I want to investigate adherence to a medication. I want to include both children and adults. For children, caregivers will answer questions on their behalf while adults will directly participate. Some variables like knowledge related will come from two different populations (caregivers and patients). Finally, can I run a regression considering the population as single population?
Thank you!
From what I understand about your situation, you can only run a single regression if you have complete data on each variable. So, you cannot enter a variable where either all the children or all the caregivers have missing data.
Thanks @ David L Morgan for your response. You are right but my concern is that, for example, knowledge as predictor will be collected from patients (for adults) and caregivers (for children). Is that possible or okay to run in a regression (considering populations as one sample population)?
From a technical perspective it is possible and it sounds like it is the only option you have.
I personally appreciate your efforts and consistent commitment. I think it would be very nice to follow what was suggested by Prof David L Morgan as he is a senior expert in the area that you are concerned with.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis