RG suggestion

This might be an unpopular opinion but I think observations should need more supporting ID's than 1 to become research grade. I know we all like it when we get a RG but I feel like only needing at least 1 other person to ID a species makes it too easy for misidentified species to become RG. Say an example being someone posts a picture of a Common Tiger Butterfly and ID's it as a Monarch, then someone else sees it and thinks it looks like a Monarch and ID's it as such without knowing? If/when these observations are used for research do these misidentified ID's can somehow be overwritten and corrected by researchers or iNat staff?
If this isn't already a feature I think there should be a ranking system in the iNat algorithm that sorts taxa by how well they're known/how easy they are to ID based off of how many observations they have or if they're just common enough. Taxa with less observations should take more supporting ID's to become RG so this doesn't mess with data of species that may be endangered.

Posted by wolfram06 wolfram06, May 15, 2022 06:12 AM


Having counts of each classification for a given observation would provide me with more options for accounting for inter-rater reliability in data analysis. Research Grade or not is helpful, but less helpful than having the empirical distribution of classifications.

Posted by galenseilis about 2 months ago (Flag)

Add a Comment

Sign In or Sign Up to add comments