Research Grade algorithm

As a fairly experienced user who does a lot of IDs of others’ observations, I’m accustomed to seeing any obs with two IDs as the same species, and no contradictions, being elevated to research grade.
Examples: https://www.inaturalist.org/observations/19442017
https://www.inaturalist.org/observations/19983046#data_quality_assessment
I am also seeing observations that have 2 of the same ID and the only contradiction has been withdrawn, elevated to research grade. But lately (within the past few months) I’ve increasingly noted cases where there are 2 IDs, no contradiction that hasn’t been withdrawn, and no higher-taxon ID that disagrees with the species ID, still not elevated to RG. Example:
https://www.inaturalist.org/observations/20900690
I’ve seen a number of others that are similar. There’s a previous wrong ID but that’s been withdrawn, and the only other ID besides the two as the same species, is for a family including that species. But no R.Grade.

Is this a bug? Or is there some subtle wrinkle to the R.G. algorithm that I’m missing?

1 Like

welcome to the forum

this is an example of an ancestor disagreement

3 Likes

You could try tagging the disagreeing user in a comment to ask them to take a look. They could withdraw their disagreeing ID (even if they don’t want to ID to species themselves). Or they could be fine with their current ID in which case, one more species level ID would be needed to go to RG.

I’ve found that in cases like this, clicking, “No, it can’t be improved” kind of wakes the computer up to what’s really going on and it makes the observation Research Grade.

1 Like

Oh yes - you have hit my sore point. Hard disagreement. Despite the way us, merely humans view that obs. If you click - What’s This - at CID - you can see that iNat nurses that hard disagreement back to family as a grudge.

The subtle wrinkle forced iNat to write a blog post - in June 2019.
https://www.inaturalist.org/blog/25514-clarifying-ancestor-disagreements
In an October 2023 comment there @loarie says - We don’t have any plans to make changes to how disagreements work. Currently, if you add an ID of Family X which disagrees with an ID of Species Y, you are disagreeing with everything in between Family X and Species Y (e.g. Genus Z).

I got tired of explaining to scientists - you and I know that both species are spiders, duh. (But I am left feeling like a monumental fool - and there are a few taxon specialists who ever so politely hate me) So I made a text expander to pair with a link to that blog post.

iNat does ‘ancestor disagreement’
You disagreed with the wrong taxon, therefore, you disagree with all subsequent IDs.
Please reconsider?

(Rant at Computer Vision algorithm over)

3 Likes

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.