(Total stats novice)

I have 8 predictor variables with only 2 levels each (denoting the presence/ non-presence for each data point) and a continuous outcome variable. Logistic regression obviously does not work because of the nature of my predictor variables.

What can I use to model this?

Vignette: I am researching the predictors of severe blood loss during liver surgery for cancer. The liver is anatomically divided into 8 segments. During surgery, different segments are removed depending on the location of cancer.

Modelling "the number of segments removed" is not a problem.

But I want to model the predictive effect of each of the different segments being removed. However, this means that for each surgery a segment is either "removed" or "not removed" - which means these variables only have 2 levels.

What can I use??

  • $\begingroup$ Welcome to CV. Since you’re new here, you may want to take our tour, which has information for new users. If your outcome is a continuous variable, why don't you use multiple linear regression? $\endgroup$ – T.E.G. Apr 9 '18 at 21:46
  • 1
    $\begingroup$ You state that logistic regression will not work because of the nature of your predictor variables, which is an incorrect statement. Logistic regression requires your outcome variable to be nominal with two categories, but can accommodate any type of predictor variable: nominal, ordinal, continuous or discrete. What outcome variable do you actually use to quantify severe blood loss? The nature of that outcome variable (as well as information on the study design) will determine what regression model you can use. $\endgroup$ – Isabella Ghement Apr 9 '18 at 23:42

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.