If you plan a medical doctorate, you soon meet terms such as experimental, clinical, statistical or retrospective. They describe where the data come from, and they decide which approvals you need and how much time you should plan. This guide explains the terms, uses the University of Marburg as an example of what faculties require, and offers a checklist for preparation.
Short answer: Experimental means that you generate the data yourself. Retrospective means that you analyse existing data, such as patient records or registers. The Faculty of Medicine at Marburg lists experimental, statistical and clinical doctorates on its page. According to the faculty ethics committee, retrospective studies still need an ethics vote. Your faculty decides whether a project is admitted. Last updated: October 2026.
The most important points at a glance
- Experimental means that you generate the data yourself, in a laboratory or in a study.
- Retrospective means that you analyse existing data, for example findings or registers.
- The Marburg faculty names experimental, statistical and clinical doctorates as options.
- According to the faculty ethics committee, a retrospective study also needs an ethics vote.
- The application for acceptance must be submitted before the doctoral work begins.
- The type of work determines effort, data protection and the timetable.
What do the terms mean?
The distinction between experimental and retrospective describes when the data are created. In an experimental project, you plan and carry out the investigation yourself, for example in a laboratory or in a series of trials. In a retrospective project, you draw on data that already exist, such as documented treatment histories, findings or entries in a register.
Clinical projects are usually tied to patients. Statistical projects analyse data using statistical methods. In practice the terms overlap, so the classification in the supervision meeting and in the doctoral regulations matters more than a dictionary definition.
What does the Marburg faculty require?
The page “Fragen über Fragen?” of the Faculty of Medicine at the University of Marburg names experimental, statistical and clinical doctorates as options. It does not define the types further. Retrospective work is mentioned only in connection with the ethics question. The page gives no date, so check the current status yourself before you register.
According to the page, for biomedical research on humans and for epidemiological projects, you must obtain advice from an ethics committee, unless the project has already been reviewed. The faculty ethics committee has also decided that a retrospective study needs an ethics vote as well. Often an application already exists if you work on a part of a larger research project.
Comparison of the two types
The table compares the two types that matter most for most doctoral candidates in medicine. The information is based on the Marburg page and the Tübingen doctoral regulations.
| Point | Experimental | Retrospective |
|---|---|---|
| Data source | You collect the data yourself | Existing data, such as records or registers |
| Ethics vote | Required before the start for research on humans | Also required according to the Marburg ethics committee |
| Time needed | Depends on the set-up of the experiment | Depends on the data situation and access |
| Typical hurdle | Set-up, participants, laboratory time | Data access, anonymisation, data protection |
The Tübingen doctoral regulations require advice from the ethics committee before the start for research on humans, and proof of it at admission. Your supervisor should clarify whether your project falls under this rule. The regulations are dated 9 December 2021 and in force since 1 February 2022.
Which steps belong to preparation?
- Agree with your supervisor which type your work is, and record it in the application.
- Check whether your project is research on humans, and submit the documents for the ethics consultation.
- Submit the application for acceptance as a doctoral candidate before the work begins.
- For retrospective data, clarify access to records or registers and anonymisation before you analyse anything.
- Document every data source so you can later explain where each value comes from.
Frequently asked questions
Is a retrospective project a full doctorate?
The faculty decides this under its doctoral regulations. The Marburg page does not treat retrospective work as a separate category but mentions it in the context of ethics. Ask your supervisor and the doctoral office before you start.
Do I really need an ethics vote for a retrospective study?
According to the Marburg faculty ethics committee, yes. Other faculties may decide differently, so clarify the rule for your university even if you work with data that already exist.
Can I change the type of my project later?
A change affects the application and should be agreed with your supervisor and the doctoral office. Note changes early and in writing, so they do not surface only at submission.
Do I have to fix the type before acceptance?
According to the Marburg page, the application for acceptance must be submitted before the doctoral work begins. A clear statement of the type in the application makes the ethics consultation easier. The page states that the application can be sent by post and that you do not need to appear in person.
When does each type make sense?
The choice follows the research question first. If you want to measure an effect, an experimental or clinical design often fits. If you examine connections in existing data, a retrospective project is a natural choice. The research question should decide, not convenience.
Data access and data protection in retrospective work
In retrospective projects, data access is often the real bottleneck. Patient data are subject to medical confidentiality and data protection law. Before you analyse records or register data, it must be clear who grants access, in what form the data are provided, and whether they are pseudonymised or anonymised.
Also plan for the ethics committee to ask about the data source, scope and analysis. A clear description of which variables you need and why speeds up the vote. Not every database suits a dissertation, so a feasibility check before the application is worthwhile.
What changes for the timetable?
Experimental work needs time for set-up, repetitions and analysis. Retrospective work often starts analysis sooner, but it can fail on missing or incomplete data. Estimate both variants realistically and build in a buffer for data cleaning.
Assumption: you have 500 findings from a register. You plan 40 hours for cleaning and 60 hours for statistics. That makes 100 hours of pure work before you write up the results. These hours do not appear in any application, but they belong in your work plan.
Related guides
- PhD in Germany: routes, titles and requirements
- PhD Funding: Position, Scholarship or Own Savings
- Health Insurance During a PhD in Germany
- Doctoral Title Abbreviations: From Dr. rer. nat. to Dr. h. c. Explained
Sources:
- University of Marburg, Faculty of Medicine, page “Fragen über Fragen?” (no date on the page, in German)
- Doctoral regulations of the Faculty of Medicine, Tübingen, for Dr. med. and Dr. med. dent., 9 Dec 2021 (PDF, in German)
Last updated: October 2026. This guide is general information, not individual advice.
