Page 25 - Regression Guideline

Underwriters"quesJon"and" response"
How appraisers are responding to underwriters question about regression
Q. Why is your search so big of an area? (Rural area)
A. The age, GLA, distance and time parameters is used to develop a large data pool to be able to extract those items that are reliable value factors. Those parameters are not excluding features, in fact by having a broad sample of homes that actually sold enables the statistical analysis to find what features can be reliable in valuating different features. The fact is, acreage sites, basement area, basement finish and walkout basements are analyzed and they may be collinear with one or more other predictors, and so, have no independent contribution to the multiple regression model. The SAVVI report search parameters is a 20 mile radius and sales up to 48 months old to be able to build a statistical base of information of at least 200 sales in order to determine what features are actually relevant to value. The 20 mile radius is within the same marketing area when considering homes similar to the subject in GLA and acreage. Homes that sold more than 12 months ago could be adjusted for their date of sale, and if used would and could be adjusted based on the time graph which was included in the SAVVI report. Just because a sale may have occurred 12-48 months ago does not mean it is not relevant, it just means that a time adjustment should be considered if used.
Q: Why didn’t you adjust for Bedroom, acreage and walkout basement
A. Closer analysis of the SAVVI report would show that features such as bedrooms, acreage and walkout basements were analyzed but found not to be, by themselves, a 95% reliable value indicator for properties similar to the subject. The market data to support the adjustments or lack of adjustment is the essence of the SAVVI Report. The support for the adjustments used is in the quantity of sales used to arrive at those adjustments. Those adjustments are based on home sales of properties 30% larger, 30% smaller, 20 years older, 20 years younger, sales within a 20 mile radius over the past 48 months. View adjustments can not be determined through the Hedonic Regression analysis. View adjustments are based on paired sales analysis over the past two years for homes in the subject’s marketing area. Golf and pond views have similar appeal and not adjustment was necessary between golf and pond view.
the analysis and market data used to determine that bedroom and bathroom adjustments were not warranted is in the SAVVI report that was made part of the appraisal report. In a Hedonic Regression analysis, bedroom or bathroom adjustments may or may not be applicable solely dependent on the statistical analysis of the similar properties that were used in the analysis.
24"