brettscaife.net

Solution to question 11.3

I will use the 'magic number' method of calculating the sample sizes.

Standard deviation Difference worth finding Power Significance Magic number Patients per group
a) 8·0 3·0 90% 5% 10·5 150
b) 6·0 2·5 90% 5% 10·5 121
c) 6·0 3·0 80% 5% 7·8 63
d) 6·0 3·0 90% 1% 14·9 120

Note that I rounded all the numbers up to the next whole number. (We would also probably round up the number of patients to the next 10, if necessary, we are, after all, dealing with approximations.) The numbers above are patients per group. You need to double the number to get the total number of patients for the full study.

Back to questions