In: Statistics and Probability
A clinical trial for hydroxychloroquine was performed and hailed
by the leader of a large nation as a huge success. Perhaps the
biggest success ever!
In this trial only 292 people who received hydroxychloroquine died,
compared to the placebo group, where 837 people died. A total of
3,024 people received hydroxychloroquine, and 11,536 received the
placebo.
Plot a graph that best communicates the results of this trial. Do
you agree that these results show that hydroxychloroquine is a
useful treatment for SARS-CoV-2?
Ho: p1 - p2 = 0
Ha: p1 - p2 > 0
sample #1 ----->
first sample size, n1=
3024
number of successes, sample 1 = x1=
292
proportion success of sample 1 , p̂1=
x1/n1= 0.0965608
sample #2 ----->
second sample size, n2 =
11536
number of successes, sample 2 = x2 =
837
proportion success of sample 1 , p̂ 2= x2/n2 =
0.072555
difference in sample proportions, p̂1 - p̂2 =
0.0966 - 0.0726 =
0.0240
pooled proportion , p = (x1+x2)/(n1+n2)=
0.0775412
std error ,SE = =SQRT(p*(1-p)*(1/n1+
1/n2)= 0.00546
Z-statistic = (p̂1 - p̂2)/SE = ( 0.024
/ 0.0055 ) = 4.393
z-critical value , Z* =
1.6449 [excel function =NORMSINV(α)]
p-value = 0.0000 [excel
function =NORMSDIST(-z)]
decision : p-value<α,Reject null hypothesis
NO, we do not agree that these results show that hydroxychloroquine
is a useful treatment for SARS-CoV-2