Development of tools for analysing statistical/image manipulations, and for training scientists to avoid misconduct

Dubious statistical tests and inappropriate image manipulation are serious problems in biomedical literature. I will contribute to the development of new tools for analysing statistical and image manipulations, and training scientists to avoid misconduct. The project will result in free tools to help scientific editors and the scientific community judge papers, and provide training on how to appropriately handle statistical and image data. 

Part of the project will focus on tools to minimise the misuse of statistical data in publications. Expanded tests will help researchers, editors and funders detect dubious inferences, and I will create open-source software to educate scientists on acceptable forms of analysis. I will also create software to detect the more obvious forms of inappropriate image manipulation, such as duplication and transformation.

The open-source software will be available for editors and the wider scientific community. This work has the potential to improve biomedical literature.