A quarterly housing report should allow a reader to identify the geography, property type, period, and source behind every statistic. A headline about Boston prices is difficult to evaluate when it silently combines the city with a larger region or compares condominiums with a different mix of homes. The reporting method belongs beside the result, not hidden after a dramatic conclusion.
This guide describes how to assemble a reproducible snapshot. It is not a report of current quarterly prices or inventory. The small numerical examples are invented to illustrate calculations. Use dated source material for actual figures and retain the definitions supplied with it. When the evidence does not support a comparison, say so rather than filling the gap with a confident estimate.
Write the reporting question first
Choose a question narrow enough to answer. For example, you might examine closed condominium sales in a defined geography during a particular quarter. That is different from asking what every home in Greater Boston is worth or whether a buyer should purchase immediately.
Write the question in the report notes before collecting numbers. It will guide which sources and fields are relevant. If the available source cannot answer the original question, either change the question openly or obtain appropriate evidence. Do not switch definitions halfway through the process and retain a headline that describes something else.
Define the geography explicitly
Specify whether the report concerns the City of Boston, particular neighborhoods, named municipalities, or a source defined regional market. Record the source’s boundary definition where available. A metropolitan area, county, association coverage area, and municipal boundary should not be treated as interchangeable.
If you create your own grouping, list its components. Explain how an address is assigned and how ambiguous records are handled. A reader should be able to reproduce the grouping without guessing what “near Boston” means. For neighborhood work, also explain whether you used an official boundary, a provider’s classification, or a clearly labeled editorial grouping.
Keep property types separate
Decide which property categories are included and preserve the source’s definitions. Mixing condominiums, single family properties, and buildings with multiple units can change a reported median even when prices within each category behave differently. A combined number may be useful for a defined purpose, but the composition should be visible.
The Greater Boston Association of REALTORS publishes housing market resources that distinguish report categories. Review the actual report and methodology relevant to your period. Cite the specific publication used rather than treating a general landing page as the evidence for a number you obtained elsewhere. Source: Greater Boston Association of REALTORS: Housing market data
Establish the dates and status rules
Record the start and end of the quarter, the extraction date, and the transaction status used. Closed sales, pending sales, and active listings describe different stages. Do not substitute one for another simply because it is the first available field in a spreadsheet.
Also decide how later revisions will be handled. A report extracted shortly after the period ends may differ from a later version. Preserve the original file and label the version. If you update a published figure, note the revision rather than silently presenting the new value as though it had always been the original result.
Build a data dictionary
Create a short definition for every field you use. Include the unit, source column, treatment of missing values, and any transformation. A price field should specify what price it represents. A time measure should explain which dates define it and whether the source already supplies a calculated value.
Use clear names in the working file. Avoid a column labeled simply “days” when it could mean time on market, time under contract, or time since extraction. A data dictionary may feel slow at first, but it prevents repeated interpretation errors when another person reviews the report or you return to it next quarter.
Preserve the original and work on a copy
Keep the source file unchanged and create a separate working copy. Record when and where it was obtained, along with any applicable access or reuse terms. Do not remove attribution or imply that a licensed or third party dataset was collected independently by Homzora.
Document each cleaning step. If you remove duplicates, define the duplicate key. If you exclude a record, record the reason. A row that looks unusual is not automatically wrong. Investigate it or explain the limitation rather than deleting it merely because it changes the result you expected to publish.
Calculate medians transparently
For a simple invented example, prices of $500,000, $550,000, $600,000, $650,000, and $900,000 have a median of $600,000. Their total is $3,200,000 and their mean is $640,000. The two statistics answer different questions and should not be swapped while retaining the same label.
This example is not a Boston price estimate. It demonstrates why the calculation and sample count matter. In an actual report, use the source’s documented methodology or explain your own. If the sample is small, make that limitation prominent rather than suggesting that a precise looking number describes every property in the area.
Calculate changes using comparable periods
Identify the earlier value and the later value before calculating a percentage change. In a hypothetical series rising from $600,000 to $630,000, the increase is $30,000 divided by $600,000, or 5 percent. State whether the comparison is with the preceding quarter or the same quarter a year earlier.
Even correct arithmetic does not establish that the samples are comparable. Review geography, property type, status, and methodology on both sides. Changes in the mix of transactions can affect the result. Avoid turning a change in the reported median into a claim that every individual home appreciated by that percentage.
Treat inventory as a defined measure
Explain exactly what the source counts as inventory and the date or period it represents. A count of active listings at a point in time is not automatically the same as the number of properties offered at any point during the quarter. Preserve the definition with the chart or table.
Do not construct a supply ratio without defining both the numerator and denominator. If a source publishes its own measure, use its explanation and avoid recreating it from incompatible figures. When the necessary inputs are unavailable, omit the metric or label the limitation rather than borrowing a familiar industry term for a different calculation.
Distinguish indices from transaction summaries
A price index and a median sale price should not be displayed as though they were the same measure. FHFA explains its House Price Index methodology and intended interpretation. Review the geographic and coverage details before using it alongside a local transaction report. Source: FHFA: House Price Index questions and methodology
An index can provide context without supplying an appraisal for a specific home. Explain that distinction in the narrative. If two series move differently, investigate their coverage and methods before calling one incorrect. Different measures can describe different aspects of the market, and a clear report helps the reader understand why rather than forcing them into a single story.
Handle survey estimates appropriately
If you add Census American Community Survey information, retain the estimate period and consider the margin of error. ACS estimates should not be presented as live asking rents or a count of current listings. The Census Bureau specifically addresses the need to consider uncertainty when comparing estimates. Source: Census Bureau: Comparing ACS estimates and margins of error
Keep contextual survey data in its own section. Explain why it is relevant and what it cannot establish about the current quarter. This protects the report from an attractive but invalid comparison between a multiyear household estimate and a recent transaction statistic. Similar labels do not make the underlying measures equivalent.
Audit the narrative against the calculations
Read every sentence containing a number and locate its source or formula. Check whether the headline describes the same period and geography as the table. Remove claims about causes unless you have evidence that supports them; a change occurring alongside another event does not by itself establish causation.
Ask a second reader to reconstruct one calculation from the notes. If they cannot, improve the documentation. Check rounding consistently and ensure that chart labels match the units. A polished visual does not compensate for an unclear denominator, an omitted sample count, or a comparison that crosses incompatible definitions.
## Preserve a publication checklist
Before releasing the snapshot, confirm that the title, tables, captions, and downloadable file all use the same definitions and version. Check that each link leads to the intended source and that any reuse conditions have been respected. A correct calculation can still mislead if its chart title names a different geography or period.
Keep a short record of the checks completed and any limitation that remains. If the report contains original calculations, label them as such and retain the formulas. This makes a future correction easier to explain. It also allows the next reporting cycle to begin from a documented method instead of reconstructing the process from the appearance of the previous page.
A checklist should support judgment rather than conceal it. If a result looks unusual, investigate the data even when every formatting box has been checked.
Publish the limits with the findings
Include the source, reporting period, extraction date, categories, sample count where available, and important exclusions. Explain whether the report will be revised. Link readers to the underlying publication or permitted source access, and distinguish your calculations from figures published directly by the provider.
The final product should make both its findings and its boundaries understandable. A reproducible Boston snapshot is valuable because another reader can follow the evidence. It should not need an unsupported forecast, a fabricated current price, or a claim that the data alone determines the right decision for every buyer and seller.
Sources and further reading
Sources reviewed October 8, 2026. Program rules, product terms and public information can change. Hypothetical examples and editorial checklists are identified in the article.