I have a set of data for 2 visits in patients and I would like to see whether there is a effect of a difference of one variable on another. So, lets say, my variables are A, B, age + gender. I want to see the effect of (amount of change in B) on (amount of change in A).

Is it correct if I model this as

A(BL-FU) = B(BL-FU) + A(BL) + B(BL) + age + gender ?

Also age and gender effects both by A and B.

Please advice on the model. Thank you.

  • 1
    $\begingroup$ What are BL/FU? $\endgroup$ – Robert Kubrick Sep 29 '14 at 15:38
  • $\begingroup$ Baseline and Follow-up values of each patients. Sorry about that. $\endgroup$ – ssm Sep 29 '14 at 15:43
  • $\begingroup$ What do you mean by "I want to see the effect of (amount of change in B) on (amount of change in A)"? Can you rephrase what the predicted variable is? And why do you have a doubt about your model? $\endgroup$ – Robert Kubrick Sep 29 '14 at 15:47
  • $\begingroup$ Thank you for your reply. to rephrase, B variable has two measurements at two time points. A has two measurements at the same time points. Other words, for all patients, baseline A and B and followup A and B values are measured. I would like to see the effect of difference of B (baseline B - followup B) on difference of A (baseline A - followup A). My doubt is in the fact that Im trying to model the 'difference of A' as a function of 'difference of B' with baseline A and baseline B as covariates. Is it statistically correct? Also my other covariates, gender and age effect both A and B. $\endgroup$ – ssm Sep 29 '14 at 15:55
  • 1
    $\begingroup$ You should edit your question and add that you intend to use OLS for this problem, if I'm assuming correctly. $\endgroup$ – Robert Kubrick Sep 29 '14 at 17:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.