Bollywood Sequels as Brand Extensions

Part 3: Multiple Linear Regression

Summary Statistics

          vars   n   mean     sd median
Sequel       2 749   0.10   0.30      0
Actor        3 749   2.85   1.94      3
Actress      4 749   2.33   1.86      2
Producer     5 749   0.48   0.50      0
Director     6 749   0.48   0.50      0
Rating       7 749   2.99   0.83      3
BoxOffice    8 749 395.70 672.18    201

Regression - Model

\[ BoxOffice= \beta_0 + \beta_1*Sequel+ \beta_2*Actor + \beta_3*Actress \] \[ + \beta_4*Producer + \beta_5*Director + \beta_6*Rating \]

Regression - Model


Call:
lm(formula = BoxOffice ~ Sequel + Actor + Actress + Producer + 
    Director + Rating, data = sequel.df)

Residuals:
   Min     1Q Median     3Q    Max 
-902.0 -250.6  -94.6   95.1 6411.3 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept) -517.005     89.255  -5.792 1.02e-08 ***
Sequel       336.853     73.676   4.572 5.66e-06 ***
Actor         54.598     14.540   3.755 0.000187 ***
Actress       88.888     15.024   5.916 5.03e-09 ***
Producer      66.630     50.116   1.330 0.184089    
Director      -5.378     50.354  -0.107 0.914975    
Rating       163.147     26.686   6.114 1.57e-09 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 599 on 742 degrees of freedom
Multiple R-squared:  0.2123,    Adjusted R-squared:  0.2059 
F-statistic: 33.33 on 6 and 742 DF,  p-value: < 2.2e-16

Inferences from Model:

1) The expected increase in BoxOffice collection of a Sequel relative to a non-Sequel is INR 336.8 million.

Inferences from Model: Star Power

2) An increase in Actor Rating by 1 unit, in a movie, increases the expected BoxOffice collection by INR 54.6 million.

3) An increase in Actress Rating by 1 unit, in a movie, increases the expected BoxOffice collection by INR 88.9 million.

Regression - Model


Call:
lm(formula = BoxOffice ~ Sequel + Actor + Actress + Producer + 
    Director + Rating, data = sequel.df)

Residuals:
   Min     1Q Median     3Q    Max 
-902.0 -250.6  -94.6   95.1 6411.3 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept) -517.005     89.255  -5.792 1.02e-08 ***
Sequel       336.853     73.676   4.572 5.66e-06 ***
Actor         54.598     14.540   3.755 0.000187 ***
Actress       88.888     15.024   5.916 5.03e-09 ***
Producer      66.630     50.116   1.330 0.184089    
Director      -5.378     50.354  -0.107 0.914975    
Rating       163.147     26.686   6.114 1.57e-09 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 599 on 742 degrees of freedom
Multiple R-squared:  0.2123,    Adjusted R-squared:  0.2059 
F-statistic: 33.33 on 6 and 742 DF,  p-value: < 2.2e-16

Inferences from Model: Experience

4) The experience level of the Director and the Producer has no significant influence on the expected BoxOffice collection

Regression - Model


Call:
lm(formula = BoxOffice ~ Sequel + Actor + Actress + Producer + 
    Director + Rating, data = sequel.df)

Residuals:
   Min     1Q Median     3Q    Max 
-902.0 -250.6  -94.6   95.1 6411.3 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept) -517.005     89.255  -5.792 1.02e-08 ***
Sequel       336.853     73.676   4.572 5.66e-06 ***
Actor         54.598     14.540   3.755 0.000187 ***
Actress       88.888     15.024   5.916 5.03e-09 ***
Producer      66.630     50.116   1.330 0.184089    
Director      -5.378     50.354  -0.107 0.914975    
Rating       163.147     26.686   6.114 1.57e-09 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 599 on 742 degrees of freedom
Multiple R-squared:  0.2123,    Adjusted R-squared:  0.2059 
F-statistic: 33.33 on 6 and 742 DF,  p-value: < 2.2e-16

Inferences from Model: Movie Rating

5) An increase in Movie Rating given by the Audience by 1 unit increases the expected BoxOffice collection by INR 163.1 million