Get in touch

Specta · Independent data analysis

Policing and public-safety data analysis.

For newsrooms, nonprofits, legal teams and oversight bodies. Methods first, findings tested against rival explanations, limits stated.

Start a conversation See published work

Services

Three fixed-scope engagements

Each starts with a scope call and a fixed-price statement of work.

01

Methods review

We check an analysis for errors and bias before it is published.

From $1,5001 to 2 weeks

What's included

Who it's for

Reporters, editors and researchers with an analysis about to go public.

What you get

  • A written review of the data, code and claims
  • Checks for wrong population counts, record breaks, miscoded values and groups too small to compare
  • Key figures rerun from source
  • A call to walk through it
02

City data analysis

Rates per resident for your city, with the plain explanations tested.

From $3,0003 to 5 weeks

What's included

Who it's for

Local newsrooms, advocates and community groups with a question about their city.

What you get

  • An audit of the source data
  • Rates per 100,000 residents for each group
  • Tests of the plain explanations: age, place, offense, time
  • Charts ready to publish, a methods note and the code
03

Litigation and oversight support

Disparity analysis of stops, arrests or charges, built to hold up under scrutiny.

Quoted per matterSet by the case schedule

What's included

Who it's for

Public defenders, civil rights lawyers, and city, county or state oversight bodies.

What you get

  • Enforcement by group, against residents and reported crime
  • Comparison choices and rival explanations tested
  • A written report with reproducible code
  • Confidential, under your agreement or protective order

Timelines start when we have the data. Half is paid up front, half on delivery.

Selected work

Published analyses

From Residents Count, our research publication. Each links to the full analysis, code and data.

Washington, DC · 2025 to 2026Oct 7, 2026

DC crime since the National Guard

167 → 94homicides, from the year before to the first Guard year
Question
Did DC crime fall after the National Guard arrived, and more where troops were posted?
What we found
Homicides fell from 167 to 94, and gunshots detected fell 44%. Areas near troop posts did no better than comparable areas on any of six measures.
Method and limits
What we did
Compared the first Guard year with the prior 10, using incident data and gunshot sensors, then tested about 800 areas near and away from troop posts under a published plan.
Limits
The data can't say what drove the drop. Post locations were rebuilt from public sources.
Read the report
59 large US cities · 2022 to 2025Oct 3, 2026

Assault victims in 59 large US cities

2,552,921assault victims in FBI records, one method for every city
Question
Does the gap in reported assault rates between Black women and other women hold across large US cities?
What we found
Yes. In all 46 cities with complete data, Black women's reported assault rate is higher than White women's: about 3.8 times in the typical city, and 2.4 times Hispanic women's. Nothing measured explains it.
Method and limits
What we did
One method for every city, chosen by a rule set before the results. Tested age, intimate partner assault and how officers record race.
Limits
Counts only assaults reported to police, and willingness to report differs by group and place. The national victimization survey shows a smaller gap; a companion report tests why.
Read the report
Washington, DC · SeriesOct 8, 2026

Policing in DC

5analyses of one police department's public records
Question
Where do DC police stops, searches and arrests land?
What we found
In every district, the Black share of people stopped exceeds the Black share of residents (1.82 times citywide). Black people who are stopped are patted down 1.7 to 4.5 times as often as others, depending on the district, while searches of both groups find something at about the same rate. Daylight doesn't change who is stopped.
Method and limits
What we did
Five tests of the department's public records, from stops by district to downgraded crime reports.
Limits
Records count stops, not people, and race is as recorded by officers. Resident comparisons miss people who work in or visit a district.
Read the series
New York, NY · 2022 to 2025Oct 8, 2026

Police response times in New York

13.8M911 calls from the public, across 197 neighborhoods
Question
Do 911 calls wait longer for police in some New York neighborhoods than in others?
What we found
Critical calls wait under a minute longer in lower-income, majority-Black and majority-Hispanic neighborhoods; less urgent calls wait 28% to 79% longer. The gaps sit between precincts, not within them.
Method and limits
What we did
Compared like calls at like hours across neighborhoods grouped by income and makeup, under a pre-set plan, with early 2026 held back as a check.
Limits
The records show how long calls waited, not why. About 40% of calls not about a crime in progress have no arrival time.
Read the report

Every report, with its code and data, is at residentscount.com

How we work

Agree on the question and the method, then follow the data

  1. Scope call

    The question, the data, the deadline. If the data can't answer it, we say so.

  2. Statement of work

    Scope, deliverables, timeline and a fixed price, on one page.

  3. Analysis

    Audit the data, compute the rates, test rival explanations. Every decision written down.

  4. Review

    Checked for errors and overreach. You see draft findings first.

  5. Delivery with methods

    Findings, charts, a methods note and the code.

What we hold to

  • Results are reported whichever way they point.

    Findings that cut against the client stay in.

  • Data is handled under agreement.

    Used only for your project, then returned or deleted.

  • No outcome is promised in advance.

    We agree on the method, not the answer.

  • Limits are stated.

    Every deliverable says what the data can't show.

About

The person you email does the work

Martin NgohSenior data scientist · M.S. Business Analytics, Georgetown

Specta LLC is Martin Ngoh's practice. He builds machine learning and data systems in production and does the analysis on every engagement.

Also from Specta

Residents Count

Our research publication: police and public records as rates per resident, city by city.

residentscount.com Lease Screener

Reads a residential lease and flags clauses for renters in DC and Virginia.

leasescreener.com

Contact

Start with an email

Tell us the question and the data you have.

Please don't attach sensitive data to a first email.

What to include in your first email

  1. Who you are and your organization.
  2. The question, in a sentence or two.
  3. The data: what it is, where it's from, how big.
  4. Your deadline, and whether the work will be published, used in a case or kept internal.
  5. Any limits on the data, such as a protective order.