is there a need to introduce weights or something similar to account for a dataset where # of controls can vary by case?

[Sakshi Rajatbhai Tewari]

I have a dataset of cases and each case having upto 3 control matches. Some cases have 1, others have 2 or 3. Before running a conditional logistic regression (which is what i plan to do), should i think about any form of weighting ? to account for this variation in number of controls?

[Rich Goldstein]

a lot depends on what you are working on but a simple answer is yes: if there is on1 control it gets a weight of 1, if 2, each gets a weight of 0.5, if three, each gets a weight of 1/3; however, there are a number of assumptions underlying this and you have not supplied enough information for anyone to assess those

[Sakshi Rajatbhai Tewari]

a lot depends on what you are working on but a simple answer is yes: if there is on1 control it gets a weight of 1, if 2, each gets a weight of 0.5, if three, each gets a weight of 1/3; however, there are a number of assumptions underlying this and you have not supplied enough information for anyone to assess those

Thank you for your response! apologies, this is my first time doing a CC analysis. If you could list a few assumptions i would be happy to supply the information, thank you!