Refer to Exercise 13.21. An additional regression model for the snack cracker data is run, incorporating products of the promotion variable with price and with category sales. The output for this model is given in the figure. What effect do the product term coefficients have in predicting sales when there is a promotion by a competing brand? In particular, do these coefficients affect the intercept of the model or the slopes?

In exercise

Transcribed Image Text:

Responae: Salee auwary ct eit Square Square dj Roct Haan Square Eror 17.791 Mean of Responae cbservat icns for sun ts 04 0.452443 0.424506 Model bror C Total 103 DF sn of quares Mean Square ati0 5175.35 16.1953 19.62 Prob P 0.0000 25901.736 31322 726 57204 462 356.1692 20 20 10 Parameta Eatinate Eetinate Sta Error t Ratio Prcb It) 26.806609 90.32649 0.27 0.7457 90.233065 47.75194 .89 0.0618 0,1335274 0.023654 .E0 0.0000 Term Intercepe Prica Category sales Promction b thr 287.6092 172.2049 1.7 01 Price*protot10-142.4326 94.15011 ブe5 0.1015 Category*Ficnotto 0.024087 0.0368160.5 0.5145 tfect Test 40 50 290 310 330 350 370 390 410 Sales Fredicted Respcnse Sales armacy of Pit Requare Square Rot Hean square E 17.9913 Mean ot BRaeponse Cbservatians for Sun e rice 0.434158 0.417182 Source Model le'sm of squares Mean square 묜Ratio 25.5759 24835.761 2368.01 53204.462 a2าจ.sa 100 323.6 Prob> F 356.7692 104 C Total 103 0.0000 40 30 20 10 Parameter Estisates Tezn Intercept Price Category sales 0.1214871 0.01824 .00 Frmoticn by other? -19.9594 3.7004.39 10 123.95375 0.66628 1.61 01106 44.849952 39.93534 11 0.261 -20 -30 -40 -30 Ettece Test 290 310 330 350 370 390 410