Which model/test should I choose? (GLM..?)


New Member

I'm new to statistics but I am getting really stuck with my data so I hope someone could help me to choose the best statistical test. I have some test (dummy) data. I have data from 64 different hospitals, for 5 different countries countries, over 5 different years. In all of the hospitals they use a certain drug that is stored at a certain temperature. They have tested the drug for effectiveness after a certain time. For other reasons they also tested for each medicine bottle, at which temperature the medicine would be stored best (with a min. temperature, maximum temperature and the calculated range). NOTE: not all samples are tested for storing conditions... Also for the number of samples per country/year/hospital differ.

This is what the data looks like (also in the attached picture):

Sample: 1....2700
Year: 2011...2015
Hospital: 64 different hospitals: A, B, C...
Country: 5 different countries: NL, AU, PH, CL, NZ
Effectiveness.score: 0...100%
Minimum Temperature: Celcius Degrees
Maximum Temperature Celcius Degrees
Temperature Range: Celcius Degrees

I want to test if there is a correlation between the effectiveness ~ Temperature data. I also want to take into account that Year, Country and Hospital could cause a side (random) effect in my data.

I was thinking to go for GLM but I can't figure out if this is the best model. I have seen that there are other generalized models but I can't find the right literature to pick the right model...

Thank you so much..!