Section 9-6 - Statistical Analysis and Standards (a) These standards apply to residential valuations performed by the Assessor with department approved software, and are directed primarily at the reliability of sales ratio study calculations performed by the Assessor for each LEA, Neighborhood or other stratum in the county. IAAO standards apply except where there is a conflict with these rules, in which case these rules shall prevail. - (i) Sales Sample Sizes for Ratio Studies. A ratio study is valid to the extent that the sample of sold properties is sufficiently representative of the population being appraised. The minimum sample size shall be 5 for any LEA, Neighborhood or other stratum. If five sales are unavailable the following methods should be used to increase sample size. If methods (A.), (B) and (C.) do not result in compliance with the statistical standards herein then method (D.) may be considered.
- (A) Restratification. If levels of appraisal are similar or properties are homogenous, broader strata containing larger samples can be created by combining existing strata or by stratifying on a different basis.
- (B) Extending the period from which sales are drawn. This is often the most practical and effective approach. Sales from prior years can be used; however, adjusting the sale price for time may be necessary and significant property characteristics must not change.
- (C) Enlarging the sample by validating previously rejected sales. Sales previously excluded from the analysis, because it was not administratively expedient to confirm them or to make adjustments, can be reevaluated.
- (D) Imputing appraisal performance. Ratio study statistics for strata with no or few sales can sometimes be imputed from the results obtained for other strata. These strata should be as similar as possible. Procedures and techniques used to appraise properties in the strata also should be similar.
- (ii) Appraisal Uniformity. If a subset of individual sales ratios within a sample is exceptionally higher or lower than the norm in the sample, the assessor should first look for common property characteristics among them and determine if they constitute justification for the creation of a new neighborhood. In the event no such common characteristics exist, the assessor may filter outliers in accordance with Appendix B of the IAAO Standard on Ratio Studies (2013). The COD for any LEA, Neighborhood or other stratum of residential and commercial, improved and vacant, shall conform with Section 9.2 of the IAAO Standard on Ratio Studies (2013).
- (iii) Level of Appraisal. The level of appraisal for any LEA, Neighborhood or other stratum of residential improved or vacant property, shall lie between 0.90 and 1.10. To protect against the likelihood of over assessment of properties the recommended Level of Assessment for residential improved properties is less than 1.00.
- (iv) Calculating Market Adjustments. The Level of Assessment for any Neighborhood shall annually be calculated by varying the Market Adjustment Factor until the final desired Level of Assessment is achieved, using software approved by the Department. One final Market Adjustment Factor per property type shall be applied to the CAMA-generated RCNLD for each sold or unsold property in the Neighborhood unless justified and documented by the Assessor. If analysis indicates no adjustment to RCNLD is necessary to meet the level of appraisal requirements, making no adjustment is appropriate. However making no adjustment to RCNLD based solely on a lack of sales is not appropriate.
- (v) Confidence Interval of Mean or Median Sales Ratio. For residential properties a 95% confidence interval will be computed around the level of appraisal for any LEA, Neighborhood or other stratum and must include one or more points in the range of .90 and 1.10.
- (vi) Land Appraisal. Within any LEA any data points of Assessed Valuation when plotted against land area prior to applying land attribute adjustments, should fit a regression line which is derived from the sold properties. The equation of the regression lines is used to assess the land value of all sold and unsold properties in the LEA. Each stratum within the LEA should have only one regression line for all its parcels.
- (vii) Undefined Statistical Calculations or Adjustments. Calculations or adjustments of a statistical nature, not addressed by these rules or the IAAO Standards, shall not be applied to any properties in a county by the Assessor without a written memorandum of clarification from the Department requested by the Assessor of the Department under WS 39-11-102. The Department's memorandum shall be distributed to all County Assessors.