I’m a bit stuck with what seems like a simple question. I have data from an experiment whereby participants had to provide ‘yes’ or ‘no’ answers (so the response mode was binary). Their scores, however, are not binary as there were multiple conditions and I took the average score for each condition. The final scores in the dataset are not binary; they take multiple values (e.g., 0, 0.10, 0.15, 0.35 etc. and maximum possible score was 1). Is it still legit to use multiple linear regression and not logistic regression? Any pointers would be extremely helpful!
Thanks!
