Question

A hotel chain decides to use data on existing hotels to determine desirable locations for new hotels. Data on 90 hotels in the chain have been collected in an attempt to understand operating margin.

The variables are:

• Operating margin – the dependent variable, in %

• Indoor pool – whether the hotel has a heated indoor pool, Yes = 1 and No = 0

• Competitor rooms – total number of competitor rooms in hotels within 3 miles

• Distance to competitor – miles to nearest competitor hotel

• Office space – amount of office space available within 3 miles, in thousands of square feet

• Distance to downtown – miles to downtown (the central part of a city or town).

SPSS output is provided in Figure 2.

(a) Analyze the regression results, making sure you first write out the full regression model, including any assumptions, and the estimated model. In light of the regression results, propose any changes you would make to the model, including any other explanatory variables which you would consider using, or any you would remove from the existing model. Keep in mind the hotel’s objective of understanding operating margin to inform its decisions about new hotel sites.
Figure 2 Model Summary Adjusted R R Square Square .782 .769 Model R Std. Error of the Estimate 2.9158% 1 .884 ANOVA Model df F Sum of Squares 2557.772 714.164 Mean Square 511.554 Sig .000 Regression Residual 60.169 5 84 8.502 Total 3271.936 89 Coefficients Standardized Coefficients Beta Model Unstandardized Coefficients B Std. Error 58.081 2.659 2.316 671 21.840 3.453 Sig. .000 .001 .184 -.007 .001 - 536 -10.263 .000 (Constant) Indoor pool Competitor rooms Distance to competitor Office space Distance to downtown -1.864 360 -272 -5.172 .000 .019 002 .512 9.717 .000 .195 099 .103 1.971 052

Answer 1

Assumptions Our dependent Should be continuous. variable ,iner, Operating margin There should be a linear relationship between dependent variable and each variable and each of the Independent variable. ( in data . There should be homo scedast There should be no multicollin In data, .IV edasticity linearity nomaly Residuals Should be noo distributed.

Model is Regression given by : let y operating margin 21 : Indoor pool 22 • competitor rooms 713 : Distance to competitor ru Office space Distance to downtown. HS: Now the model is given by y = Bo + Bini Bo + Birr & Bella + 8383 + Bunlu + ßsts where, pils saree unstandardized coefficient ß: Now, IS Estimated model given by 58.081 + y 2.316 H1 + (-0.007) H2 + (-1-864) U3 0019 ry 0-195 x5 + Now in the See that sig. Coefficient Model output we can sig value for 15 is 0.052 5% level which means variable 15 is differenti ller than which is greater Significance Statistically signifcantly of so we can doop variable us not from owe 2 Model.