If we use the multiplicative decomposition method to analyze the quarterly bicycle sales data given in Table 15.3 below, we find that the quarterly seasonal factors are .49, 1.19, 1.64, and .63. Furthermore, if we use a statistical software package to fit a straight line to the deseasonalized sales values, we find that the estimate of the trend is

tr_t = 22.61 + .59t

In addition, we find that the half-lengths of 95 percent prediction intervals for the deseasonalized sales values in the four quarters of the next year are, respectively, 2.74, 2.79, 2.84, and 2.96.

(a) Calculate point predictions of bicycle sales in the four quarters of the next year.

Point Predictions = sn tr
t	sn	tr = 22.61 + .59t	sn tr
17	.49
18	1.19
19	1.64
20	.63

(b) Calculate approximate 95 percent prediction intervals for bicycle sales in the four quarters of the next year. (Round your answers to 2 decimal places.)

95% Prediction Intervals:

yy^

₁₇ : [ , ]

yy^

₁₈ : [ , ]

yy^

₁₉ : [ , ]

yy^

₂₀ : [ , ]