Dear network!
I have a question and hope that you will share with me your experience in this field.
We have conducted a study (review of patients medical records) to assess if abnormal hemoglobin (Hb) levels are associated with stroke severity in our settings. We are interested in both; high and low Hb levels. As for me, I though the right thing is to dichotomize Hb levels into normal and abnormal status in the Regression Analysis. Upon submitting the work to a journal, the reviewers mentioned that Hb levels should have been continuous variable. As both extreme can be a determinant of stroke severity, how the results would be interpreted in this way!
Thank you for your time, Majeda
Majeda -
Not a subject matter field for me, but can you use polynomial regression for that? Regardless, don't forget heteroscedasticity. I have an example of that, borrowing a BMI regression example in here:
https://www.researchgate.net/publication/320853387_Essential_Heteroscedasticity
Is this multiple regression where Hb is just one regressor? That should work too.
Cheers - Jim
James R Knaub and Zeynel Cebeci
: many thanks for your kind responses.Mr. Cebeci: No, the opposite! Stroke severity is the dependent variable (continuous scale scores), Hb level is an independent variable.
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation