Suppose you have time series data at the quarterly frequency, and wish to regress yt on xt allowing for constant or intercept. You also wish to allow for the possibility that the intercept depends on the quarter of the year. How might you do this? i) Include a constant term and 4 dummy variables - one dummy for each quarter of the year. ii) Exclude the constant term, and just include 4 dummy variables. iii) Include the constant term and dummy variables for the first 3 seasons only. iv) Include the constant term and dummy variables for quarters 2,3 and 4, only. Any of i), ii), iii) or iv) would be fine. Only ii), iii) or iv) would work. iii) only iv) only