Question

An analyst must decide between two different forecasting techniques for weekly sales of roller blades: a linear trend equation and the naive approach. The linear trend equation is Ft = 124 + 2.1t, and it was developed using data from periods 1 through 10. Based on data for periods 11 through 20 as shown in the table, which of these two methods has the greater accuracy if MAD and MSE are used? 

t	Units Sold
11	145
12	145
13	152
14	144
15	152
16	149
17	156
18	155
19	162
20	165

Answer 1

Units sold Forecast Error Abs(Error) Error 2 2.1 4.41 2.1 145147.1 17.64 4.2 149.2 -4.2 145 12 0.49 0.7 0.7 151.3 152 4 9.4 88.36 9.4 153.4 144 14 3.5 12.25 3.5 155.5 152 15 8.6 73.96 8.6 149 157.6 16 13.69 3.7 3.7 159.7 17 46.24 6.8 6.8 161.8 155 18 3.61 162 163.9 19 166 165 20 41.9261.65 12 Total 4.19 13 MAD 26.165 14 MSE 15

Units sold Forecast Erro A(rror Error 2 145 145 145 49 145 152 152 144 144 152 152 149 16 49 149 17 1 1 49 162 162 165 20 Total 4.888889 MAD 32.66667 MSE

MAD (Naive) = 4.888888889
MAD(Linear) = 4.19

MSE (Naive) = 32.66666667

MSE(Linear) = 26.165

Linear trend provides forecasts with less average error and less average squared error