Independent data analysis, research verification and evidence-based writing. Built around the question, not the answer you hope to get.
Need a dataset analysed, a study checked before publication, or a story investigated using primary data? The Null Institute accepts paid research commissions. We define the scope, test the evidence and explain the result in language your audience can use.
Commissioned services
What can we help you find out?
We offer statistical analysis services and data analysis consulting alongside independent research reviews and commissioned articles. The work differs by client, but the standard is the same: evidence that can be examined.
For researchers & organisations
You have data, a result, or a decision that needs a sound quantitative basis.
- Statistical analysis: descriptive analysis, modelling, regression, uncertainty estimates and robustness checks where appropriate.
- Independent statistical review: examine assumptions, reproduce calculations and test alternative interpretations.
- Research verification: check published claims against available datasets, methods and original sources.
For journalists & editors
You need a reliable answer, not another confident quotation.
- Claim checking: trace an eye-catching statistic to its data and test what it really means.
- Focused mini-research: investigate a question using public datasets or source documents.
- Evidence-based articles: research and draft explainers, data-led features and technical briefs with traceable references.
Evidence, not testimonials
Three examples of how we work
These are the Institute’s independently published investigations, not previous paid-client projects. Each illustrates a type of work we can discuss commissioning.
16.41%
One measurement changed the category
In a survey-weighted NHANES analysis of 6,100 US adults, the first blood-pressure reading and the three-reading average placed 16.41% in different categories. That does not mean 16.41% were misdiagnosed.
Read the analysisHR 2.33
A strong association needed another look
An earlier mortality study reported an adjusted hazard ratio of 2.33. Our replication examined the original result and found that the association weakened substantially in a later mortality linkage.
See the replication42 / 48
Completed is not verified
Our audit of an experimental AI research workflow found 42 projects marked completed among 48 registered. Some outputs still had unresolved questions. We examined the difference between workflow status and verified evidence.
Read the auditResults and limitations are reported in their original context. Published examples do not establish that the same outcome will occur in a new project.
What you receive
A usable answer. And a record of how we reached it.
Not every question permits a definitive answer. When the data are weak, contradictory or incomplete, that becomes part of the result rather than an inconvenient footnote.
Founder and Director of The Null Institute. His work combines quantitative analysis, independent checks of published findings and research methods that can be audited. Background and selected publications.
- Defined deliverable: a report, verification memo, analysis package or article draft agreed in advance.
- Traceable work: source links, methods, key choices, calculations and figures as appropriate to the project.
- Interpretation: a plain-language summary of the findings, uncertainties and competing explanations.
- Reproducibility where possible: annotated data or code when access rights, tools and project scope allow it.
Working together
A clear scope before the work starts
Tell us the question
Share the objective, existing public sources or a description of available data, desired output and target date.
Agree the work
We assess feasibility and propose a scope, deliverables, project fee and schedule. No obligation to proceed.
Inspect the result
We deliver the agreed report, analysis or article, explain what survived checking and flag unresolved limitations.
Pricing: paid commissions are quoted individually. A tightly bounded check and a multi-source research project require different amounts of work. Where the full scope is uncertain, we may propose a smaller feasibility stage first.
Confidentiality and independence: data handling, attribution, publication rights and confidentiality are discussed before work begins. Payment never guarantees a favourable or statistically significant finding. We do not fabricate sources or provide undisclosed academic ghostwriting.
Commission a project
Request a project quote
Tell us what you need analysed, verified or written. Include any relevant public sources and your preferred deadline. We’ll assess feasibility and propose a scope and fee.
Compose a project enquiryContact address: Kappelikuja 6, 02200 Espoo, Finland