Platform(s), such as mobile, website, API, other: Website
Description of need: When replacing a higher-level disagreeing ID with a lower-level ID on the path between the old ID and the initial ID, the new ID no longer disagrees with the initial ID, since it does not disagree with the CID.
So as an example, if the initial ID was Dandelions and I originally made a disagreeing ID of Dicots but now want to improve it to Asteraceae, the CID of Dicots disagreeing with Dandelions turns into Dandelions with a non-disagreeing ID of Asteraceae.
I imagine that few people improving their previous IDs would expect (or want!) this sort of behaviour. It can be avoided by taking the extra step of manually withdrawing the first ID before adding a second, but if you donāt realise it will happen, you may not even notice the CID ends up wrong.
Feature request details: When an ID is being replaced, the personās old ID should be withdrawn before the new ID is added so that the new ID considers the ārealā state of the observation ID and can question whether the ID disagrees with the initial ID if relevant.
Edited to add, based on discussion in the comments: An alternative proposal could be to make it so that adding an ID will always prompt the ādisagreementā popup if any of the individual IDs is below the ID being added, regardless of the CID, allowing explicit disagreement with any individual ID(s) or all lower IDs. This would solve not only the current problem but also other issues raised in discussion.
For ālogicalā consistency, I agree that the final result should be the same with or without the extra step of manually withdrawing the first ID. So, I vote for the feature request.
whether the new ID disagrees with the Community Taxon resulting from the prior removal of the previous disagreeing ID, if relevant.
I was certainly caught out by this a few times early on. Iād be inclined to say that it should respond in exactly the way it would if the original ID had not been made - i.e. giving you the pop-up choice again - rather than automatically keeping the disagreement. I think that is what is being suggested here, but I wanted to make it more explicit
I agree with the need here but there could be a slightly better solution, because there are situations (albeit not very common ones) where the opposite outcome happens - i.e. when withdrawing before adding a new ID would remove the opportunity to add a disagreeing ID.
The easiest example is if your own ID is the only ID on a sighting. For example, letās say I have a sighting of species XY and my ID of XY is the only ID on the sighting, or at least the lowest ID. Now letās say that before anyone else adds an ID, I realise that actually there is another very similar species that often gets confused with XY, and you cannot distinguish them from my photos. I would like to change my ID from species XY to genus X, but in the process I would like to explicitly disagree with the old species ID because I know that we cannot tell, to pre-empt future species IDs. This works fine currently, but if my ID were withdrawn before I added the new ID then I would lose the option to disagree with the old ID.
This situation is perhaps not very consequential but there are other similar scenarios where it has been very useful to me, mostly when there is a maverick sighting etc. and three other IDs (one of which is mine) which I now want to disagree with.
Iām not certain if there is a way to solve both problems at once. Maybe we could have the system check for disagreements both before and after the ID is withdrawn, and then give an option for disagreement if either cause a disagreement?
Thatās certainly more correct - I was trying to keep things simple, but maybe I went too far.
I guess thatās what I assumed would be the case if itās done as āwithdraw then add new IDā - but yes, itās probably worth making it clear.
I had to think about your examples for a minute before I got your point, but I see what you mean. Iām not sure whether your suggested solution is possible, but it sounds good if so, to avoid those downsides.
I agree this case is possible and it would be a bit nice to be able to mark as disagreement if so, but it seems much more fringe than the current issue.
Itās also not an entirely novel issue, there are already plenty times where I might be IDing at e.g. genus level and want to make note that I do not think it is some particular species but cannot do so via the disagreement system since no such identification is present.
For example, I could see an Examplegenus in my backyard, and a subtle field mark let me rule out E. commonspecies. For this reason I might want to make special note that I can rule out E. commonspecies due to this field mark even if I donāt know the exact species. I canāt mark as disagreeing since no E. commonspecies ID exists, but I can leave a comment explaining why I donāt think it is E. commonspecies and I imagine that would work basically just as well. If someone does later ID as E. commonspecies I can just go back and disagree anyways.
The potential issue you point out seems similar enough that I imagine it can be handled the same way.
If iNat is going to silently remove the active disagreement - and thereby change the CID effect - they should at least ask us - agree or disagree? - first.
I agree that the simple example is pretty niche and your workaround is probably all that is needed. But Iāve come across it in more complex situations where there is a maverick ID involved and it is not always so easy. E.g. there is an initial ID of Correctgenus wrongspecies, then three people (including myself) add IDs of C. closespecies. I then realise that oops, actually itās not C. closespecies at all - maybe itās something different that I donāt know, or maybe we canāt tell from the photo - so I want to add a disagreeing ID of just Correctgenus. Currently, I can actively disagree with C. closespecies when I add my genus ID, but if the old ID was withdrawn first this wouldnāt be possible. This isnāt always an issue but it becomes an issue in the frustratingly frequent times when the only thing that happens is just that the identifier who put C. wrongspecies retracts their ID. Nobody gets notified of any change and the CID updates to C. closespecies, which is less than ideal. I have a couple of dozen sightings bookmarked that I check on every now and then for just this type of situation (but they have only two disagreeing IDs because of the current system) and when I find one where the above situation has happened I have to add another disagreeing ID.
To be fair, I think the initial solution suggested by @grampianshiker is much better than the current situation because that happens far more often, but I think it would still be great if we could avoid this other issue as well.
The current behavior is consistent with iNat expectations from an identifier:
the identifier is expected to ID at the lower possible rank (in your example, genus), and thatās it,
if the current Community Taxon is as good as it can be (in your example, genus), then you can check the DQA box āNo, it canāt be improved, it is as good as it can beā.
So:
you donāt need to explicitely disagree with XW, XY, XZ (species) when IDing as X (genus), if there is no prior ID XW, XY, XZ,
but you can state that this ID genus canāt be improved (indicating that the ID X is not an effect of ignorance, but an effect of species indistinguishability).
An ID with disagreement is not for that.
A possible improvement would be a combobox to choose the taxonomic rank (genus) associated to the DQA āit canāt be improvedā, so that the box could be checked and could be meaningful even before the Community Taxon is at rank genus, or after the Community Taxon was at rank genus if it get changed by future IDs. I think this issue about this DQA flag, when the C.T. changes, has been mentioned already on the forum.
Ah I misunderstood the circumstances you were describing!
Iām now realizing Iām not sure if I completely understand the disagreement system.
In the case where the CID is at genus level, but there exist lower taxa IDs do you get prompted to ask if you disagree?
Like if thereās one C. wrongspecies ID and one C. closespecies ID and you add a Correctgenus ID does it prompt you for disagreement at all?
If not, it seems all of this could be resolved by simply letting you disagree with any combination of the existing lower rank IDs when you ID at a higher rank. E.g. in the aforementioned case giving you checkboxes for whether you disagree itās C. closespecies and if you disagree itās C. wrongspecies regardless of CID.
Before your ID and after - it is CID at Genus. Your ID makes no difference - and you cannot make your ID disagree with the wrong one. Unless you support the right one. But you can comment to explain for the next identifier. Working as intended, but not as expected.
I agree that this is a scenario that can easily be dealt with in other ways, but itās the āsimplestā scenario that shows the current behaviour so thatās why I used it as an example. The more useful and more common scenario is the one I described in a later comment where there are prior IDs of XW, XY, XZ etc. but the CID is at a higher level because of disagreeing IDs. In these situations itās very useful to be able disagree using the current system.
No, the system doesnāt prompt for disagreement in that scenario. It will only prompt you if your ID is of a higher taxon than the community ID. So in the scenario Iām talking about, if the ID is withdrawn before the identification was added, the CID would get bumped back up above species and there would be no opportunity to choose disagreement. For the initial scenario detailed by @grampianshiker itās the opposite - if the ID is withdrawn before the identification was added, the CID drops down to a lower level which would enable a disagreement to be chosen.
I had a look for a good example to hopefully explain it better. So take the following scenario:
Now imagine that actually Iāve changed my mind, and Iāve come to the conclusion that this is actually an undescribed species of Onchestus, or maybe I know that it is not O. rentzi but I am not confident on which of the other species it actually is.
To start with, the CID is O. rentzi. Currently (Scenario A), if I add an ID of Onchestus it will prompt me asking whether I am disagreeing, because the CID does not change before I add my ID, and my genus ID is coarser than the CID. I would choose yes, and the CID would be bumped back up to Onchestus.
If we change it so that my identification is withdrawn before my new ID is added (Scenario B), I do not get the option to disagree. This is because the removal of my identification will bump the CID all the way back up to Infraorder Anareolatae, and my ID of Onchestus is finer than that. So I add my new ID without disagreeing, and the CID gets bumped back up to Onchestus as well.
Now, something that happens quite often with these situations is that the Red identifier realises they are wrong and simply withdraws their ID. I would probably consider that ideal behaviour from them - I donāt want them to agree with my ID just because they think that Iām an expert. But this causes problems because nobody is notified of anything.
In Scenario A, this is not a problem. We are left with two IDs of O. rentzi and one disagreeing ID of Onchestus, so the CID stays as Onchestus. But in Scenario B this becomes a problem because we now have two IDs of O. rentzi and one non-disagreeing ID of Onchestus, so the CID gets pushed down to O. rentzi. I donāt get notified of this so I canāt add in a new disagreeing genus ID, and the sighting is now one that I have reviewed but there is still more that I could do to correct the ID.
This situation does currently happen when there are only three IDs total, so in the above example imagine that Purpleās ID is not there. In these cases I bookmark the sighting and check on it every now and then to see if Red retracts their ID, which happens quite often. If we change the system, it would take four opposing IDs for this to not happen as opposed to three.
Now I agree, this is not as common a problem as the initial scenario given by @grampianshiker. And if we had to choose between one or the other, I definitely vote that we should go with the more common issue. But itās not a negligible issue and it does happen fairly frequently, at least in the taxa that I ID. So if we can find a fix that solves both problems, I think that would be ideal. I think my suggestion of checking for disagreements both before and after the ID is withdrawn would be ideal, or simply checking whichever one causes the CID to become coarser. But there is probably a better solution out there!
I suppose realistically the ideal situation from an identifierās point of view would be to be given the chance to explicitly disagree with any lower-level ID that has been suggested, rather than only focusing on the CID, but Iām not sure whether this is feasible? Because youāre right that itās not ideal to be unable to disagree with a species-level ID in the right genus/family if thereās a previous incorrect lower-level ID pushing the CID to a higher level.
Yes, I was typing something similar and agree wholeheartedly. I am not sure why the CID is involved at all in the disagreement process, it seems always useful to have the option to explicitly disagree with finer IDs on an observation regardless of CID, and it seems as though this would fix all mentioned issues.
I agree, ideally it would be fantastic to be able to choose what to disagree with. I think it has been mentioned a couple of times here in the past and it might be in the works, but my understanding is itās a lot to implement so it might be some way off.
