Table in the page margin presents quarterly sales of the TRK-50 mountain bike for the previous 16 quarters at a bicycle shop in Switzerland. The time series plot under the sales data shows that the bike sales exhibit an upward linear trend (that is, an upward straight line movement over time) and a strong seasonal pattern, with bike sales being higher in the spring and summer quarters than in the winter and fall quarters. If we let yt denote the number of TRK-50 mountain bikes sold in time period t at the Swiss bike shop, then a regression model describing y t is
yt = β0 + β1t + βQ2Q2 + βQ3Q3 + βQ4Q4 + εt
Here the expression (β0 + β1t) models the linear trend evident in the time series plot, and Q2, Q3, and Q4 are dummy variables defined for quarters 2, 3, and 4. Specifically, Q2 equals 1 if quarterly bike sales were observed in quarter 2 (spring) and 0 otherwise; Q3 equals 1 if quarterly bike sales