What adjustment support does DataMaster provide?

Lenders increasingly expect adjustments to be backed by market evidence rather than judgment alone. DataMaster derives adjustment values from the closed sales using your own data and shows you the math behind each one.

DataMaster analyzes the market comparable sales already in your report and estimates what the market is paying for each property feature — in dollars per unit (eg. /sqft,/sqft,/Bedroom, $/Bathroom).

• It uses your data, not national averages. Every figure comes from the closed sales you import into your own report that pass your search filters. Active, pending, and expired listings are excluded.

• It shows three independent estimates. Each adjustment is computed using three different methods (below) so you can see whether they agree and pick the best one. Agreement is strong support; disagreement tells you to dig deeper.

• It shows the supporting evidence. For every value you can open the underlying sale pairs, the price distribution, the regression scatter and equation, and the summary statistics — so the number is transparent and defensible to your clients, not a closed box.

• It rates confidence. Each estimate carries a Support badge (how much data backed it) and a Stability badge (how consistent that data was), rated High / Medium / Low.

You'll find these under Market Analysis in a UAD 3.6 report, on the Feature Adjustments and Time Adjustments tabs.

Once you've decided on a value, the Power Adjust tool on the Sales Comparison Grid lets you enter an adjustment rule once for a grid row — for example $2.50 per Sq. Ft. — and apply it across every comparable automatically, with per-comp override where you need it.

DataMaster provides support for time and several common feature adjustments.

DataMaster estimates a dollar-per-unit value for:

Yes/no features (pool, basement, heating, cooling, attached) are priced as the market's dollar difference between an otherwise-comparable home with and without that feature.

DataMaster fits a price trend over your market's sales history and measures the movement from each comparable's date to your effective date, returning both a percentage and a dollar amount per comp. You choose whether to trend on contract date or sale date, and on median or average price.

Condition, quality, view, site influence, neighborhood, amenities, design, functional utility, garage and similar ratings aren't statistically derived — the market data doesn't isolate them cleanly. Those rows are yours to judge, and Power Adjust will apply your chosen adjustment consistently across the grid.

DataMaster runs three common statistical approaches on every feature, side-by-side.

The classic textbook method. Two closed sales form a valid pair for a feature only when they differ on that feature alone and match on everything else within a tolerance band. The reported adjustment is the median across all valid pairs.

pair adjustment = (price of higher comp − price of lower comp) ÷ (feature difference)

Tolerances define "close enough to match" and default to 10% on GLA and lot size, 30 days on sale date, 5 years on year built, and 1 on bedrooms and bathrooms. You can override them on the Tolerances tab. Cleanest evidence, but strict rules mean some features find few pairs.

Solves the main weakness of strict pairs: features that always move together (size, baths, luxury) rarely appear in isolation, so in a typical pool most features find no clean pairs at all. Grouped Pairs needs no twins — it splits the sales into groups and reads the adjustment off the difference between the groups' typical prices, the way an appraiser would with grouped data.

• Form the groups. For a yes/no feature, compare homes with it to homes without it. For a count, compare one observed level to the next (3 bedrooms vs 4). For a size feature, compare the larger-home half to the smaller-home half.

• Compare like with like. Groups are compared within strata of similar homes — for example within bands of similar size — so a feature isn't credited for things that travel with it (a pool home is usually also a larger home).

• Read the value. Each feature's adjustment is the difference between the groups' typical (median) prices, combined across the comparisons, with a confidence range reported alongside it.

Uses every comparable at once instead of two at a time. An ordinary least-squares model estimates each feature's marginal dollar value while holding the others constant; each coefficient is that feature's adjustment.

Reported alongside it: R² and adjusted R² (how much of price the model explains), p-value (whether the feature's effect is statistically meaningful), standard error and t-statistic, and VIF (a warning that two features are too correlated to separate cleanly — GLA and above-grade area, for instance). Best when you have a large comp set; weaker with only a handful of sales.

Time adjustments use a price index fitted to the same closed sales in your report — your own market data, not a regional or national average.

A few key points on how it's built:

• It compares like with like. A simple "average price by month" is misleading, because a month that happened to sell larger or nicer homes looks like a price jump even when the market didn't move. The index accounts for the differences between the homes that sold each month, so it measures a change in price level, not a change in the mix of what sold.

• The index is non-linear. The index tracks a market that accelerated, cooled, or turned, rather than being locked to a single straight-line rate. At the same time it's built to stay stable on thin or noisy data, so a small pool of sales doesn't produce wild month-to-month swings.

• It reads like an index. Just like a stock market index, the time adjustment index is piublished starting at 100 and moving up or down from there. The headline is the trailing change (12 months by default), and each comp's suggested adjustment is the movement from its own date to your effective date. A comp that sold on or after your effective date gets no time adjustment — the index never projects past the data it was built from. Ideally, 12+ months of sales history are required.

In short, here's a quick table summarizing the support:

Compare all three, choose the value you can defend, then use Power Adjust to apply it across the grid.