0
$\begingroup$

any suggestions how to count confidence intervals from zero biased data? I've counted generalized linear model using glmer.nb function. I have to make graph but I was told that confidence intervals cannot be counted as usualy due to random effect, which is not included in calculation. Thank you in advance!

Improve this question
$\endgroup$
2
  • $\begingroup$ What do zero biased data look like? $\endgroup$
    – user158565
    Commented Jul 27, 2019 at 21:42
  • $\begingroup$ I shoud use term - zero inflated data. I beg your pardon. I have continuous response variable (length of trajectory in metres/hour). There are about 1500 hours of experiment, bud only cca 200 are with data, others are full of zero. It means that more than 1200 rows are with ZERO length of trajectory. There are 2 predictors: sex and weight of animal. I've already counted influence of weight and sex on length of the trajectory using glmer function but confidence intervals for length~sex are probably counted incorrectly. $\endgroup$
    – Jakub Horák
    Commented Jul 29, 2019 at 11:55

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.