Skip to content
datanalyze Statistics & data science
Language FR EN
Get in touch
Services Training About Get in touch
Language FR EN

Service

Data analysis

You have collected data and you need a defensible answer from it. I take the analysis on end to end and deliver interpreted results, not raw software output.

Discuss your project
01

Who it is for

This service is for those who already have their data in hand and lack the time, the method or the confidence to analyse it themselves.

  • Doctors and clinical researchers with a patient database to exploit
  • PhD and postdoctoral researchers facing a methods chapter
  • Labs and research units with no statistician in house
  • Project managers who need to back a decision with evidence
  • SMEs sitting on data they have never used
  • Non-profits and public bodies with survey data
02

The problem it solves

A table of data answers nothing on its own. Between collection and a solid conclusion sits a series of choices — which test, which variables, what to do about missing data, which assumptions to check — and any one of them can invalidate the final result if it is settled badly.

The risk is not failing to get a number: software always produces one. The risk is getting a wrong number, or a correct one you cannot defend when a reviewer, a committee or a client asks why that method.

  • You have a p-value but you cannot justify the test you used
  • Your data visibly fails the conditions of the test applied
  • A reviewer rejected your analysis and you do not know where to restart
  • You redo the same calculations by hand at every update
  • Your results change depending on who produces them
03

What you get

An assignment ends with a set of deliverables you can use, review and reuse without me.

  • A written analysis report with the results and their interpretation
  • Tables and figures ready to publish or present
  • The written justification of the methods chosen
  • Commented R scripts that run in a single command
  • Assumption checks and sensitivity analyses
  • A debrief session so you can make the results your own
04

How it works

  1. Scoping

    A conversation to understand your question, your data and your deadline. Free and with no commitment.

  2. Proposal

    I write down what I propose to do, how, and by when. You approve before I start.

  3. Analysis

    I work on your data and come back to you whenever a methodological choice is yours to make.

  4. Handover

    Delivery of the report and scripts, then a session so you can defend every result.

05

Concrete examples

A few examples of the kind of questions handled. These are generic methodological illustrations, not real assignments.

Comparing two treatments

Does one treatment outperform another, at 6 and 16 weeks? Repeated measures on the same patients, so a model that accounts for the dependence between them.

Identifying risk factors

Which variables are associated with an event occurring? Logistic regression or a survival model, depending on whether time matters.

Analysing a survey

What do 800 questionnaire responses actually say? Descriptive analysis, association tests and grouping of profiles.

Testing a field hypothesis

Is the gap observed between two groups real, or attributable to chance? A test matched to the data type and the sample size.

Sizing a study

How many subjects are needed to detect a given effect? Power calculation up front, before collection.

Taking over an existing analysis

A reviewer disputes a method. Diagnosis of what is wrong, correction, and a rewritten statistics section.

Frequently asked questions

What format should I send my data in?
Excel, CSV, SPSS, Stata, SAS, Access or a database: all of these work. Do not spend time reorganising a file before sending it — cleaning and reshaping are part of the job, and an “already tidied” file sometimes hides useful information from me.
My data is confidential. How do you protect it?
I work with health data routinely. Data never leaves my workstation, is used for no other purpose, and is deleted at your request when the assignment ends. A non-disclosure agreement can be signed before anything is sent. For patient data, send it already pseudonymised where possible: I almost never need identities.
How long does an analysis take?
It depends mostly on the state of the data and the number of questions. A focused analysis on a clean dataset will take less time than a full study with cleaning, modelling and writing. I give you a firm estimate after scoping, before you commit.
Can I list you as a co-author on a publication?
That is possible and it happens regularly when the methodological contribution is substantial. It is discussed at scoping: some prefer a straightforward service, others a genuine scientific collaboration. Both work for me.
What if my data cannot answer my question?
I will tell you, and I will tell you early. It is a common situation — too few subjects, a missing variable, an unsuitable collection design. You would rather hear it at scoping than after several weeks of work, and I would rather say it than produce a result I do not believe in.

Other services

Support for researchers

You run the analysis, I secure the method

Learn more

Study design and sample size

Design it right before you collect

Learn more

Consulting and decision support

From dormant data to evidence-based decisions

Learn more

R development and automation

An analysis that runs itself, identically

Learn more

Training

Making your teams self-sufficient in R and data analysis

Learn more

A dataset to analyse, a method to validate?

Describe your need in a few lines, or book a first no-commitment call. I will tell you straight if I am the right person for it.

Describe my need Book a call