Question

you are testing a hypothesis that the mean value of blood pressure in one cohort is lower than that in the second cohort. what steps should you take to solve this problem? assume that sampling was done in manner to produce a simple random sample. also assume that cohorts include different individuals. you will need to calculate the value of standard deviations of blood pressure measurements in each cohort based on experiment data available. both cohorts contain 50 individuals .

Answer 1

Answer)

First step is to set up hypothesis

Null hypothesis Ho : u1 = u2

Alternate hypothesis Ha : u1 < u2

Where u1 = mean blood pressure of first cohort

u2 = mean blood pressure of second cohort

Second step is to choose the type of test

As here we need to calculate the sample s.d so it is obvious that, population s.d is unknown

So we will use two sample t test

Third step is to estimate the test statistics

Test statistics = (x1-x2)/standard error

X1 = sample mean of cohort 1

X2 = sample mean of cohort 2

Standard error = √{(s1^2/n1)+(s2^2/n2)}

Here n1 = n2 = 50

S1 = sample s.d of cohort 1

S2 = sample s.d of cohort 2

Step 4 is to find Pvalue

First we will estimate the degrees of freedom which is equal to smaller of n1-1, n2-1

Then by t distribution we will estimate the p-value

Fifth step is to make the conclusion

If p-value is less than the given significance level reject Ho