I have collected data over 30 days and want to assess if there is a difference between the two groups (control and treatment) over each day. Is it appropriate to carry out linear regression on this type of time series data? My outcome variable is continuous.
Asked
Active
Viewed 245 times
1
-
1I would look into panel data (longitudinal) methods. Difference-in-difference is a popular method for this class of problems if you have pre-treatment observations. – lmo Jan 01 '17 at 15:03
-
Here's an early post on CV that provides a quick intro. – lmo Jan 02 '17 at 16:13