This questions is about what is the right way to include time variable in model.

I will start here.

I have two outcomes, 1) Continuous 2) Discrete(binary). I have several other adjustment variables and one of which is “time”.

Specifically, my time variable is as follows: Jan 1990, Feb 1990, March 1990......Jan1992. Two years time span with monthly intervals. Right now this is coded as a factor variable.

I am afraid that if I include this without thinking deeper about this variable, the effect estimates for each month may be gobble-de-goup i mean, i am worried it may show up as March 1990, June 1991, Jan 1990 etc in any random order that is not correct.

So my question is , Is it okay to include time (month-year) variable as-is which is currently a factor variable or should I apply additionally transformation (change to numeric or integer) etc such that the effects are interpretable as effects over time ? Please advise. Thanks.