Table 1:

YEAR SALES(in lakhs) PRODUCTION(in lakhs) ADEXPENDITURE(GRPs)

2008 7 11 900

2009 8.5 12 900

2010 9.2 12.4 1000

Following are the Z-scores for the above:

Table 2:

ZSALES ZPRODUCTION ZADEXPENDITURE

-1.09729 -1.30854 -0.86603

0.23725 -0.11896 -0.86603

0.86004 0.35687 0.86603

I run Linear Regression on the data of Table1, with Sales as the dependent variable and Production and AdExpenditure as the independent variables.

Following is the output in SPSS:

**Model Summary**

Model R R Square Adjusted R Square Std. Error of the Estimate

1 1.000a 1.000 . .

a. Predictors: (Constant), ADSPEND, PRODUCTION

**ANOVA**

Model Sum of Squares df Mean Square F Sig.

1 Regression 2.527 2 1.263 . .a

Residual .000 0 .

Total 2.527 2

a. Predictors: (Constant), ADSPEND, PRODUCTION

b. Dependent Variable: SALES

**Coefficient**

Model Unstandardized Coefficients Standardized Coefficients

B Std. Error Beta

1 (Constant) -10.400 .000

PRODUCTION 1.500 .000 .962

ADSPEND .001 .000 .051

As per the coefficients, the regression equation is:

SALES = -10.4 + 1.5*PRODUCTION + .001*ADSPEND

**Verification:**

To verify the above linear equation, let us try to predict the sales of 2010,and 2011.

Sales (2010) = -10.4 +1.5*12.4 + 1

Sales (2010) = 9.8 (seems reasonable, atleast for demonstration)

**Verification (Z Score Coefficients):**(Constant is always zero in this case)

Sales (2010) = 0.962*0.35687 + 0.86603*0.051

Sales (2010) = 0.3433 + 0.04 = 0.3833 ( no where close to the actual value 0.86 from Table 2)

Please let me know why am I not able to get the Z Score coefficient for Sales 2010 correctly. This forum has been excellent in resolving my doubts.

I am sorry for the untidy math here:shakehead