You are on page 1of 7

The economist “meat loaf is

counter-cyclical”
A lot of data, a lot of opinions

Data Analytics for Leaders


Quality Kitchens
She said “Promotion affects future
People eat meat loaf in winter sales”
Team 3 members:
Azza Hmissi
Belgacem Gnichi
Dorsaf Ben Cheikh
Habiba Dridi
Mohamed Ben Braham He said “advertising results are
Walid Ben Ayed confusing”

1
Adopted Steps

• How correlated variables are


• Check linearity of Sales versus Promotion, Advertising and Index
• Consider Dummy variables (mutually exclusive and collectively exhaustive)
• Iterative Regression Models (regression 1,2,3)

2
Models Simulations

Regression Statistics Regression Statistics Regression Statistics


Multiple R 0,9509 Multiple R 0,9495 Multiple R 0,9493
R Square 0,9042 R Square 0,9016 R Square 0,9012
Adjusted R Square 0,8495 Adjusted R Square 0,8556 Adjusted R Square 0,8641
Standard Error 52,1203 Standard Error 51,0407 Standard Error 49,5169
Observations 23 Observations 23,0000 Observations 23,0000

Standard Lower Upper Coefficie Standar Lower Upper Coefficien Standard Lower Upper
  Coefficients Error t Stat P-value 95% 95%   nts d Error t Stat P-value 95% 95%   ts Error t Stat P-value 95% 95%
Interce Intercept 826,57 235,59 3,51 0,00 324,42 1328,72 Intercept 822,32 227,93 3,61 0,00 339,14 1305,50
pt 808,28 242,37 3,33 0,00 288,43 1328,12 prom 6,57 0,78 8,39 0,00 4,90 8,24 prom 6,54 0,75 8,71 0,00 4,95 8,14
prom 6,51 0,81 8,07 0,00 4,78 8,24
prom-1 -3,53 0,75 -4,72 0,00 -5,12 -1,93 prom-1 -3,49 0,71 -4,90 0,00 -5,00 -1,98
prom-1 -3,41 0,79 -4,34 0,00 -5,09 -1,73
adv 2,60 0,71 3,66 0,00 1,08 4,11 adv 2,52 0,62 4,06 0,00 1,21 3,84
adv 2,64 0,73 3,63 0,00 1,08 4,20
adv-1 2,78 0,75 3,70 0,00 1,17 4,38 adv-1 2,80 0,73 3,81 0,00 1,23 4,36 adv-1 2,87 0,65 4,44 0,00 1,50 4,24
index -4,92 2,33 -2,11 0,05 -9,92 0,07 index -4,99 2,28 -2,19 0,04 -9,85 -0,14 index -4,95 2,20 -2,25 0,04 -9,63 -0,28
q1 20,36 32,81 0,62 0,54 -50,02 90,74 q2 -71,01 29,19 -2,43 0,03 -133,23 -8,79 s1 -66,62 22,25 -2,99 0,01 -113,80 -19,45
q2 -61,24 33,71 -1,82 0,09 -133,54 11,05 q3 -62,51 28,51 -2,19 0,04 -123,28 -1,75
q3 -52,65 33,17 -1,59 0,13 -123,79 18,49

Sales = 808,28 + 6,51prom - 3,41prom(-1) + 2,64adv + Sales = 826,57 + 6,57prom - 3,53prom(-1) + 2,6adv + Sales = 822,32 + 6,54prom - 3,49prom(-1) + 2,52adv +
2,78adv(-1) - 4,92index + 20,36q1 - 61,24q2 - 52,65q3 2,8adv(-1) - 4,99index – 71,01q2 - 62,51q3 2,87adv(-1) - 4,95index – 66,62s1

3
Spending on Advertising or Promotion

Sales = 822,32 + 6,54prom - 3,49prom(-1) + 2,52adv + 2,87adv(-1) - 4,95index – 66,62s1

Impact of advertising on sales = 2.52 + 2.87 = 5.39


Impact of promotion on sales = 6.54 - 3.49 = 3.05
 1k$ to spend on advertising

4
Is meat loaf mix a Counter-cyclical item?

  Coefficients Standard Error t Stat P-value Lower 95% Upper 95%


Intercept 822,32 227,93 3,61 0,00 339,14 1305,50
Prom 6,54 0,75 8,71 0,00 4,95 8,14
prom-1 -3,49 0,71 -4,90 0,00 -5,00 -1,98
adv 2,52 0,62 4,06 0,00 1,21 3,84
adv-1 2,87 0,65 4,44 0,00 1,50 4,24
index -4,95 2,20 -2,25 0,04 -9,63 -0,28
s1 -66,62 22,25 -2,99 0,01 -113,80 -19,45

 index variable :
coefficient < 0, t stat>2 , p-value<0,05, CI doesn’t include 0

 The meat loaf mix is a counter-cyclical item

5
Seasonal effects on Sales

  Coefficients Standard Error t Stat P-value Lower 95% Upper 95%


Intercept 822,32 227,93 3,61 0,00 339,14 1305,50
Prom 6,54 0,75 8,71 0,00 4,95 8,14
prom-1 -3,49 0,71 -4,90 0,00 -5,00 -1,98
adv 2,52 0,62 4,06 0,00 1,21 3,84
adv-1 2,87 0,65 4,44 0,00 1,50 4,24
index -4,95 2,20 -2,25 0,04 -9,63 -0,28
s1 -66,62 22,25 -2,99 0,01 -113,80 -19,45

 q2 and q3 grouped on a single variable S1


 Quality of the Model improved, Adjusted R2 = 0,8641 (from 0,8556)
 t stat increased to -2,99
 Significant seasonal effect

6
Thanks

You might also like