Service · Research analytics

Statistical consulting for defensible research.

Founder-led20 yearsReply within 24 hFixed fee in writingNo ghostwriting

For research teams, PhD candidates, and independent researchers who need a statistician for a defined study: explicit assumptions, appropriate methods, reproducible code, honest uncertainty, and a technical record that withstands peer review.

High-level context is enough; do not send confidential data. We reply within 24 hours with a fit assessment and the next useful step — no sales pitch, no obligation. We can sign an NDA before reviewing non-public data or materials. Prefer email? hello@statgazer.com

Clear starting scope

Choose the smallest useful engagement.

Statistical consulting and quantitative research support in three scopes: start with a focused review or scope a reproducible analysis. The exact question, deliverables, timing, and fee are agreed in writing before paid work begins. Send a one-line brief and we reply within 24 hours with a fit assessment and the next useful step. No sales pitch, no obligation.

For client-owned research only · no ghostwriting or coursework completion.

Founder-led delivery. Twenty years of private teaching and consulting across econometrics, statistics, machine learning, and applied quantitative methods.

Three ways to work with us

In every mode the research stays yours: your data, your decisions, your authorship.

Focused review

Methods & Analysis Review

From $750 USD

Know what to change, and why.

Perfect if: reviewers or your committee questioned your statistics, or you want an independent read of your methods before you submit.

Request a Methods Review
  • One focused research question.
  • Systematic desk review of the agreed study design, methods, and prepared outputs.
  • Prioritized written findings: the evidence, why it matters, and what to change.
  • One consolidated round of clarification by email is included within the agreed scope. Send your questions within 14 calendar days of delivery.
  • Five to seven business days from complete materials and payment.
Read sample review notes

Synthetic example · no client data

What you send · what you get · not included

You send

  • Your manuscript, protocol, or analysis plan.
  • Your existing outputs — tables, figures, model summaries — and the code or notebook if you have one.
  • One focused question you want answered.

You get

  • Written review notes: design and methods appropriateness, assumptions, threats to validity, and whether your outputs support your interpretations — prioritized, with what to change and why.
  • One consolidated round of clarification by email is included within the agreed scope. Send your questions within 14 calendar days of delivery.
  • Delivery in five to seven business days from complete materials and payment.

Not included

  • New analysis, data cleaning, code repair, independent re-running of computations, new materials, and an expanded scope are excluded; any additional work is quoted separately before work begins. We do not write or edit text for submission.
Request a Methods Review

Inside a Methods Review

A clear path from finding to next step.

See how we turn existing methods and outputs into review notes you can act on: a clear assessment, prioritized findings, and questions for written follow-up.

Read the sample review

Illustrative notes on a synthetic study. No client data or new analysis.

Sample review notes01 / 03

Finding 01 · Material · Excerpt

Document why the adjustment supports the intended claim.

Evidence
The reported difference falls from +5.55 points unadjusted to +0.96 after adjustment for baseline score, need, age, and site.
Why it matters
Readers need the rationale and measurement timing to assess whether the model addresses the intended question.
Action
Explain the adjustment set and measurement timing, and identify the record supporting prespecification.

Your own research

Written Research Consultation

From $500 USD

Perfect if: you need a recommendation on one specific design, method, or interpretation question in your own study.

Working on your dissertation, thesis, or research paper? Send one focused question and only the excerpts needed to answer it. Receive a concise written memo with a recommendation, rationale, assumptions, limitations, and practical next steps. A systematic review of agreed methods and prepared outputs belongs in Methods & Analysis Review.

Request a written consultation

Ethical boundary: no ghostwriting, coursework completion, fabricated data, guaranteed significance, or submission-ready thesis or manuscript writing.

What you send · what you get · where we draw the line

You send

  • A short description of your study and where you are in it.
  • Only the methods or output excerpts needed to answer the agreed question.
  • One focused question you want answered.

You get

  • A concise written memo on one focused question: recommendation, rationale, assumptions, limitations, and practical next steps.
  • One consolidated round of clarification by email is included within the agreed scope. Send your questions within 14 calendar days of delivery.
  • The delivery date is agreed in writing before payment.

We will

  • Assess the specific design, method, or interpretation choice in your question.
  • Explain the recommendation, assumptions, and limitations in writing.
  • Set out practical next steps you can carry out yourself.
  • Work with your advisor's knowledge — and, on request, describe our role for your committee in one paragraph.

We won't

  • Provide a systematic Methods & Analysis Review or carry out new analysis as part of this consultation. New analyses, new materials, or an expanded scope are quoted separately before work begins.
  • Write or edit your thesis or manuscript for submission.
  • Perform the analysis for a dissertation or any assessed academic work and hand you results to present as your own.
  • Complete coursework or assignments.
  • Fabricate, "fix," or select data.
  • Promise significance, acceptance, or a passed defense.

You remain responsible for the research, analytical decisions, writing, disclosures, and compliance with your institution's rules.

If our input materially shaped your methods, we are glad to be acknowledged — and equally fine not to be. Co-authorship is only ever at your and your advisor's discretion, and never a condition of working with us.

Working model

Reviewable at every stage.

We agree the decision, analysis boundary, available evidence, and review audience first. Work then proceeds in written increments so your team can challenge assumptions before they become expensive to change.

1. Frame

Define the question, outcome, data-generating process, decision context, constraints, and what evidence would alter the conclusion.

2. Analyze and challenge

Build the analysis, test assumptions, compare reasonable alternatives, document uncertainty, and resolve material discrepancies.

3. Hand off

Deliver the code, evidence package, written findings, limitations, and operating notes your team needs to reproduce the work.

Commercial fit

Built for teams accountable for the result.

StatGazer works best where the analysis will inform a publication, policy, grant milestone, product decision, or technical review and the team needs an inspectable record of how the conclusion was reached.

Good fit

  • A research team needs statistical depth that is not available internally.
  • An existing analysis needs independent review before external scrutiny.
  • The data or model is complex enough that reproducibility and handoff matter.
  • The team wants direct senior involvement from scoping through delivery.

Not a fit

  • You need fabricated data, a predetermined result, or a guaranteed significant finding.
  • You want ghostwritten academic work or completion of a student assignment.
  • You need clinical, regulatory, legal, or research-ethics approval from StatGazer.
  • You cannot provide enough context to evaluate data provenance and study design.

Practical details

Questions researchers ask before they write to us.

How much does it cost?

Three published starting points: a Written Research Consultation from $500, a Methods & Analysis Review from $750, and a Statistical Analysis Engagement from $1,500. "From" means the published scope; anything larger — more questions, more models, messier data — is quoted in writing before paid work begins, and you can stop at the quote.

How long does it take?

A Methods & Analysis Review is delivered in five to seven business days from complete materials and payment. For a Written Research Consultation, the delivery date is agreed in writing before payment. A Statistical Analysis Engagement gets its timeline in writing after a short data-readiness review; most run two to four weeks. If you have a hard deadline, say so in the brief — we tell you in the first reply whether it is realistic.

Who actually does the work?

The founder. There is no bench, no junior analyst, and no handoff between the person who scopes your project and the person who delivers it.

Which software do you work in? Do I have to switch?

Analysis is delivered in R or Python with the code, so you can rerun every step. We review SPSS, Stata, and SAS outputs as well; if your department requires one of those, tell us in the brief.

Can you review an analysis I have already done?

Yes — that is exactly what the Methods & Analysis Review is for. Send the methods, the outputs, and the question that worries you. You get prioritized written findings and the included clarification round by email. Within the agreed Methods Review scope, we can flag technical-accuracy and interpretation issues in a draft you wrote; you make every edit. Draft review is not included in the $500 Written Research Consultation.

My data is messy. Is that a problem?

Not a problem, but it is a separate line. The published scopes assume analysis-ready data. If cleaning, merging, or reshaping is needed, we quote it separately after seeing a sample, so the analysis fee stays predictable.

Is it ethical to hire a statistician for my dissertation or paper?

Yes — as long as you remain the author and follow your institution's rules. Written Research Consultation and Methods & Analysis Review provide explanations in writing within the selected package and agreed scope. Neither includes new analysis, live teaching, calls, or writing or editing text for submission. We do not complete assessed academic work for you. We can describe our role to your advisor or committee on request.

Will my paper be accepted? Will I pass my defense?

No consultant can ethically promise that, and you should be wary of anyone who does. What we can promise is that the methods we review or deliver will be appropriate, documented, and defensible — and that you will understand them well enough to defend them yourself.

How is my data handled? Do we need an NDA?

Keep confidential data out of the first message. If your materials are sensitive, we sign an NDA before you send anything. Access is founder-only unless agreed otherwise, materials are used solely for your engagement, are never used to train third-party AI models, and are deleted on request when we are done.

Can I put your fee in a grant budget?

Yes. Funders routinely accept "statistical consulting services" as a direct cost. We provide a written scope and quote you can attach to a budget justification, and we can invoice the institution directly.

Should I acknowledge you, or list you as a co-author?

Follow your venue's norms. If our input materially shaped the methods, an acknowledgement is welcome and never required. Co-authorship is at your and your advisor's discretion and is never a condition of working with us.

How do we pay?

By Stripe invoice — card or bank transfer. Consultations and Methods & Analysis Reviews are paid before work starts. Statistical Analysis Engagements are 50% on the signed scope and 50% on delivery. Institutions can be invoiced directly.

What if I need more changes after delivery?

For Written Research Consultations and Methods & Analysis Reviews: One consolidated round of clarification by email is included within the agreed scope. Send your questions within 14 calendar days of delivery. New analyses, new materials, or an expanded scope are quoted separately before work begins. A Statistical Analysis Engagement includes one defined revision round.

Sample work product

See what the headline estimate is hiding.

A research memo should name the question, estimand, assumptions, uncertainty, and the checks that change the conclusion. This deliberately confounded synthetic program-evaluation example shows that record — it is not a client result.

StatGazer / Research Analysis Memo Ref. SYNTHETIC-RESEARCH-001

Illustrative · synthetic data · not a client study

Analysis and robustness extract

Program evaluation — the unadjusted comparison is not decision-ready.

Question
Among eligible participants, how does voluntary program participation relate to the 12-week outcome score?
Estimand
Covariate-adjusted mean difference among the synthetic eligible records.
Sample
1,200 synthetic records · 573 participant records · 627 comparison records · 4 synthetic sites. Complete 12-week outcomes by construction; no simulated attrition or missing data.
Primary method
Prespecified adjustment for baseline score, need, age, and site · heteroskedasticity-consistent HC1 95% intervals.
Specification results from an illustrative synthetic program evaluation, not a client study or evidence of real-world program impact.
Specification Estimate HC1 95% interval
Unadjusted comparison +5.55 +4.47–+6.63
Prespecified adjustment +0.96 +0.09–+1.82
Flexible covariate form +0.95 +0.09–+1.82
Leave site A out +1.10 +0.05–+2.14
Leave site B out +0.89 −0.09–+1.88
Leave site C out +1.24 +0.23–+2.25
Leave site D out +0.66 −0.30–+1.62

Flexible form adds baseline², need², and baseline × need. Each site check refits the prespecified adjustment after omitting one synthetic site.

Swipe or scroll to inspect every specification

Forest plot of seven synthetic program-evaluation specifications. The unadjusted estimate is plus 5.55; adjusted estimates are near plus 1, and two leave-one-site-out intervals cross zero.
Code-rendered synthetic example from a deterministic repository script. Not client data, a published study, or evidence of real-world program impact. RNG_SEED 20260818. Open the full-size chart in a new tab

Scope

A complete evidence path, not an isolated calculation.

The engagement follows the result from question to source data to code to conclusion. That makes the work easier to review, revise, and reuse after the immediate deadline.

Typical workstreams

  • Research-question framing, analysis plans, and data-readiness review.
  • Power or sample-size analysis where the design and assumptions support it.
  • Statistical modeling, forecasting, time-series, panel, causal, or ML analysis.
  • Diagnostics, robustness checks, sensitivity analysis, and error analysis.
  • Reproducible Python, R, or SQL workflows with documented data lineage.

Deliverables

  • A written analysis plan and explicit assumption register.
  • Clean, reproducible code or notebooks and an execution guide.
  • Review-ready figures, tables, diagnostics, and sensitivity results.
  • A methods and findings memo separating evidence, interpretation, and limitations.
  • A technical handoff for the researchers who will maintain or extend the work.

Professional boundary. Statistical, quantitative research, model review, and engineering consulting — not clinical, regulatory, legal, research-ethics, or publication assurance. We do not guarantee statistical significance, acceptance, or a preferred result.

Independent review

Already have a model or analysis?

Use independent model validation when the main question is whether an existing methodology, implementation, or evidence package holds up under technical challenge.

See model validation

Next step

Start with a high-level brief.

Share the research question, the decision it informs, timing, and data status. Do not include non-public data or identifiable research records. We reply by email within 24 hours with a fit assessment and the next useful step.

Send a project brief