- Back-testingAn AVM Back Testing environment is different from an AVM test environment. For an AVM test, by definition the benchmark value is based upon a recent verified sales price, but for AVM Back Testing the benchmark value is based upon an available (reconciled) appraised value that is (usually) not supported by a recent sale. The AVM(s) predict the appraised value and standard test metrics are applied. This outcomes analysis is compared to how the AVMs are employed in a production environment in lieu of the traditional valuation process.
- Base-Home ApproachA method of appraising single - Family residential properties whereby each residence to be appraised is compared with one having common or typical characteristics and of known value, called the base home, and differences between the two in terms of condition, size, number of garages, and the like are weighted by the appraiser in the determination of the value of the property to be appraised.
- Base-Lot MethodA method of appraising land parcels whereby each parcel to be appraised is compared with a parcel having common or typical characteristics and of known value, called the base lot, and differences between the two in terms of location, size, shape, topography, and the like are analyzed by the appraiser in estimating the value of the lot to be appraised.
- Benchmark(1) A term used in land surveying to mean a known point of reference. (2) In property appraisal, a property of known value and of known effective age and replacement cost. (3) By extension, a model property to be used in determining by comparison the grade or quality class of other properties.
- Benchmark ValueThe market value against an AVM’s value predictions will be compared for performance testing. Recent, arm’s length transactions are preferred.
- BiasThe degree to which an AVM or cascade tends to overvalue or undervalue subject properties. Analysts quantify an AVM’s bias by calculating the mean or median valuation error.
- Binary (Dummy) Variable (1) Binary variables are qualitative data items that have only two possibilities – yes or no (for example, corner location). (2) A variable for which only two values are possible, such as results from a yes-or-no question; for example, does this building have any fireplaces? Used in some models to separate the influence of categorical variables. Also called a dichotomous variable or dummy variable.
- Blended ModelAn AVM that incorporates a combination of valuation methods including but not limited to: Hedonic, home price Index, Hybrid, Tax Assessed value multipliers, statistical regression or an Appraisal Emulation approach.
- Blind TestingAVM performance testing where analysts remove any valuations for which the model may have had access to the benchmark value. Blind testing provides users with a better understanding of how a particular AVM or cascade will perform in the production setting.
- BootstrapA computer-intensive method of statistical inference that is based on a repeated resampling of data to provide more information about the population characteristics. The bootstrap is a data-driven procedure that is particularly useful for confidence interval approximation when no traditional formulas are available or the sample has been drawn from a population that does not conform to the normal distribution.
- Bump LogicUser-defined methods or criteria to select from or combine various classes of valuation services, e.g., AVM, BPO, hybrid valuation, drive-by or full appraisal.
- Bundle of RightsThe six basic rights associated with the private ownership of property: right to use; sell; rent or lease; enter or leave; give away; and refuse to do any of these.
b) AVMs 201: A Practical Guide to the Implementation of Automated Valuation Models, Jim Kirchmeyer, 2008.
d) Collateral Assessment & Technologies Committee, Summary of Definitions & Terms, 2006.
e) Joint Industry Task Force on AVMs, IAAO Standard on AVM Glossary, September 2003. https://www.iaao.org/media/standards/AVM_STANDARD.pdf
f) Appraisal Institute, Joint Industry Task Force on Automated Valuation Models, Work Group Terminology, 2005.
g) Merriam-Webster (https://www.merriam-webster.com/)