of sample 1 , p̂ 2= x2/n2 =
0.147007636
difference in sample proportions, p̂1 - p̂2 =
0.051036499
pooled proportion , p = (x1+x2)/(n1+n2)=
0.1815
std error ,SE = =SQRT(p*(1-p)*(1/n1+
1/n2)= 0.0028
Z-statistic = (p̂1 - p̂2)/SE =
18.2609
p-value = 0.0000
decision : p-value<α=0.05,Reject null hypothesis
so, there is enough evidence to conlcude that true proportion of people who survived after cardiac arrest in day is greater than that of night time at α=0.05