It's one of the questions I get most by email, almost always quietly: "Is it okay to have someone help me with the statistics of my thesis?" Behind it there's usually fear: of cheating, of the committee seeing it as a shortcut, or of a reviewer calling it fraud. If you're reading this with that doubt, let me reassure you straight away: getting statistical support for your thesis or paper is not only legitimate, it is a common and recognised practice in science. What separates legitimate help from fraud is not that you get help, but how it's done and how it's acknowledged.
The right question isn't "is it allowed?" but "how do I declare it?"
No researcher works alone. Papers are signed by several authors precisely because science is collaborative: someone designs, someone collects data, someone analyses and someone writes. Statistical analysis is one more specialty, just like academic English or figure design. Not mastering a multilevel structural equation model doesn't disqualify you as the author of your thesis, just as it doesn't disqualify you to ask for help with a lab technique that isn't yours. The question has never been whether you can lean on someone, but whether that collaboration is acknowledged transparently.
What the standards say: ICMJE and COPE
The two reference frameworks in research integrity are clear. The ICMJE (International Committee of Medical Journal Editors) defines four criteria for authorship: (1) substantial contribution to design, acquisition, analysis or interpretation; (2) drafting or critically revising the manuscript; (3) approval of the final version; and (4) accountability for the work. Someone who contributes but does not meet all four criteria, for example, someone who runs a specific analysis, is not an author: they are acknowledged in the acknowledgements.
The COPE (Committee on Publication Ethics) condemns ghostwriting, but it's worth understanding what that means: a ghost author is someone who would deserve authorship, because their contribution was conceptual and substantial, and is deliberately hidden. That is what's forbidden. A consultant who runs the analysis you designed and appears in the acknowledgements is the opposite: transparency. Acknowledging help is not hiding it.
In short: if someone advises on or runs the statistical analysis of a study you lead, they go in the acknowledgements. If their contribution is conceptual and substantial (they designed the analytical approach, interpreted it, wrote that section), then they deserve declared co-authorship. Either way, out in the open.
Where the line is: legitimate help vs. "thesis mills"
Here's the nuance that really matters. Legitimate help and fraud look very different when you put them side by side:
- Legitimate help: you design (or co-design) your study, you know your data and your question. An expert runs or supervises the analysis, explains it to you, and hands you the syntax so you know exactly what was done. You come out understanding your results and able to defend them.
- Fraud (the "thesis mills"): someone writes the whole thesis for you, hands you a result you don't understand, that you couldn't reproduce or explain, and sells it as a "turnkey thesis". That isn't support: it's impersonation, and it's exactly what committees and reviewers penalise.
Red flags that you're dealing with a mill and not a consultant: they offer to write your full thesis or paper, they charge by word or by chapter, they don't explain the method, they don't give you the syntax or the outputs, and they vanish after delivery. A good consultant does the opposite: they leave you understanding your own work.
How to acknowledge it correctly (with examples)
Declaring it is simple and takes the fear away. Some common wordings:
- In the thesis, in the acknowledgements: "I thank [name] for statistical advice on the data analysis."
- In a paper, in the Acknowledgements section or the Author Contribution Statement: "[Name] provided statistical analysis support." If the contribution was substantial, they are listed as a co-author with their role described.
My advice: ask your supervisor and, if you're publishing, check the journal's authorship policy (almost all defer to the ICMJE). On the projects I support, I help you draft that sentence so it meets exactly what your university or journal requires. There's nothing to hide, and declaring it well protects you.
And you're still the author (and you learn along the way)
This is the point that reassures PhD students most: with a good consultant you don't delegate your thesis, you reinforce your authorship. You come out knowing which analysis was run and why, with the reproducible syntax in hand, able to answer your committee with confidence when they ask why a confirmatory and not an exploratory factor analysis, or how you justify measurement invariance. That's the opposite of cheating: it's doing your work better and with judgment.
If after this you want to take the step, the key is choosing well who you hire. I cover it in detail in how to choose a statistician for your doctoral thesis, with the quality signals and the red flags.
Want statistical support done with judgment (and acknowledged properly)?
I'm a PhD in psychology. I run the analysis of your thesis or paper, explain it so you can defend it, hand you the reproducible syntax, and help you word the acknowledgement. No mills: you remain the author.
Get a free consultation →In short: hiring statistical help for your thesis or paper is ethical, legal and common, as long as it's acknowledged transparently and you understand and can defend your work. The cheating isn't asking for help; the cheating is hiding it or handing in something that isn't yours. Do it right and you're not doing anything wrong: you're doing science the way it's really done, as a team. If you like, let's talk about your case in a no-commitment consultation.