Page 50 - Regression Guideline for AMC

SUMMARY
Summary
This overview of evalua6ng mul6ple regression modeling applied to real estate appraisal has a.empted to highlight important aspects of the modeling. As men6oned at the beginning, regression is a powerful but complex sta6s6cal technique for making real estate market predic6ons. Used correctly, and with near complete and good quality data, it can produce accurate es6mates of the value of proper6es in a given market. However, used incorrectly and with poor or suspect quality data, it can just as easily produce inaccurate results, the inaccuracy of which will be masked by the sophis6cated analy6cs that underlie regression analysis.
To minimize that prospect, we have presented 5 key ques6ons that should be asked about any appraisal results based on sta6s6cal regression. These are not foolproof but they will help reduce the chance that the regression model has produced an inaccurate es6mate for a given subject property. To review, these ques6ons are:
50