This is a two-part training course designed for scientists to develop a robust understanding and skill set for effective scientific image processing.
Data misrepresentation in scientific literature is becoming a pertinent issue as more advanced tools for collecting and processing captured scientific images become available. The integrity of and public trust in scientific research rely on proper representation of visual data. Currently, there are no clear guidelines or training resources available to the scientific community on how to extract meaningful information from raw visual data responsibly, without misinterpreting, skewing, or obscuring the data. Furthermore, this understanding is crucial in helping scientists better recognize scientific image manipulation and misrepresentation.
This technology is a training course developed by Columbia University’s Office of Research Compliance and Training, termed “Handling Scientific Images: Dos and Don’ts.” This is a two-part course designed to effectively teach graduate students, researchers, journal editors, and others the ethical framework for approaching scientific image editing. This includes learning how to balance visual data clarity without compromising data integrity, efficiently managing digital assets and documentation, and recognizing cutting-edge data manipulation attempts, particularly with regard to AI-generated content. The primary takeaways from this educational experience are a theoretical understanding of processing scientific images and applicable guidelines and practical expertise in handling and analyzing visual data.
IR CU26018
Licensing Contact: Dovina Qu