Biostatistics consulting for health research: dissertations, clinical studies, scale validation and reviewer responses. A PhD statistician and a fixed quote in 24 h.
You hand me the analysis and I hand it back done: run, reported in APA 7 or Vancouver, and defensible in front of a reviewer or a committee. Clinical and observational studies, nursing, physiotherapy, dentistry, public health and psychology. A fixed per-project quote, never an hourly rate, and I reply myself in under 24 working hours.
from €300 + tax. The closed analysis plan. Execution from €500, response to reviewers from €800, the full statistical section of a paper from €1,500.
Nobody arrives out of methodological curiosity. People arrive stuck, and almost always in one of these four places.
Collection went well, the sample is there, and now you have to choose between a mediation model, a multilevel one and a mixed one. Each option changes what you can conclude, and choosing badly does not show until a reviewer sees it.
An adjusted Cox model, measurement invariance across groups, bootstrapped indirect effects, a fit index you did not know existed. You are one step from the accept, with a deadline on top.
EFA, CFA, invariance, McDonald’s omega and some item analysis. Your supervisor wants Mplus and you have taken three SPSS courses.
The numbers come out, but the methodological justification is not written down, and the question that sinks a defense is never "what is the beta?".
This is the decision that shapes what working with me feels like, so it goes here and not in the small print. On an hourly rate, every model that fails to converge and every assumption that breaks is an invoice for you; on a fixed quote, the risk that I am slow is mine.
Five project types cover almost everything that comes in. Each has its own page with the detail; here is the summary and the link.
From the analysis plan to the Results chapter and the statistical questions at your defense or viva. Dissertation statistics help
Methods, analysis, results, tables and figures for the manuscript you are about to submit. The Q1 journals service
EFA, CFA, invariance across groups, reliability with omega, content validity and item analysis. Questionnaire validation
The point-by-point document, the reanalyses they ask for and the revised text. Reviewer response service
An audit of an analysis you have already done, with the problems ordered by severity. The statistical review
Methods repeat from one journal to the next, but every field has its typical reviewer and the test it never forgives. These are the ones that come in most.
Survival analysis with Kaplan-Meier and Cox regression, with the proportional-hazards check; logistic regression with adjusted odds ratios; ROC curves with the AUC and its confidence interval; and the STROBE or CONSORT reporting the journal requires. Statistician for medical research
Scales: adaptation and validation with EFA, CFA and omega; content validity from an expert panel (Aiken’s V and the content validity index); and models for patients nested in units or centers. Statistician for nursing research
Repeated-measures trials analyzed with mixed models rather than an ANOVA on change scores; ANCOVA with the baseline; effect sizes with their interval; and the minimal clinically important difference (MCID) to say whether the change matters to the patient. Statistics for physiotherapy research · Statistics for sport science
Inter-examiner reliability with kappa (weighted when the scale is ordinal) and ICC with its model declared; and models for teeth or surfaces nested in patients, the step most often forgotten. Statistics for dental research
Full psychometrics (EFA, CFA, invariance, IRT), mediation and moderation with bootstrapping, multilevel and structural equation models, and meta-analysis. Statistician for psychology research
Statistical consulting is a methodological and analytical partnership between a researcher and a specialist. It is not ghostwriting and it is not a copy-paste of SPSS output. It is a joint effort in which the consultant contributes judgment to design the study well, choose the right tests, run them rigorously and report the results in a way that holds up in front of a dissertation committee or a reviewer.
The difference between professional consulting and a low-cost service shows up at three moments: when a reviewer demands a sample-size justification, when an editor asks for an analysis no one had planned, and when the defense turns into a serious methodological discussion.
A statistical consultant turns a research question into a defensible analysis plan and runs it end to end: chooses the test or model that fits the design (not the other way round), checks assumptions and power, runs the analysis in whatever software fits, interprets the results in substantive language and reports them in APA 7 or Vancouver.
What sets a good consultant apart is not knowing many tests but knowing which one not to use and why: when a repeated-measures ANOVA falls short of a mixed model, when a correlation hides a moderation, or when the sample size will not support the model the supervisor wants, and being able to defend those decisions in front of a committee or a reviewer.
The main profile is the health-sciences or psychology researcher who needs solid analyses but wants to spend their time on what they do best: the research question, the interpretation and the writing.
Most requests for biostatistics consulting and biostatistics services from the United States fall into four groups, and each one gets a different deliverable.
Starting points in US dollars: analysis plan ≈ $340 (€300), analysis executed ≈ $570 (€500), response to reviewers ≈ $920 (€800), statistical section of a paper ≈ $1,700 (€1,500). Quoted and invoiced in euros.
MSc, MRes and MA dissertations are assessed as your own work, so for them I run and explain the analysis, give you the SPSS or R output tables and syntax, and review the draft you write. PhD theses get the full analysis and a written Results chapter to review with your supervisor. The detail, the timing for the dissertation season and the prices in pounds are on the dissertation statistics page.
Dissertation and thesis statistics help
Every project is closed with a written scope, but these are the usual blocks:
A full dissertation analysis takes 4 to 8 weeks from the moment I have clean data; a response to reviewers, 1 to 3. If your deposit or journal deadline falls inside that window, write to me and I will tell you whether I can make it; if I cannot, I will tell you that too.
The two ends of that range are not a negotiating trick: they depend on how the data arrive. A clean file with a finished codebook sits at the low end; a three-wave database with half-written labels and undocumented missing values eats the first weeks just becoming usable. That is why the clock starts at clean data, not at signature.
I work per project, not by the hour: you know what you will pay from the start, with no surprise extra hours. What moves the figure above the starting point is the number and complexity of the analyses, whether an instrument has to be validated from scratch, the urgency, and whether you also need the Methods and Results written. A mediation with n = 200 and clean data is not the same job as a longitudinal multilevel model with missing data.
The starting points are the same everywhere; only the currency you see them in changes.
Everything is online, in English or Spanish, and the base is Spain: the GDPR covers your data from the first email, quotes and invoices are in euros, and there is no need to align time zones just to find out what something costs.
Your data is handled pseudonymized, over a secure channel, used for nothing but your project and deleted when it ends. I sign a confidentiality agreement before you send anything if you ask for one, and universities, hospitals and research groups in Canada, Germany, the Netherlands, Switzerland or the Nordic countries can attach the data processing agreement (GDPR article 28) their legal office asks for.
Read the data processing agreement (GDPR art. 28)
If your project is small, you might not need consulting at all. These calculators are open and used in universities:
These are not rate cards: they are the starting points for each kind of job, published so you do not have to write in just to find out whether you are out of budget. The exact figure arrives fixed, within 24 working hours, after I have seen your case.
Quotes and invoices are in euros. The pounds, US dollars and Australian dollars are an approximate conversion at a recent exchange rate, so you can tell at a glance whether it fits.
If a first-round reviewer knocks down the statistical section, I redo the analysis and write the response at no extra cost. I do not guarantee your paper gets accepted: I guarantee you are not left on your own.
And if all you want is a critical read before submitting, with nothing executed, there are two short steps: the Method review at €49 + tax and the full human review of the manuscript at €149 + tax.
This part usually decides more than the previous one. If something you need is on this list, better to know now than after two emails.
Not everyone needs me to run the analysis. If what you want is to know what a reviewer will object to before you submit, there is a tool on this site that does exactly that on its own, in minutes.
They are two different things and it is worth not mixing them: that is software and it gives you a read of the manuscript; this is human work that runs the analysis, answers in writing and signs what it delivers. Plenty of people start with the tool and only write in when the report surfaces something they cannot solve alone.
Try Previum · AI peer review, free
It depends on scope, which is why the exact figure comes after I see your design and your data, not before. As a reference: a closed analysis plan starts at €300, running the analysis and delivering it in APA 7 tables at €500, a response to reviewers at €800, and the complete statistical section of a paper at €1,500, before tax. Those are starting points. The real quote reaches you fixed, in writing, in under 24 working hours.
Because charging by the hour hands you the risk that I am slow. If a model refuses to converge and three parameterizations have to be tried, on an hourly rate you pay for that afternoon; on a fixed quote I do. You know the number before we start and I have the right incentive, which is getting it right the first time.
A methodological and analytical partnership between a researcher and a specialist, from study design to the response to reviewers. It is not ghostwriting: the contribution is acknowledged in the Methods, the acknowledgments or the author contribution statement, as the journal or the university asks.
No: the health sciences and psychology. That is where I hold my PhD, where I publish and where I know what a reviewer will object to. If you write to me from engineering, economics or political science I will tell you on day one rather than take the job and learn on your data.
R (lavaan, psych, lme4, metafor, survival), SPSS, JASP, jamovi and Mplus. Psychometrics (EFA, CFA, IRT, invariance), regression of every kind, survival analysis, ROC curves, mixed and multilevel models, SEM, mediation and meta-analysis. You always get the commented syntax, without having to ask: that is what lets you reproduce the result, answer a reviewer who asks for a reanalysis, and carry on without me if you ever want to.
No. For a doctoral dissertation I run the analysis and deliver the Results chapter written in APA 7, which you review with your supervisor and make your own; the introduction, the literature review and the discussion are yours. For graded master’s or undergraduate work (an MSc dissertation, an Honours thesis, a capstone) I run and explain the analysis, give you the tables and syntax, and review the draft you write. If your university’s rules do not allow outside help with the write-up, I stick to running and explaining the analysis.
The ICMJE recommendations do not say an external statistician can never sign: they say that whoever meets the four authorship criteria is an author, and whoever falls short of all four goes in the acknowledgments, with written permission from the person named. On most projects the contribution is declared as methodological support in the acknowledgments or the author contributions; if it did meet all four criteria, the honest thing is to discuss co-authorship with your supervisor before the writing starts.
Yes. I am based in Spain, so the GDPR applies to everything I do. Data travels pseudonymized over a secure channel, is used for nothing but your project and is deleted from my systems when the project ends. I sign a confidentiality agreement before you send anything if you ask, and the data processing agreement your institution needs: there is a ready-to-sign one on this site, or I sign yours.
Rarely. The work is delivered in writing, questions are answered by email within one working day, and when a call is genuinely faster it is scheduled in your time zone. Researchers in North America usually get my reply at the start of their working day.
If you have one and they have time, use them: they are free and they know your context better than I do. People write to me when one of three things happens. There is nobody, because many departments have no methodologist. There is someone, but they are busy, and your analysis joins a two-month queue that does not fit the journal calendar. Or the specific method you are asked for — invariance, multilevel, IRT, bootstrapped indirect effects — is not their area, which is different from not knowing statistics. What I bring is not knowing more about your construct: it is that this is what I do all day, that there is a delivery date in writing, and that if the reviewer knocks it down I redo it at no charge.
The deadline goes with the scope and is fixed in writing in the quote, next to the price: it does not move afterwards. As an order of magnitude, a response to reviewers usually takes one or two weeks, a full dissertation analysis four to eight weeks from clean data, and a questionnaire validation with three studies was delivered in two (Psicothema, 2026). If you have a journal or deposit date, tell me in the form: it is what decides whether I can take it, and I would rather tell you I cannot make it than say yes and deliver late.
A short description of the design, what your supervisor or the journal is asking for and your deadline. The data comes later, pseudonymized (no names or direct identifiers), and I explain how to prepare it. If you have not collected data yet, even better: the design stage is the cheapest moment to fix a problem.
This page covers what statistical consulting is, how to hire it and what it costs. If you are after the technical detail of which analyses get done in psychology — psychometrics, SEM, mediation, multilevel — that lives on the statistician for psychology research page. And every service, one by one, is on the services page.
The full breakdown, on the pricing page