At some point I found an observation with the opposite problem - it was casual for no obvious reason (it had one ID and meet all the Research Grade criteria). I agreed with the ID and it went from casual to RG. I’m mentioning it in case it helps find the cause.
When I remove the “Southern Africa” filter I still see 181 observations with this issue. And it must not be a legacy thing because some were created as recently as today.
Probably an indexing glitch, since it looks like the quality grade changes once you perform any action (like faving) that causes the observation to be re-indexed. We’ll take a look.
What reindexing? Is this related to the recent major outage, i think i remember reading something about reindexing there? If so i wouldnt consider it a job for the community to manually click through >1000 observations. This is a problem of iNats internal workings and should be moved to bug reports then.
I’ve reindexed all these observations. The results at the URL posted may be cached for a little bit, and when that expires it should should no results (you can modify the search by adding a date param to confirm in the mean time). It’s likely there is some action that is not properly setting quality grade or reindexing observations afterward, so we can look into tracking that down. Thanks for the report.