The procedures undertaken to investigate a research misconduct are usually dictated by research ethics and integrity policy, prescribed either by the institute or by the national agency overseeing research. This policy would typically contain information on how an investigation should be conducted, as well as a non-exhaustive list of what constitutes research misconduct. Typically lacking from these policies would be a precise prescription of how the degree of severity of research misconduct could be determined. Adjudication of severity may often be left to the discretion of individual research integrity officers, or a committee of enquiry. Owing to the subjectivity of this process, the conclusion reached could vary between investigating officers/committees, even when adjudicating based on similar evidence. This variation would likely have an impact on the sanctions delivered. We hereby propose a research misconduct severity matrix, which considers eight independent ethical elements with different weightage, each assigned a numerical score by factoring against five different shades of severity (from minor to major). The sum of the scores associated with these elements returns the research misconduct severity score, a numerical value which would aid investigating officers/committees in reaching a consensus on misconduct severity, and better standardize sanctions meted out.
Keywords: Research misconduct investigations; research integrity; research misconduct severity matrix.