In: Statistics and Probability
he data on sheet #2 represents data various economic factors. We want to use a subset of the remaining variables to predict Interest Rates on US Treasury Bonds (Interest). Construct a table of correlations and chose the two variables that have the strongest correlation to Interest and build a multiple regression model from those variables. Write your regression equation, value, and the meaning of each variable used. State the overall hypothesis test for your model
| CRUDE | FOREIGN | DJIA | GNP | PURCHASE | CONSUMER | INTEREST | 
| 10.9 | 31 | 974.9000244 | 1718 | 1.756999969 | 234.3999939 | 7.61 | 
| 12 | 35 | 894.5999756 | 1918 | 1.649000049 | 263.7999878 | 7.42 | 
| 12.5 | 42 | 820.2000122 | 2164 | 1.531999946 | 308.2999878 | 8.41 | 
| 17.7 | 54 | 844.4000244 | 2418 | 1.379999995 | 347.5 | 9.439999 | 
| 28.1 | 83 | 891.4000244 | 2732 | 1.215000033 | 349.3999939 | 11.46 | 
| 35.6 | 109 | 932.9000244 | 3053 | 1.09800005 | 366.6000061 | 13.91 | 
| 31.8 | 125 | 884.4000244 | 3166 | 1.034999967 | 381.1000061 | 13 | 
| 29 | 137 | 1190.300049 | 3406 | 1 | 430.3999939 | 11.11 | 
| 28.6 | 165 | 1178.5 | 3772 | 0.961000025 | 511.7999878 | 12.44 | 
| 26.8 | 185 | 1328.199951 | 4015 | 0.927999973 | 592.4000244 | 10.62 | 
| 14.6 | 209 | 1792.800049 | 4240 | 0.912999988 | 646.0999756 | 7.68 | 
| 17.9 | 244 | 2276 | 4527 | 0.879999995 | 685.5 | 8.38 | 
For the given data , the correlation table is shown below:
| CRUDE | FOREIGN | DJIA | GNP | PURCHASE | CONSUMER | INTEREST | |
| CRUDE | 1 | ||||||
| FOREIGN | 0.304276 | 1 | |||||
| DJIA | -0.14611 | 0.861947 | 1 | ||||
| GNP | 0.374804 | 0.991314 | 0.806587 | 1 | |||
| PURCHASE | -0.61967 | -0.91289 | -0.60792 | -0.95052 | 1 | ||
| CONSUMER | 0.152951 | 0.973337 | 0.883626 | 0.969579 | -0.85455 | 1 | |
| INTEREST | 0.969381 | 0.152756 | -0.30097 | 0.227549 | -0.48812 | 0.003861 | 1 | 
The two variables that have the strongest correlation to Interest are CRUDE & PURCHASE.
Therefore the multiple regression analysis summary using this variable is given below:
| SUMMARY OUTPUT | ||||||||
| Regression Statistics | ||||||||
| Multiple R | 0.979934 | |||||||
| R Square | 0.960271 | |||||||
| Adjusted R Square | 0.951442 | |||||||
| Standard Error | 0.501724 | |||||||
| Observations | 12 | |||||||
| ANOVA | ||||||||
| df | SS | MS | F | Significance F | ||||
| Regression | 2 | 54.75873 | 27.37936 | 108.7661 | 4.97E-07 | |||
| Residual | 9 | 2.265542 | 0.251727 | |||||
| Total | 11 | 57.02427 | ||||||
| Coefficients | Standard Error | t Stat | P-value | Lower 95% | Upper 95% | Lower 95.0% | Upper 95.0% | |
| Intercept | 2.259987 | 1.126809 | 2.005653 | 0.075858 | -0.28903 | 4.809005 | -0.28903 | 4.809005 | 
| CRUDE | 0.282557 | 0.022094 | 12.78899 | 4.47E-07 | 0.232577 | 0.332537 | 0.232577 | 0.332537 | 
| PURCHASE | 1.348011 | 0.624444 | 2.158737 | 0.059189 | -0.06458 | 2.760602 | -0.06458 | 2.760602 | 
The regression equation is,

 The
increase a CRUDE by a unit will increase the Interest by 0.282557
units ,when other variable remain constant.
 The
increase a PURCHASE by a unit will increase the Interest by
1.348011 units ,when other variable remain constant.
For overall hypothesis test,
The p-value of F is quite small which is 0.000000497 . so we have sufficient evidence to reject the null hypothesis .
Hence the at lest of the the slope parameter is not equal to zero.