The general rules of scientific integrity have been layed out in the Statutes for Safeguaring Good Research Practice at KIT and are legally binding. The statutes apply to all persons working scientifically at KIT and obligate them to comply with the general principles of good research practice. Depending on the research discipline, additional rules of good research practice might be relevant to the research process - and have to be taken into account.

Researchers bear a special responsibility for the persons they supervise as well as for all their subordinated employees. Supervisors have to make sure, that their subordinates know and apply the standards of good research practice to their work. You can find futher information on this topic in the Guidelines for the Doctorate at KIT as well as the Guidelines for the Postdoc Phase at KIT and the KIT Code of Conduct for external use (the version for internal use provides references to the corresponding intranet pages).

Below, the paragraphs of the KIT Statutes on Safeguarding Good Scientific Practice as well as the corresponding guidelines of the DFG Codex are summarized thematically.

Researchers at KIT have to comply with legal framework conditions as well as regulations and guidelines of KIT in order to conduct research. If necessary, they have to obtain approvals or ethics votes for their research proposals.

Relevant legal framework conditions are, e.g.:

• Data protection
• Copyright law
• Labor law
• Employee Invention Act
• Rights and obligations from contracts with third parties

Relevant KIT-internal regulations and guidelines are, e.g.:

• KIT Code of Conduct [internal version][external version].
• Underwriting guidelines [internal version (only avaliable in German)].
• Guidelines for Ethical Principles of KIT.

Additional framework conditions should be set by the scientist at the earliest time possible, such as determining the rights of use for the research data. Please apply special diligence to the subject, when the research projects requires the participation of several - academic as well as non-academic - institutions, or when scientists change their institution.

At KIT, there are a number of persons and institutions to give you advice about the applicable framework conditions:

• The KIT Ethics Committee comments on individual research projects, when there might be reason for ethical concerns or if the wish for an ethical review is expressed. The two Ombudspersons for Ethical Principles advise KIT members and staff on ethical issues concerning everyday research and concrete research projects.
• FOR offers help with the initiation and application of third-party funded projects.
• Help on data protection issues is available from the Data Protection Office (website only avaliable in German).
• Help on copyright law for publications is provided by the KIT Library.
• Help for other legal matters is available at RECHT (website only avaliable in German).
• Help on all questions concerning research data can be found with the Serviceteam RDM∂KIT.

At every stage of a research project, it must be clear which people are involved and what their roles and responsibilities are. This also applies to PhD students, student and research assistants, laboratory staff and other persons involved. If conditions change (e.g. the focus of work of one of the persons involved), roles and responsibilities are adjusted, if necessary.

Every research process is to be carried out lege artis (= "according to the rules of the art"), wich means, researchers have to observe the standards related to their research field and are asked to use established methods in their research. The application of a method sometimes requires specific expertise, which may be ensured by cooperative agreements with persons and institutions within KIT or with external partners. If new methods are developed or applied, quality assurance and the establishment of standards are of particular importance.

Thorough work is a basic prerequisite for a high quality research processes, for example, in calibrating equipment, collecting, processing, and analyzing research data, selecting and using research software as well as developing and programming it, and keeping laboratory notebooks.

Good planning facilitates quality assurance: creating a research data management plan before starting work is recommended. Assistance with research data management is provided by the Serviceteam RDM∂KIT.

When planning a research project, the current state of research must be considered and acknowledged. This requires a comprehensive search for already published research results, established standards, and field-tested applications in order to build new, relevant research questions on them.

Assistance with finding published research results and research data is provided by the KIT Library.

All information relevant to the development of a research result must be documented in a comprehensible manner, to ensure, that the results can be properly reviewed and evaluated. This includes information on the research data generated or used, methodology, evaluation and analysis. In some cases, the origin of the research hypothesis must be recorded. Citations must also be documented. If you develop research software, you have to record the source code along with all other relevant information.

The documentation of the mentioned information has to be done according to professional recommendations and standards. Individual results that do not support the research hypothesis must also be documented!

If necessary, electronic laboratory notebooks (ELNs) can support the data collection. The Serviceteam RDA∂KIT provides advice on ELNs as well as the documentation of research data in general. Basic information on how to get started with research data management is provided on the English Pages of the information platform forschungsdaten.info.

For the documentation of citations, a reference management program can be helpful: The KIT Library offers training and advice on Citavi and Zotero.

It must be ensured that (unconscious) preconceptions do not lead to bias in the interpretation of research results. To prevent bias, the application of appropriate methodology (e.g. blind studies) is necessary.

In addition, it must be examined whether gender and diversity may be significant for the research project and to what extend. This applies with regard to the methods, the work program, the objectives, etc.

When interpreting findings, the respective framework conditions are taken into account and appropriate methods are used in order to avoid (unconscious) bias in the results.

As a general rule, research results obtained are to be published and introduced into the scientific discourse, insofar as legal framework conditions and ethical reasons do not prevent publication. As far as possible, third parties should be granted access to all relevant information necessary for a possible replication of the research results. Scientists decide on their own responsibility - taking into account the customs of the relevant field - whether, how and where their research results are to be made publicly available. This decision must not depend on third parties.

Scientific publications should:

• Describe results in a complete and comprehensible manner.
• Always outline the quality assurance mechanisms used, especially when new methods are developed.
• Identify the origin of data, organisms, materials and software used in the research process and provide evidence of subsequent use.
• Provide complete and accurate evidence of own and others' prior work through citations and references.
• Repeat previously published results only in clearly identified form and only to the extent necessary for understanding the context.

Furthermore, in keeping with the idea of "quality before quantity", inappropriately small publications are to be avoided.

KIT Library provides information on publishing for KIT scientists.

As a basic principle, the scientific quality of a contribution does not depend on the publication organ in which it is made publicly available. Nevertheless, the quality of the publication organ shoud be taken into account, when publishing research results and data.

Authors carefully select the publication organ - taking into account its quality and visibility in the respective field of discourse. Publication organs include books, journals, professional repositories, data and software repositories, as well as blogs, workshops and scientific conferences. A key criterion in the selection decision is, whether the publication body has established its own regulations to ensure good scientific practice (e.g. peer review).

Scientists, who assume the function of editors, carefully consider for which publication organs they assume this task.

Particular caution is required, if publication organs do not or only insufficiently perform the necessary quality assurance prior to publication ans use aggressive marketing: The KIT Library provides information on its pages on how to check a journals integrity. The Helmholtz Open Science Office maintains a FAQ on the topic of Predatory Journals and Predatory Publishing.

Research results, the central materials they are based on and, if applicable, the research software used, must be retained for at least ten years in accordance with current standards. In particular, this applies to: Measurement results, software codes, simulation results and analytical calculations, collections, study surveys and questionnaires, as well as cell cultures, material samples or archaeological finds. If there are reasons for not retaining certain data (e.g. for protection of personal data), or if shortened retention periods are appropriate, the reasons must be documented in a comprehensible manner. The retention period begins on the date on which public access to scientific knowledge is established.

The data shall be stored on durable and secure media at the institute where the data originated, at other reliable institutions (esp. libraries or archives), or in renowned repositories (e.g. KITopen and RADAR4KIT). Repositories can also be used to publish research data. The repositories used for this purpose should be listed in re3data.org. When archiving research data, the FAIR principles (Findable, Accessible, Interoperable, Re-Usable) should be taken into account. If research software, developed in-house, it is to be made publicly available to third parties. This is usually done by providing the corresponding source code and using an appropriate license. The source code must be persistent and citable.

The heads of the scientific working units are responsible for ensuring the preservation of research data. They are urged to issue suitable regulations für preserving research data on the basis of legal provisions and the principles of scientific work, recognized in the respective field.

KIT supports the storage of research data by providing the appropriate infrastructure. The Serviceteam RDM∂KIT advices on suitable storage options for research data. Subject-specific standards for the documentation and storage of research data are developed by the consortia of the National Research Data Infrastructure (NFDI). The KIT Library advises on publication in the KITopen repository and the use of the RADAR4KIT repository.

As a general rule, an author is someone who has made a genuine, traceable contribution to the content of a scientific text, data or software publication. We assume, that researchers make a coprehensible and genuine contribution to a publication, when they put substantial work into at least one of the following steps of a publication:

• Development and conception of the research project
• Development, collection, acquisition and provision of data, software and sources
• Analysis/evaluation or interpretation of the data and sources, and the conclusions drawn therefrom
• Writing of the manuscript

Inadmissible is an "honorary authorship": A person, that made no contribution to the publication in the sense of the above points, ist not an author. A supporting role may be appropriately acknowledged in footnotes, the Foreword, or Acknowledgement. Contributions that are not sufficient to warrant authorship include, but are not limited to:

• Mere organizational responsibility for obtaining grant funding
• Provision of standard investigative materials
• Instruction of staff in standard methods
• Merely technical assistance in data collection
• Mere technical support (e.g. mere provision of equipment and experimental animals)
• Mere provision of data sets (usually)
• Mere reading of the manuscript without substantial contribution to the content
• Management of the research work unit in which the publication was created (definition of the term Research Work Unit: §5 (2) KIT Statutes GWP).

On the basis of the criteria mentioned above, the persons involved agree on who will be named in the publication and in what form. In doing so, they ensure that no one is left out.

The order of authors aligns with the conventions of the research field. Agreement on the order of names is reached in good time, i.e. usually no later than when the manuscript is being formulated.

Authors of a scientific text, data or software publication always bear joint responsibility for its content. It must be ensured, that all authors have been named and have agreed to the respective submitted version of the publication.

Without sufficient reason, a required consent to a publication of results may not be refused. A refusal of consent must be justified with a verifiable criticism of data, methods or results.

Furthermore, authors shall ensure, as far as possible, that their research contributions are marked by the publishers or the infrastructure providers in such a way, that they can be correctly cited by users.

If researchers made findings publicly available and subsequently notice discrepancies or errors, they correct them. If the discrepancies or errors are the reason for the retraction of a publication, the scientists shall work as quickly as possible with the relevant publisher or infrastructure provider etc. to correct or retract the publication and mark it accordingly. The same principle applies, if the scientists are informed of such discrepancies or errors by third parties.

The evaluation of scientific performance requires a multidimensional approach: In addition to scientific performance, the individual person must also be considered. Therefore, individual characteristics can also be included in the performance evaluation.

Originality and quality are key in evaluating for examinations, awarding of academic degrees, promotions, hiring, and appointments and have priority over quantity. This applies specifically to the performance- an load-based allocation of funding in research. High-quality in science is measured by discipline-specific criteria. Quantitative indicators should be included in the overall assessment with appropriate differentiation and reflection only.

For the inclusion of individual characteristics in the performance evaluation, the General Act on Equal Treatment (AGG) must be observed. The following aspects may be taken into account, when evaluating the performance of scientists (provided that the applicable legal framework does not prevent it): commitment to teaching, academic self-administration, public relations, knowledge and technology transfer, or contributions in the interest of society as a whole, as well as the person's scientific attitude (e.g., with regard to openness to knowledge and willingness to take risks).

If voluntarily indicated, individual characteristics in CVs can also be included in the assessment. These are, for example, personal, family or health-related downtimes (and training or qualification periods extended as a result), alternative career paths or comparable circumstances.

Reviewers of scientific papers must use a transparent evaluation system and maintain their independence as examiners. This also includes, that they carry out the evaluation unbiased. Researchers shall immediately report any conflicts of interest to the responsible office. In examining a research project, conflicts or bias can occur regarding the project itself, the person(s) involved or the matter discussed.

Strict confidentiality is mandatory when reviewing and assessing submitted manuscripts, funding applications or the credentials of persons. The same applies to working in advisory and decision-making bodies. The disclosure of third-party content to which the reviewer or committee member gains access to another third pary, as well as the use of the same for one's own purposes, is therefore forbidden.