I got a little question.
I need to build a regression model to check the relation (no need for prediction) between Y, and a number of X's (X1,...,X5).
The problem is, that Y is a time series !
I checked Y, and it has a seasonal trend. It is also correlated to it's Lag(1), meaning, every case is correlated to the previous one, but again, that can be due to the seasonality, because some cases follow a case which is in the same season.
What should I do to be able to run a "normal" regression model ? Is it enough to create a categorical variable to represent the season, and to include also the variable of the 1st lag ? If I do that, will my analysis be more or less accurate ?