0
$\begingroup$

How can I calculate the power or the adequate size of my population in a Cox multivariable regression analysis where I'm testing the association of different independent variables with my dependent variable (0/1)? I have 843 patients 79 events and in my final multivariable model I have 4 variables (3 are 0/1; 1 is 0,1,2,3). I'm familiar only with SPSS and MedCalc.

$\endgroup$
1
  • $\begingroup$ It sounds like you have already collected your data. Why are you conducting a power analysis now? $\endgroup$
    – TPM
    Dec 4, 2019 at 21:03

1 Answer 1

0
$\begingroup$

SPSS Statistics currently doesn't offer anything designed to do any type of power or sample size analysis for a Cox regression model. Since the program has some general mathematical functions, distribution functions, and a matrix programming language, if you know the math you can program pretty much anything, but I assume that's not what you're looking to do.

I'm not very familiar with MedCalc, but based on a quick web search it looks like it offers power analysis (specifically calculation of required sample sizes) for logrank tests for survival analysis, but nothing for general Cox regression models. I might have missed something, but it doesn't look like it'll do what you're looking for either.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.