The document describes using multiple regression analysis to build a model for predicting cake sales using 6 factors. It provides data on monthly sales and factors like advertising costs, delivery staff, varieties etc. for 25 observations. Regression is performed using SPSS with enter, forward and backward methods. The enter method finds 69% of sales variation is explained and identifies significant factors. The forward method identifies advertising and delivery staff as most significant in explaining 60% of variation. The backward method finds a similar level of explanation with fewer factors.
The document describes using multiple regression analysis to build a model for predicting cake sales using 6 factors. It provides data on monthly sales and factors like advertising costs, delivery staff, varieties etc. for 25 observations. Regression is performed using SPSS with enter, forward and backward methods. The enter method finds 69% of sales variation is explained and identifies significant factors. The forward method identifies advertising and delivery staff as most significant in explaining 60% of variation. The backward method finds a similar level of explanation with fewer factors.
The document describes using multiple regression analysis to build a model for predicting cake sales using 6 factors. It provides data on monthly sales and factors like advertising costs, delivery staff, varieties etc. for 25 observations. Regression is performed using SPSS with enter, forward and backward methods. The enter method finds 69% of sales variation is explained and identifies significant factors. The forward method identifies advertising and delivery staff as most significant in explaining 60% of variation. The backward method finds a similar level of explanation with fewer factors.
The document describes using multiple regression analysis to build a model for predicting cake sales using 6 factors. It provides data on monthly sales and factors like advertising costs, delivery staff, varieties etc. for 25 observations. Regression is performed using SPSS with enter, forward and backward methods. The enter method finds 69% of sales variation is explained and identifies significant factors. The forward method identifies advertising and delivery staff as most significant in explaining 60% of variation. The backward method finds a similar level of explanation with fewer factors.
Download as DOCX, PDF, TXT or read online from Scribd
Download as docx, pdf, or txt
You are on page 1/ 7
NAME:S.
JAYAGOKUL REGNO:22MBA0168 BUSINESS ANALYTICS AND INTELLIGENCE
DIGITAL ASSIGNMENT 2
1. Write a case on Multiple Regression Analysis, create a data set, get the output tables from statistical software, perform enter and forward and backward regression analysis. Interpret the tables.
PROBLEM
The owner of darling bakery, Vellore would like to build a regression model consisting of six factors, to predict the sales of cakes. Data for the past twenty months on sales and six different factors were collected for the purpose.
METHODOLOGY
The data is self-generated.
There are 25 observations. The three regression models are created using SPSS
INPUT
The variables for which the data have been collected are as follows:
Dependent Variable:
Y = Monthly sales (in Rs. 000s)
Independent Variables:
X1 = Number of delivery boys
X2 = Cost of advertisements in Rs. 000’s
X3 = Number of outlets
X4 = Varieties of pizzas
X5 = Competitors activities index
X6 = Number of existing customers (‘000s)
NAME:S.JAYAGOKUL REGNO:22MBA0168 DATASET
S. SALES ADCOST DELBOYS VARIETIES OUTLETS COMPETITOR CUSTOMER
OUTPUT NAME:S.JAYAGOKUL REGNO:22MBA0168 1. REGRESSION: ENTER
Model Summary Mode R R Adjusted R Std. Error of l Square Square the Estimate 1 .831 a .690 .587 13.969 a. Predictors: (Constant), CUSTOMER, COMPETITOR, DELBOYS, VARIETIES, ADCOST, OUTLETS
ANOVAa Model Sum of df Mean F Sig. Squares Square 1 Regression 7822.304 6 1303.717 6.681 .001b Residual 3512.336 18 195.130 Total 11334.640 24 a. Dependent Variable: SALES b. Predictors: (Constant), CUSTOMER, COMPETITOR, DELBOYS, VARIETIES, ADCOST, OUTLETS
Coefficientsa Model Unstandardized Coefficients Standardized t Sig. Coefficients B Std. Error Beta 1 (Constant) -61.890 31.229 -1.982 .063 ADCOST 2.324 .652 .653 3.564 .002 DELBOYS 1.642 .805 .337 2.039 .056 VARIETIES 3.258 2.236 .289 1.457 .162 OUTLETS -.249 .853 -.086 -.292 .774 COMPETITO 5.187 2.816 .291 1.842 .082 R CUSTOMER -.178 .372 -.118 -.478 .638 a. Dependent Variable: SALES
INTERPRETATION: NAME:S.JAYAGOKUL REGNO:22MBA0168 The value of R-square is .690, which represents that there is 69 percent of variation of sales which can be explained by considering all the independent variables. The linear equation commonly used for a regression analysis is Y = a + bX1 + cX2 + dX3 + ....., Where, Y is the dependent variable and X1, X2, X3,..... are the independent variables, and b, c, d are the coefficients of the respective independent variables. From the coefficients table, the above equation can be written as follows: Sales = -61.890+ 2.3249 (Ad. cost) + 1.642 (del boys) + 3.258 (Varieties) + -0.249 (Outlets) + 5.187 (Competition) + -0.178 (Customer). Anova: The significant value is 0.001 which is below then 0.05. Therefore, the Regression model is a good fit.
2. REGRESSION: FORWARD
Model Summary Mode R R Adjusted R Std. Error of l Square Square the Estimate 1 .703 a .494 .472 15.795 2 .774 b .599 .563 14.369 a. Predictors: (Constant), ADCOST b. Predictors: (Constant), ADCOST, DELBOYS
ANOVAa Model Sum of df Mean F Sig. Squares Square 1 Regression 5596.900 1 5596.900 22.435 .000b Residual 5737.740 23 249.467 Total 11334.640 24 2 Regression 6792.353 2 3396.176 16.449 .000c Residual 4542.287 22 206.468 Total 11334.640 24 a. Dependent Variable: SALES b. Predictors: (Constant), ADCOST c. Predictors: (Constant), ADCOST, DELBOYS NAME:S.JAYAGOKUL REGNO:22MBA0168
Coefficientsa Model Unstandardized Coefficients Standardized t Sig. Coefficients B Std. Error Beta 1 (Constant) .984 7.247 .136 .893 ADCOST 2.500 .528 .703 4.737 .000 2 (Constant) -3.723 6.877 -.541 .594 ADCOST 2.107 .507 .592 4.154 .000 DELBOY 1.672 .695 .343 2.406 .025 S a. Dependent Variable: SALES
INTERPRETATION: The value of R-square is .599, which represents that there is 59.9 percent of variation of sales which can be explained by considering all the independent variables. The linear equation commonly used for a regression analysis is Y = a + bX1 + cX2 + dX3 + ....., Where, Y is the dependent variable and X1, X2, X3,..... are the independent variables, and b, c, d are the coefficients of the respective independent variables. From the coefficients table, the above equation can be written as follows: Sales = -3.723 + 2.107 (Ad. cost) + 1.672 (del boys). Anova: The significant value is 0.001 which is below then 0.05. Therefore, the Regression model is a good fit.
3. REGRESSION: BACKWARD
Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .831 a .690 .587 13.969 2 .830 b .689 .607 13.628 3 .821 c .674 .609 13.585 4 .802d .642 .591 13.893 5 .774e .599 .563 14.369 a. Predictors: (Constant), CUSTOMER, COMPETITOR, DELBOYS, VARIETIES, ADCOST, OUTLETS b. Predictors: (Constant), CUSTOMER, COMPETITOR, DELBOYS, VARIETIES, ADCOST c. Predictors: (Constant), COMPETITOR, DELBOYS, VARIETIES, ADCOST d. Predictors: (Constant), COMPETITOR, DELBOYS, ADCOST e. Predictors: (Constant), DELBOYS, ADCOST NAME:S.JAYAGOKUL REGNO:22MBA0168
The value of R-square is .599, which represents that there is 59.9 percent of variation of sales which can be explained by considering all the independent variables. The linear equation commonly used for a regression analysis is Y = a + bX1 + cX2 + dX3 + ....., Where, Y is the dependent variable and X1, X2, X3,..... are the independent variables, and b, c, d are the coefficients of the respective independent variables. From the coefficients table, the above equation can be written as follows: Sales = -3.723 + 2.107 (Ad. cost) + 1.672 (del boys). Anova: The significant value is 0.001 which is below then 0.05. Therefore, the Regression model is a good fit.
