Question

VERY URGENT

A world wide fast food chain decided to carry out an experiment to assess the influence of advertising expenditure (A) on sales (S) or vice versa. The table shows, for 8 randomly selected countries,  percentage increase in advertisement expenditure (A) and percentage increase in sales (B) collected.  The data presented in the following table are the sums and sum of squares. (use 2 digits after decimal point)

  

∑ A = 22.5	∑ A2 = 64.91	∑ ( A-Abar )2 = SSA = 1.63
∑ B = 135	∑ B2 = 2595	∑ ( B-Bbar )2 = SSB = 316.88
∑ AB= 394.4		∑ [( A-Abar )(B-Bbar )] = SSAB = 14.71

a. Compute sample correlation coefficient. r=Yanıt

b. Test correlation coefficient for significance. (hint: ρ is population correlation coefficient)

Null Hypothesis is

Ho: ρ>0

Ho: ρ≥0

Ho: ρ<0

Ho: ρ≤0

Alternative Hypothesis is

H1: ρ<0

H1: ρ≥0

H1: ρ>0

H1: ρ≤0

Test statistic  is Yanıt

Critical Value of the test at α=5% is

Does your findings support the hypothesis that as advertisement expenditures increases, sales also increases at 5% significance level?

May be

Yes

No

c. Find the linear regression equation of sales based on advertisement.

Estimated sales = + Advertisement

d. Find the linear regression equation of advertisement based on sales.

Estimated advertisement = + Sales

e. Which model is more deterministic? (hint: use coefficient of determination to decide)

Model in part (c)

Model in part (d)

They are equivalent

Answer 1