geopii 3 circles logo geopii

location / value / technology

facebook   twitter  

Spatialest®


 

everything you ever
wanted
to know about
Spatialest
but were afraid
to ask

 

Spatialest is a unique software package which uses comparable sales information, property comparison features, and location to produce estimates of value. It offers a comprehensive range of GIS and statistical functionality enabling both novice and advanced users to generate high quality accurate appraisal models. The functionality ranges from EDA, time trending, non-spatial and spatial modeling to equity and post modeling analysis.

 

 

Exploratory Data Analysis

There is a wealth of functionality to enable the analyst to quickly identify outliers, find invalid information, look for invalid sales patterns or trends in the data. These can all be performed via simple analytical steps. More advanced users can utilize the full spectrum of SQL scripts to further aid analysis.

 

In addition to a comprehensive list of statistical functionality there is an easy to use graphing option which provides the ability to visualize data via scatterplots, boxplots, pie charts, or histogram charts. These all provide an interactive interface which enable on-screen selections to be performed which can then be mapped or tagged for further analysis.

  • Quickly identify key patterns and trends in the data
  • Stratify or remove outliers, geographically or statistically
  • Visualize relationships in the data and interact with the results
 

EDA Functionality

Descriptives Tabulated data reporting descriptive statistics on numeric variables such as Min, Max, Count, Mean and Std Dev
Frequencies Describes distribution of data into a number of classes with a count of occurrences in each
Crosstab Describes the distribution of two or more variables simultaneously
Correlations Describes the relationship between pairs of variables
Graphs Scatterplots, boxplots and histograms
Grids Return descriptive statistics on variables based on a specific geographic region.
Missing Value Analysis Accommodate for missing values in data by treating the missing values as base, a suitable alternative, or by ignoring those values altogether so that results are not adversely affected or skewed
Group Variables Combine similar attributes within a variable for the purpose of achieving an optimum number of sales for modeling and analysis purposes. When there is a lack of representativeness or lack of sales for certain attributes, where possible, they should be combined with another attributes.
Create new variables Create new variables and to modify the values of existing variables. Compute statements follow a standard syntax based on VB.Net
Find Duplicates The Find Duplicates analysis will return the number of occurrences of duplicated records up until the Limit set by the user.
Property Summaries View summary statistics of a collection of occurrences per unique combination of selected properties/attributes. For example, report upon the average SPPSF in each neighborhood, categorized by grade or quality of property.
 
 

Comprehensive suite of GIS Functionality

Spatialest has been engineered with GIS at the forefront of its interface. No other solution in our industry can claim this. Consequently the user has access to all the visualization and analytical capabilities of a GIS combined with the more traditional techniques and functions of CAMA modeling, such as Regression Analysis.

Our GIS-CAMA integration extends much further than simply linking to a property parcel. The Spatialest ethos is about enabling geography to be fundamental to analysis and this is facilitated through our powerful interface. Simple spatial queries can be performed as well as advanced features such as surface interpolations. All these features are linked to the property data file so that they are interoperable with the exploratory analysis, graphing toolsets, or model building applications.

 

Spatialest is a unique software package which uses comparable sales information, property comparison features, and geographic location to produce estimates of value. The software has been designed specifically for the assessment industry and avoids the need for multiple applications making complex analysis easy for both novices and experts alike. spatialest® offers the complete GIS-CAMA integration.

  • True GIS-CAMA integration
  • Visualize and understand spatial patterns and trends
  • Leverage existing GIS data
 

GIS Functionality

Add GIS data Utilize ESRI shapefiles, SDE layers, and raster image layers
Select Data Geographically Write select queries in the map interface or use spatial select tools to identify features or generate models based on geography.
Use Buffers Create distance buffers around features. For instance, this might be used to identify properties which fall within a flood plain.
Utilize online map data Review comparables using Google online maps and street view imagery.
Create interpolated surfaces Create a value prediction surface in order to identify key patterns and trends in the data.
Visualization Use Quantiles, Graduated Symbols, Ratio Maps, and Value Maps to symbolize quantitative or qualitative data
Spatially Disparate Regions Model relatively homogeneous data from spatially disparate regions
Temporal Patterns & Trends Assess and review trends and patterns which are often not evident from tabular analysis
 
 

Time Adjustments

Data that has been collected over a period of time, such as house sales, will normally be affected by trends within that time period. Consequently, a Time Adjustment(sometimes referred to as Seasonal Adjustment) must be made to data to align all records to a particular date; the Valuation or Antecedent date.

The time adjustment method computes a new sale price figure from the existing sale price based on the overall price trend over the date period selected.

 
 
 

Regression Analysis

Model Builder Wizard

The Model Builder Wizard empowers the user with the most intuitive and powerful model building solutions in the industry and enables the user to quickly produce a model without any prior knowledge of statistics. . At the click of a button the user can produce an MRA model (either Additive or Multiplicative) complete with performance statistics for review.

The intelligent interface allows even the most novice user the ability to analyze data quickly, easily and appropriately. At the click of a button the user can produce an MRA model (either Additive or Multiplicative) complete with performance statistics for review. Additionally, the 'Model Builder Wizard' has built in checks to ensure that models are of sound statistical quality. For instance, the issue of multi-colinearity is addressed whereby the wizard will perform correlation checks and drop redundant variables.

 
  • Easy to use
  • No programming required
  • Specifically designed for Assessment Industry

More advanced users have the opportunity to customize features or, if they prefer, generate their own MRA model outside of the wizard. However, we would still recommend that they utilize the functionality in this toolset as it provides a wealth of exploratory information about the key market drivers in the data that might be ignored in more independent analysis.

Multiple Regression Analysis (MRA)

Spatialest can create a regression via the Wizard or through the Regression script. The Regression script offers the user more flexibility.

  • Additive Form
  • Multiplicative Form
  • Manage Nulls
  • Force Attributes
  • Output and Report Results
  • Analysis often carried out in SPSS can be easily replicated in spatialest but the processes are carried out much more easily and quicker in a more intuitive environment. For example, the creation of binary variables, irrespective of the number of distinct values, is completed in seconds within spatialest. The image below shows a comparison between SPSS and spatialest and illustrates the work involved in handling qualitative variables (such as Quality, Style, Neighborhoods etc) in the regression model.
 
 

Comparable Model

Spatialest contains the most powerful and advanced method of property value prediction. The Comparable Model process introduces location to the valuation process by selecting the optimum comparables. The patented process is unlike other 'comp selectors' in that it does two things:

  • It selects visually similar properties, those which exhibit the most similar property characteristics to that of the subject property, such as property type, size, grade or number of bedrooms
  • It selects nearby similar properties

 

 

An estimate is generated where a specified number of comps which meet the comp rules are found within a specified search distance. This means that every property undergoes a process similar to that undertaken for a Fee Appraisal and has the benefits of:

  • Reducing the level of appeals
  • Providing an easy to understand process for staff and taxpayers alike