Unexpected? Expected? Explict Disagreement to Previous ID Affects Following IDs

Curious to know if this is intended behavior. If it is, it’s no wonder there are some weird “community” IDs at unacceptable taxon levels with the concomitant need to unnecessarily “recruit” more IDers or specifically ask for a Withdrawal. Yes, I’ve seen the obfuscating verbiage of branch and whatever disagreements, and some of the forum topics addressing them, but they simply make no sense to me (my limited intelligence I guess).

The user explicitly disagreed with a first ID, as was appropriate. Why should that specific disagreement continue to show as a disagreement to following IDs? That disagreement was intended for one preceding ID, not any IDs that followed. I would logically think that a disagreement to a previous ID should then end with that disagreement, not continue to be applied to ensuing IDs. In a perfect iNat world, if an IDer then disagreed with the ensuing IDs, then another disagreement would be appropriate. Even odder, in this case, is that the following IDs were actually refinements of the disagreeing ID, not a totally different superfamily or order or other higher taxon.

The system obviously “knows” the ID to which the disagreement applies. I’d assume it could ignore it when additional IDs are made. So what am I, or the system, missing?

Thanks.

1 Like

It doesn’t. The cumulative count of the agreements for the species ID are exactly 2/3, and it needs to be greater than 2/3

3 Likes

A disagreement all the way up to superfamily Pyraloidea will actually count as a disagreement with Crambidae and all its descendants too.

It’s called “branch disagreement” and yeah, it can be super confusing. There was a note on the iNat blog last year that the disagreeing ID pop-up might be updated to be clearer with a more nuanced descriptor of what kind of disagreement is meant by the ID, but no updates have been made. Long posts and reading here:

https://www.inaturalist.org/blog/25514-clarifying-ancestor-disagreements
https://forum.inaturalist.org/t/change-wording-used-by-the-system-when-downgrading-an-observation-to-an-higher-level-taxa/3862

4 Likes

Thanks, but my question was more about why a disagreement that has nothing to do with following IDs is still considered a disagreement forever.

1 Like

I think the link from @bouteloua explains this. To some extent, no?
Or at least the history and reason for it.
Did you check it out already?

https://www.inaturalist.org/blog/25514-clarifying-ancestor-disagreements

(Skip to the bottom for the proposed changes)
I’m still not 100% sure if I fully understand the change though, or if the proposed changes actually solve instances like this…

It’s not. One more species level ID will tip it. Think of the explicit disagreement as being an ascertion that everything from that level down is questionable, so it counts as one vote against anything down from there. It’s only one vote, and can be countered with more weight for the other IDs

2 Likes

As Cassi points out, it is a problem with how the explicit disagreement modal is worded. And to be fair, it’s kinda hard to come up with a “fits every situation” version of the wording!

If you are saying “I (personally) can only be confident that it is this super family, am certain that it is not that species, but I would admit that it could still be in the same family as that suggestion”, then the wording of the question and outcome of the choice don’t exactly square up.

It does cause some heated debates!

Personally, if you are sure to superfamily, but NOT certain it can’t be anything else on that branch, then I don’t think it should be an explicit disagreement that you are making. However, as I point out, it can be over-ridden with more input by other IDers…

6 Likes

This is always a thorny problem! If I am certain that it is not Taxon A, but am not confident that I know what genus it belongs to, this seems to be the only available option.

As you mention, it doesn’t matter all that much. Either way, the correct ID will still need three “votes” to overcome the original incorrect ID.

1 Like

I proposed it back in June last year, and again recently… it would be useful to be able to challenge an ID. I don’t mean to challenge what the CID is, etc, but to challenge the specific ID that an identifier has made.

an example:

A species level ID, probably based on CV suggestion, of Agenus specifica by Anew Inatter

Dr Ultima Specialist chips in with an ID of Agenus (non explicit), drops a comment that there are 2 or 3 species that it could be, and that the genus level ID is more realistic. Note in particular that Dr Ultima Specialist doesn’t rule out the possibility of Agenus specifica being correct, but because s/he thinks that a realistic response from Anew would be to withdraw or change the overconfident ID, Dr Ultima also puts on a challenge to that ID.

The “challenge” would be like a flag, and would appear in the alerts stream for the user who made that ID. It would have no effect initially, but if left unresolved for more than 12 months, then that ID automatically withdraws. The owner of that ID can view that alert and choose to clear it, effectively reasserting their belief that it is that species, perhaps even being generous with a statement of why they still hold that position.

[as an alternative, it could have a much shorter timeframe, like 1 month… given that it’s only likely to be applied in genuinely challengeable situations, ie those where explicit disagreement is currently made to bump CIDs]

Anew never returns to the site, having completed the course assignment, and the ID and flag are never reconsidered. 12 months later (or 1 month) the ID would revert to withdrawn, and should Anew ever return or become active again, they could re-instate the IDs, possibly by reviewing alerts that might be generated when they were auto-withdrawn.

However, before the 12 months is up, Keen Amatar sees the challenge on that ID, looks at the photo and spots details that support the species level ID, even though it is not in the literature as a determining character, Keen may have worked a great deal with this species and be able to confidently support that ID, so can make a supporting ID, but perhaps also could clear that challenge flag so that Anews ID stands as valid.

I would expect that any application of a challenge flag would involve a clear statement of what the effect would be, and warning the applier to exercise that option with caution. It shouldn’t become a tool to annoy identifiers, and perhaps even should have limits to how many such challenges each user can make in a 24 hour period. Another safeguard might be where a flag is applied, and a second user needs to confirm it’s placement, so it effectively takes two to challenge. And remembering that it only withdraws the ID… the identifier can always reinstate the ID at any time, even after multiple challenges!

I see this “challenge” flag as being a sort of “are you sure you meant to make that ID”, and in the minority of such cases that they did mean to, the impact and roll back is as simple as clearing the flag “yes, I am sure!”

The wording on it could be “Dr Ultima Specialist has requested you reconsider this ID”, which softens it a lot from the “challenge” terminology I use in describing it.

2 Likes

This would have the advantage that if there were 5 “incorrect IDs”, one CV with 4 “agrees” by classmates, then ALL FIVE could be challenged and overturned in similar fashion, so there would no longer be the need to rope in 10 other identifiers besides your self to tip it. The Geralds observations just wouldn’t need to occur…

This could be great.
So many of the maverick IDs which hold things up just seem to be people who have come and gone.

1 Like

I imagine that if the option to “challenge” an ID was tucked away in the “upside down carrot” menu (which is a down-caret, btw :) ) along with the delete and edit options, then it won’t tend to be over-used. I would perhaps feel more comfortable knowing that use of such a feature could be taken away from someone if they abuse it, though!

1 Like

definitely gets my vote :slightly_smiling_face: