0
$\begingroup$

I am fairly new to statistics and hoping someone can provide some advice on some analysis I'd like to do. I am looking to analyse the effects of covariates on the repeatability/variability of a continous variable. The continous variable is a blood test and is repeated on a yearly basis for 5 subsequent years for each subject. to try and make it clearer an example of a covariate might be gender of patient, presence of comorbidity, taking a certain medication- does taking this medication influence the variability of the test result for subsequent years etc.

Thanks in advance

Improve this question
$\endgroup$
3
  • $\begingroup$ So you have a blood test which you are administering every year for 5 years and you want to determine if a medication has an effect on the blood test? Repeatability is confusing here, this is not standard terminology. Anyway, this looks like a regular medical study, you will also need a control group for this. You should do more research on the topic. $\endgroup$
    – user2974951
    Commented Feb 14, 2019 at 13:30
  • $\begingroup$ Sorry, I do not think I've explained it properly. We expect the result of the test to be very similar each year for each patient. We want to assess if there are any factors which could be influencing the variation in results for each patient. So for example does having a certain medical condition make it more likely that the test results vary from year to year. $\endgroup$
    – S.Strike
    Commented Feb 15, 2019 at 12:44
  • $\begingroup$ I don't really understand, but I think you are asking for a mixed model, which would return the various factors variances, so which factors are responsible for the most variability in the results. Does this look related stats.stackexchange.com/questions/15768/…? $\endgroup$
    – user2974951
    Commented Feb 15, 2019 at 13:46

0

Reset to default

Your Answer

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

Browse other questions tagged or ask your own question.