Good Scientific Practice

The principles of good scientific practice require that all scientists work "lege artis" (i.e. according to the rules of art).

This includes, among other things:

•     the correct handling of data,
•     critical questioning of all results
•     to avoid and prevent scientific misconduct,
•     maintaining strict honesty with regard to one's own and third parties' contributions.

Possible violations of the rules of good scientific practice are manifold. For example, scientific misconduct occurs when deliberately or negligently

•     Data or sources are invented or falsified,
•     you compromise or interfere with the research activities of others,
•     violate intellectual property (e.g., through plagiarism or theft of ideas),
•     falsely accuse third parties of scientific misconduct,
•     know about scientific falsification or misconduct and do nothing about it,
•     neglecting the duty of supervision.

A detailed list is available in Section 9 of the Bylaws, titled "Ensuring Good Scientific Practice, Preventing Scientific Misconduct, and Handling Violations."

• TUD Dresden University of Technology has adopted the Bylaws on Ensuring Good Scientific Practice, Preventing Scientific Misconduct, and Handling Violations.
• All members and affiliates of TUD are required to comply with these bylaws, to make them the foundation of their academic work, and to actively contribute to the prevention of academic misconduct within their sphere of influence.
• Scientific misconduct is not tolerated at TUD. Suspected cases are investigated carefully and with respect for those involved. If the suspicion is confirmed, appropriate measures will be taken.

The ombudsperson is the contact person, advisor, and mediator in all cases of suspected scientific misconduct. If necessary, he or she will be supported by the Review Board for Scientific Misconduct.

Your report will be treated confidentially. Confidentiality serves to protect the whistleblower as well as the person suspected of misconduct.

The ombudsperson shall maintain a regular exchange with the liaison officers of the faculties, the Review Board for Scientific Misconduct, as well as the other advisory bodies of TUD. Conflict cases that are not related to scientific misconduct can be forwarded confidentially to the responsible offices of TUD (e.g., Personnel Representation Council, conflict mediator of the Graduate Academy, psychosocial counselling, etc.) with the consent of the informant. 

If there is reasonable suspicion of academic misconduct in the view of the ombudsperson, this suspicion will be reviewed by the Investigative Committee or, in suspected cases where the misconduct relates to academic examinations (e.g. Bachelor's, Master's, Diploma examinations) or graduations (doctorates, habilitations), by the regular Examination Board provided for in the respective examination and graduation regulations.

Contact

Prof. Dr. Alexander Kemnitz | Ombudsperson

Each faculty appoints one female and one male scientist as a liaison officer for early-career researchers. 

List of contact details of the faculty liaison officers (download) (in German only)

These liaison officers serve as contact persons that can be easily approached by early-career researchers (especially doctoral candidates). These contact persons can also mediate in problematic situations. If necessary, and only with the consent of the person seeking advice, the contact persons can pass the conflict case on to the ombudsperson.

This shall not affect the right of direct recourse to the ombudsperson.

Furthermore, certified contact persons are available in all five Schools, as well as at the CFAED, the CRTD, and the Graduate School DIGS-ILS. They offer subject-specific GSP workshops for master's students and doctoral candidates.

List of contact details of all GSP contact persons and trainers (in German only)

In cases of suspected scientific misconduct, the Office for Good Scientific Practice  supports the ombudsperson, the Investigative Committee, and the regular review boards.

The Office for Good Scientific Practice accepts reports of suspected misconduct confidentially and provides information on possible procedural steps. This shall not affect the right of direct recourse to the ombudsperson or the Investigative Committee.

In cases of suspected scientific misconduct, the Office for Good Scientific Practice offers to check final theses (e.g. Bachelor's, Master's, Diploma examinations) or graduations (doctorates, habilitations) for plagiarism using plagiarism detection software.

Contact

Sigrid Flade

The e-learning “Export Control in Science” (modules 1 to 3) is available in OPAL for all employees of TU Dresden:

German: https://bildungsportal.sachsen.de/opal/auth/RepositoryEntry/45092896771 English: https://bildungsportal.sachsen.de/opal/auth/RepositoryEntry/45448036355

In its bylaws on ensuring good scientific practice, Section 3 (3), TUD has stipulated the following: “All academic staff at TU Dresden, as well as all doctoral students, are required to complete at least one online training course or in-person session on the topic of ‘Good Scientific Practice.’”

The Center for Continuing Education (ZfW) regularly offers workshops on “The Rules of Good Scientific Practice” for TUD professors.

Dates and registration

Workshops on Good Scientific Practice for Professors
 

In its Living Guidelines on the Responsible Use of Generative AI in Research (03/26), the European Commission recommends that researchers adhere to the following principles:

Responsibility for Research Integrity
Researchers remain fully accountable for the integrity and accuracy of all research outputs, including those generated with AI tools. They are aware of the limitations of generative AI, such as bias, hallucinations, and inaccuracies.

Transparency in AI Use
Researchers ensure transparency by clearly disclosing when and how generative AI tools are used in the research process. They explain the limitations of the tools employed.

Data Protection and Intellectual Property
Researchers pay particular attention to data protection, confidentiality, and intellectual property rights, especially when sharing sensitive or protected information with AI tools. They handle any personal data output in compliance with applicable laws.

Legal Compliance
Researchers comply with all relevant national, EU, and international legal requirements. Outputs generated by generative AI may be especially sensitive with respect to intellectual property rights and personal data protection.

Continuous Learning
Researchers continually update their understanding of the capabilities, limitations, and best practices for using generative AI tools, including their environmental impact and the evolving landscape of responsible AI use.

Restrictions on Sensitive Activities
Researchers refrain from using generative AI tools in sensitive activities that may affect other researchers or organizations, such as peer review or the evaluation of research proposals.

Prohibition of Forgery, Falsification, Plagiarism, and Ethical Prompting
Researchers do not use generative AI to fabricate, falsify, or plagiarize content. They apply ethical prompting and do not manipulate or alter original research data using AI tools.

Awareness of Hidden Prompts
Researchers are aware of the risks associated with "hidden prompts"—instructions for AI systems that are not visible to human users—and are taking steps to mitigate them.

Guidelines

Artificial Intelligence (AI) is evolving rapidly and is also a relevant technology for science and research. The DFG website brings together information on all areas of activity in which AI plays a role: grant applications, peer review, and specific funding opportunities for AI.

• DFG AI Topic Page
• DFG Statement on AI in Research (2023)
• DFG Guideline on the Use of AI in Peer Review ( 2026)
• ERA Living Guidelines 2026 (Responsible Use of Generative AI in Research)

• FAQs on "Artificial Intelligence and GWP" 
  Ombuds Committee for Scientific Integrity in Germany (OWID)

TUD conducts comprehensive research into AI applications and actively incorporates the topic into its teaching. In addition, various centers of expertise advise faculty and students on leveraging AI's potential for their work.

In March 2023, the TUD University Executive Board acknowledged the potential of AI:
"TUD recognizes the potential, but also the challenges, posed by AI-powered text-generation programs such as ChatGPT. It encourages discussion on the classification and use of such programs, particularly in an educational context, but also about legal framework conditions."
~ Rectorate of TUD, March 28, 2023

• Artificial Intelligence (AI) at TU Dresden
• Use of AI in Teaching (Intranet)

The TU Dresden follows the DFG’s statement on the responsible use of AI tools in research (2023), which also serves as a guideline for funding under DFG programs.

Use of AI in Doctoral Theses

• When writing a doctoral dissertation, you are required to disclose any aids you have used. The use of AI tools is considered an aid and must therefore be disclosed accordingly.
• Please review the rules regarding the disclosure of aids in your faculty’s doctoral regulations.
• If in doubt, contact the relevant doctoral committee and disclose any tools used for the sake of transparency.
• Further information on the use of AI in theses can be found on the TU Dresden website.

Information and Continuing Education

AI Campus
The AI Campus is the learning platform of the Stifterverband for Artificial Intelligence—featuring free online courses, videos, podcasts, and tools.

HRK AI-LOTSE
AI-LOTSE is the hub for guidance, technology, services, and expertise in artificial intelligence at universities.

Literature Search

• SciSpace
  All-in-one tool for literature research, PDF analysis, AI summaries, and citation suggestions
• NotebookLM (Google)
  An AI-powered note-taking tool that analyzes your own documents, PDFs, websites, and videos
• Semantic Scholar
  Free, AI-based search engine for scientific literature with semantic search

• AI Usage Cards

  Transparent AI Documentation (Universität Göttingen)

• Note: Even though AI tools provide summaries and citations, it remains your academic responsibility to verify the sources. Never upload unpublished manuscripts or confidential data to public AI systems.

Academic Writing and Grant Proposals

AI-powered writing tools assist with style, grammar, argument structure, and compliance with journal guidelines—particularly helpful for non-native speakers.

• Thesify
  AI writing assistant for academic texts and grant proposals; offers feedback, a journal finder, and citation suggestions.
• Paperpal
  AI-powered writing and proofreading tool for manuscripts, grant proposals, plagiarism checks, and journal compliance checks
• SciSpace
  Also suitable for manuscript preparation and writing support

• Note: AI can never be listed as a (co-)author. The use of generative AI in writing must be transparently disclosed in the Methods section or in the acknowledgments. Responsibility for content, originality, and integrity always remains with you.

Visualize ideas and identify research gaps.

AI helps map research fields, identify gaps, and generate new hypotheses, a crucial advantage when developing your own research questions.

• Research Rabbit
  Visualizes citation networks, author relationships, and topic clusters; identifies research gaps
• Allen Institute for AI
  Open-source tools for hypothesis generation and literature analysis, among other uses

• Note: Use AI-driven ideas and visualizations as inspiration, but always critically evaluate the relevance and quality of the suggestions. The responsibility for selection and interpretation remains with you.

Coding and Statistics

AI-based coding and statistics tools are indispensable for data-driven research. They accelerate analysis processes, assist with debugging, and make open-source code more accessible.

• Cursor
  AI code editor with agents for code generation, debugging, and multi-file edits
• GitHub Copilot
  An AI pair-programming tool for code completion, debugging, and code reviews
• CatalyzeX
  A platform for searching for and using open-source code in scientific publications

• Note: If AI has written significant portions of your analysis code, this must be clearly disclosed. You bear full responsibility for methodological correctness and traceability.

• Isabella Buck (2025): Wissenschaftliches Schreiben mit KI
• Zdravko Tretinjak (2026): KI‑Tools für Studium und Forschung : Prompts für effizientes wissenschaftliches Arbeiten

During the planning phase, the research questions and methodology are developed. To ensure transparency and avoid conflicts, the division of labour within groups and the order of authorship for publications should be set out in writing as early as possible. 

A wide variety of data is generated throughout the research process. This research data can originate from both qualitative and quantitative research projects, e.g., through interviews, surveys, experiments, or observations.

This research data may be in analog or digital form, including text documents, tables, logbooks, questionnaires, audio and video recordings, samples, collections, database contents, or protocols.

All sources used must be documented in full and in a traceable manner in accordance with the FAIR principles (Findable, Accessible, Interoperable, Reusable). Manipulation or fabrication of data (data falsification and fabrication) constitutes a serious violation of the GWP and may result in criminal consequences.

Guidelines for Handling Research Data

The long-term archiving of research data is a prerequisite for the traceability of scientific results.

To establish general guidelines for the handling of research data - that is, beyond primary data - TU Dresden has adopted guidelines for the handling of research data.

Research Data | Support & Advice

• The Service Center Research Data supports all researchers at TU Dresden and at DRESDEN-concept institutions with individual challenges related to research data:
  • Data Management Plan for Grant Applications
  • Advice and Information on Research Data Management
  • Support with DMP implementation
• The ZIH offers a range of technical services for research data management. Further information and guidance on using these technical services as research data management tools are available from the Service Center Research Data.

All individuals who have made a genuine, substantial contribution to the design, data collection, analysis, interpretation, or preparation of the manuscript, and who have consented to publication, must be listed as authors.

Such a contribution exists, in particular, when a researcher has made a scientifically significant contribution to

• the development and design of the research project or
• the development, collection, acquisition, or provision of data, software, or sources, or
• the analysis/evaluation or interpretation of the data, sources, and the resulting conclusions, or
• the development of scientific findings, or
• the drafting of the manuscript.

If a contribution is insufficient to justify authorship, this support should be appropriately acknowledged in footnotes, the preface, or the acknowledgments

Honorary authorships are not permitted. Generative AI systems cannot assume authorship; responsibility for content and scientific integrity always remains with the participating researchers.

Determining Authorship and Author Order
Ombuds Committee for Academic Integrity in Germany

AI Tools in the Writing Process

When using generative AI (e.g. for research, draft texts or translations), the purpose, scope and models used must be disclosed transparently, for example in the methods section or in the appendix.

Responsibility for the content and adherence to scientific integrity remain entirely with the researchers.

The rules governing the use of AI are not always consistent. For example, different academic journals have different policies on the use of AI. The Ombudsman’s Committee for Academic Integrity in Germany (OWID) provides an overview of this.

• AI in the Writing Process | TUD

Citation

Any ideas or text passages borrowed from other sources must be accompanied by precise citations. This applies equally to research data and software, as well as to your own previously published texts (to avoid self-plagiarism).

Saxon University and State Library (SLUB)

• SLUB: Writing and Citing
• SLUB Citation Guidelines

TU Dresden Publication Guidelines

Publications and citations are key indicators of academic success. Correct and complete attribution of publications to their authors and to TU Dresden is essential, as missing or incorrect information regarding institutional affiliation can result in publications not being recognized as achievements of the university or its researchers.

TU Dresden Publication Guidelines

Research Information System (FIS)

The Research Information System (FIS) serves as the single point of information for all key metrics regarding research achievements and activities of the academic staff and associate members of TU Dresden. To increase the visibility of TU Dresden’s research profile, publications are recorded in relevant databases and consolidated in the university’s FIS.

Research Information System (FIS)

"Good Scientific Practice" (GWP) and professional research data management (RDM) are closely intertwined. Both principles require that research data be collected, documented, stored, and archived in a manner that is transparent, traceable, and tamper-proof throughout its entire lifecycle.

Ideally, before publication, the text should undergo an internal review or an external peer review process. Furthermore, the underlying research data – including unpublished data – must be archived for at least ten years in accordance with the FAIR principles (‘Findable, Accessible, Interoperable, Re-Usable’).

• Service Center Research Data | TUD
  The Research Data Contact Point supports every researcher at TU Dresden and at DRESDEN-concept institutions with individual challenges related to research data:
  • Data Management Plan for Grant Applications
  • Advice and information on RDM
  • Support with DMP implementation

• Research Data Management Services | TUD
  The ZIH offers a range of technical services for research data management. For more information and guidance on using these technical services as research data management tools, please contact the Service Center Research Data.

• Research Information System (FIS)

• Publication Guideline of TU Dresden

Related Links and Resources 

• forschungsdaten.info

  With practical, topic-specific, and discipline-focused articles, the site introduces readers to research data management and provides concrete, practical tips.
  
   

Important Resources and Guidelines

Deutsche Forschungsgemeinschaft (DFG)

Online Portal "Research Integrity"
Guidelines for Safeguarding Good Research Practice. Code of Conduct

Ombuds Committee for Research Integrity in Germany (OWID)

Ombuds Committee
International Codes of Conduct and Research Integrity Reports
FAQs on "Artificial Intelligence and Research Integrity"

Federal Ministry of Research, Technology, and Space (BMFTR)

Urheberrecht in der Wissenschaft. Ein Überblick (in German only)
Copyright in Academic Work - An overview

Contacts and Advisory Services

Network against Abuse of Power in Science (MaWi)

MaWi Network

Additional Resources & Links

Game-based learning | Gamification on academic integrity

Dilemma Game| Erasmus University Rotterdam
Seneca's Integrity Matters |  University of Waterloo
Jeopardy on Academic Integrity | Jeopardy Labs

Goethe University Frankfurt developed this one-hour online course for individual work, especially for doctoral candidates. It offers a brief introduction to the standards and principles of good scientific practice.

How does it work?

You're a doctoral candidate at TU Dresden, and you would like to complete the e-learning course. Please register via Opal!

Module 1 | Registrierung via Opal

Important

• Please make sure that your browser allows pop-up windows.
• If you have already completed the course on an earlier date, just upload your certificate to Promovendus.
• Please note that this tool was developed by Goethe University Frankfurt; therefore, Goethe University's data protection regulations apply to course participation.

The e-learning course "Good Scientific Practice at TU Dresden," developed by the Graduate Academy of TU Dresden, teaches doctoral candidates the specific guidelines, procedures, and support services related to good scientific practice. Participants also learn about ethical standards, documentation requirements, and institutional support structures.

How does it work?

• Have you completed the first e-learning module successfully?
• Then, please register for the second module of our certificate course via OPAL.
  You will need a valid ZIH log-in.

Module 2 | Registration via Opal

Important note

• Please make sure that your browser allows pop-ups.
• After successful completion, please download your certificate of participation and upload it to Promovendus.

This 2-hour interactive online workshop for doctoral candidates addresses case studies on good scientific practice. Participants analyze hypothetical scenarios from everyday research, discuss the dilemmas presented in small groups, and work together to develop possible approaches. The interactive course promotes exchange and strengthens competence in dealing with ethical challenges in research.

Registration for Module 3 | Prerequisites

• Have you been admitted as a doctoral candidate at your faculty by the doctoral committee?
• Have you successfully completed Module 1 and 2 of the certificate course and generated both certificates?
• Have you signed up for a session for Module 3 in Promovendus? Please note that there are also waiting list spots; this does not mean you are automatically registered to participate. Free places are allocated to people on the waiting list. We will contact you should a spot become available!
• Have you uploaded both certificates of participation in Promovendus?

Upload Certificates | How does it work?

• After registering for module 3, click on the "My Profile" tab in Promovendus and navigate to your course participant record (abbreviated with "V").
• In the left-hand menu under "Documents," click on "Add document" to upload the proof of successful participation in modules 1 and 2 as PDF files.

Module 3 | Dates

24.08.2026 | 9:00 – 11:00 am | Online (English) | Link for Registration                   09.09.2026 | 9:00 – 11:00 Uhr | Online (Deutsch) | Link zur Anmeldung                 24.09.2026 | 9:00 – 11:00 am | Online (English) | Link for Registration             14.10.2026 | 9:00 – 11:00 Uhr | Online (Deutsch) | Link zur Anmeldung         29.10.2026 | 9:00 – 11:00 am | Online (English) | Link for Registration          04.11.2026 | 9:00 – 11:00 Uhr | Online (Deutsch) | Link zur Anmeldung         10.12.2026 | 9:00 – 11:00 am | Online (English) | Link for Registration          13.01.2027 | 9:00 – 11:00 Uhr | Online (Deutsch) | Link zur Anmeldung            14.01.2027 | 9:00 – 11:00 am | Online (English) | Link for Registration       03.02.2027 | 9:00 – 11:00 Uhr | Online (Deutsch) | Link zur Anmeldung              

Furthermore, the Qualification program offers a broad spectrum of courses
that focus on different questions closely linked to this topic, such as:

• Visualization of Research Data
• Scientific Writing
• Research Data Management

Doctoral candidates in medicine

The Faculty of Medicine provides its own courses on the topic of “Good Scientific Practice” particularly for doctoral candidates in medicine.

More information on this can be found on the website of the faculty of medicine.
Find  dates and registration for courses of Dresden university medicine here.

Workshops for university lecturers

Workshops on "Good Scientific Practice" for university lecturers
will be available on the website of the Center for Continuing Education (ZfW).

Introductory workshops for doctoral candidates

Certified course instructors: At the request of supervising university lecturers and subject to time availability, the Graduate Academy offers interdisciplinary introductory online workshops on Good Scientific Practice (GWP) for doctoral candidates.

• Participants: min. 8 / max. 16 doctoral candidates (and postdocs)
• Language: German or English
• Format: online
• Duration: 3.5 hours

Please note:

• Requirement for doctoral candidates: Acceptance at a faculty of the TUD

• Requirement for Postdocs: Membership at the Postdoc Center

• The successful implementation of this depends on the availability and capacity of the GA trainers. While we make every effort, we cannot always guarantee it within a specific timeframe.

Are you interested?

In order to offer you the best support and ensure effective planning, we kindly ask for the following information from you:

• Preferred date
• Number of participants
• Workshop language (German/English)

Please email us at

Doctoral candidates & postdocs with questions regarding "Good Scientific Practice" can also make individual appointments with specifically qualified advisors at the Graduate Academy.

Would you like to book an appointment?

Please use our GA contact form.