This program allows either the determination of sample size, or the calculation 

of the power of a comparison of two percentages, two means, or two survival curves.

For the TI-82:

----begin documentation----

This program allows either the determination of sample size, or the calculation 

of the power of a comparison of two percentages, two means, or two survival curves.

It has been written after Lachin JM "Introduction to sample size determination 

and power anlysis for clinical trials." Controlled Clinical Trials. (1981) ; 2 : 

93-113.



As it, it allows to calculate sample size for alpha=.05 (one or two tailed) and 

for beta=.80 or .90. Of course, other values can be added but this will inflate the 

program. You also must indicate the control to case ratio (generally 1).

The value computed for the sample size (N) is the total of cases + control.

The values computed for survival curves are for two independent groups with 

censoring and continuous recrutment for the whole duration of study. I will perhaps 

write later for other situations.



Feel free to modify this program and propose improved versions.