We stated in the beginning of the chapter that regression can be used to understand the way the world works. That is, you can look at the regression coefficients (their signs and magnitudes) to see the effects of the explanatory variables on the dependent variable. However, is it possible that apparently small changes in the data can lead to very different-looking equations? The file P10_57.xlsx lets you explore this question. Columns K–R contains data on over 100 (fictional) homes that were recently sold. The regression equation for this original data set is given in the range T15:U21. Columns C–I contain slight changes to the original data, with the amount of change determined by the adjustable parameters in row 2. The regression equation for the changed data appears in the range T6:U12. It has been calculated through special matrix functions (not StatTools), so that it changes automatically when the random data change. Experiment by pressing the F9 key or changing the adjustable parameters to see how much the two regression equations can differ. After experimenting, briefly explain how you think housing pricing works—or can you tell?