Table 4 Linear regression models explaining logaritmized CPD with the branch information and combined models with multiple influencing factor

(Intercept)

2.65 (0.18)^∗∗∗

1.77 (0.46)^∗∗∗

1.82 (0.59)^∗∗

branche C

2.85 (0.35)^∗∗∗

3.05 (0.49)^∗∗∗

branche D

1.25 (0.31)^∗∗∗

2.42 (0.54)^∗∗∗

branche F

1.24 (0.19)^∗∗∗

1.13 (0.32)^∗∗∗

branche G

1.56 (0.21)^∗∗∗

1.27 (0.34)^∗∗∗

1.26 (0.33)^∗∗∗

branche I

2.17 (0.21)^∗∗∗

1.94 (0.35)^∗∗∗

1.83 (0.34)^∗∗∗

branche J

1.08 (0.24)^∗∗∗

1.19 (0.37)^∗∗

branche K

1.04 (0.23)^∗∗∗

1.26 (0.38)^∗∗∗

branche L

1.15 (0.20)^∗∗∗

1.46 (0.36)^∗∗∗

branche M

0.65 (0.26)^∗

0.85 (0.44)

branche O

0.88 (0.21)^∗∗∗

0.39 (0.39)

branche Q

0.90 (0.24)^∗∗∗

0.99 (0.36)^∗∗∗

0.98 (0.34)^∗∗

opening hours per week

0.00 (0.00)

0.01 (0.00)^∗

combined number of ratings

0.01 (0.01)^∗

0.01 (0.00)^∗

log(area + 1)

0.13 (0.05)^∗

0.09 (0.08)

R²

0.09

0.15

0.19

Adj. R²

0.08

0.13

0.18

Num. obs.

1810

700

298

RMSE

1.61

1.57

1.51