Code Of Ethics

• ONN Leadership Practices
  • Leadership shall provide guidance to staff through annual training and in performance evaluations.
  • The Director of ONN has the final say on what information is published in the event that an ethical dilemma is identified.
• ONN Research Practices
  • Collection of information: ONN staff are committed to:
    • Only soliciting or collecting open source information.
    • Using primary sources and original data wherever possible. Citing those who have first hand knowledge of the situation, local experts and authors.
    • Assessing source material to remove undue bias based on gender, nationality, race or ethnicity, religion, political views, age, ableness or sexual orientation.
    • Respect individuals’ reasonable expectation of privacy, anonymising data that might reveal personal information, but keeping original files secure and intact should the validity of the information be challenged.
    • Always obeying the laws of the jurisdiction in which the staff member is located, as well as the laws of Austria, the United States and, to the extent relevant, the extraterritorial laws that may apply by virtue of that person’s citizenship(s).
    • Never soliciting information that may jeopardise the safety or security of an individual or an organisation.
  • Analysis of information: ONN staff are committed to:
    • Identifying subjective assumptions and possible personal biases that may affect the analysis.
    • Treating all data sources critically; never assuming that information is error- or bias-free.
    • Never spreading misinformation or disinformation.
    • Using techniques to structure hypotheses and test them.
    • Not favouring one party over others.
    • When weighting data, declaring why the relative weights are assigned.
    • Never plagiarising; always attributing and identifying sources clearly.
    • Providing context that does not misrepresent or oversimplify developments; never stereotyping.
  • Internal peer review: ONN staff are committed to:
    • Vigorously peer reviewing analyses internally before publication.
    • Check sources are accurate, timely and reflect the views of those it is intended to represent.
    • Forming red-teams to test hypotheses, where appropriate;
      While there are no penalties for an individual’s incorrect assertion internally, ONN’s analysts, as a group, should guide the final publication under the direction of the Director.
  • Dissemination of information: In disseminating information, the following principles shall be respected:
    • Speed should not take precedence over accuracy.
    • Publishing information that escalates conflict should be avoided; guidance from peers and leadership to guard against doing so should be sought.
    • Fact and opinion should be distinguished and identified as such.
    • Ethical choices and processes should be explained to audiences; civil dialogue with the public about best practices should be encouraged.
    • Mistakes should be acknowledged and corrected promptly and prominently; corrections and clarifications should be carefully and clearly explained publicly.
  • Quantitative Analysis: When performing quantitative analyses, the following principles shall be respected:
    • Mistakes, both analytical and procedural, are inevitable.  Acknowledge and correct mistakes as soon as they are identified. Corrections should transparently be reported to the public.
    • Source data is often biased or selected on non-randomized factors. Quantitative analyses should always include a brief discussion of these limitations and how they may distort the analytical conclusions of a study. Analysts should also publicly disclose any gaps in their own technical understanding of a quantitative analysis.
    • Avoid the language of causality or deterministic predictions when a quantitative analysis does not support such conclusions. Instead use more conservative and probabilistic language when framing quantitative results. (Correlation does not equal causation)
    • Quantitative research should be accompanied with publicly accessible replication materials that would allow the public to replicate both the research design and specific statistical analysis performed.
    • Do not lie with data. Explore the distributional structure of any measure used before presenting a quantity of interest. For example, avoid reporting mean values when skewness distorts the central tendency and instead report the median. Likewise, avoid crafting data visualizations that present an incomplete or distorted picture of reality.
    • Properly attribute labor no matter how small. If someone provides coding help or designs a visual, attribute the creator appropriately. If someone gives feedback that makes the analysis better, acknowledge their help. If someone other than you provides a substantial amount of labor that makes the analysis possible when it otherwise would not be, consider making them a full co-author.

WEIGHING SOCIAL GOOD AND POSSIBLE HARM

• The public’s need for information should always be balanced against potential harm.
  • An evaluation of whether publication of the information or analysis could contribute to escalation of conflict, hatred or disinformation should be conducted;
  • Publication of the information or analysis should not contribute to discrimination on such bases as geographic, social or ethnic origin, political views, age, race, gender, sexual orientation, language, religion or ability.
  • Legal access to information does not translate into an ethical justification to publish.
• Do not exercise influence — unintentionally, out of inexperience or with malicious intent — in ways that risk ethical harm.
  • Consider the long term implications of the extended reach and permanence of publication.
  • Consult with peers or leadership to determine the best course of action.

INDEPENDENCE

• Conflicts of interest, real or perceived, should be avoided, and unavoidable conflicts disclosed to ONN leadership.

ACCOUNTABILITY AND TRANSPARENCY

• Information regarding an observed event, site or publication should be gathered, updated and corrected in a timely, explicit and transparent manner throughout the life of an observed event, site or publication.
• Staff may not engage in professional misconduct, such as plagiarism, distortion of facts, slander, libel, defamation or making unfounded assertions.