But that does not solve the issue of many possibly wrong IDs flooding the map… and worst part of all, drowning the few interesting legid observations ( e.g. because there are epigynal photos for a spider provided) in it’s way..
5 Likes
but, that is what ajott - spider specialist - has explained. For an iNat user, whether identifying or not, we have only the info in photos or notes or comments or … We can only use what is available to us on iNat. John is The Fundi and knows what it is, is only useful to iNatters if John is willing to share that info.
6 Likes
Since it sounds like this is connected with observations by one particular user (rather than many different users), it might make sense to start a conversation with that user to find out what the context of these observations is (i.e., how they arrived at the ID and whether there is additional evidence they did not provide) and what the user’s expectations are (i.e., is it just for their own record-keeping or do they expect confirmation)
As I the variety of responses here suggests, people use iNat in different ways and there may be a variety of reasons why a presumably knowledgeable user is uploading observations with IDs that do not seem justified by the evidence. I would probably respond differently depending on what these reasons are.
My own position is that the ID system on iNat is based on the principle that IDs should be supported by the evidence provided. In other words, if microscopy or DNA sequencing was carried out, it is the responsibility of the observer to document this in some way that allows other users to assess the correctness of their conclusions. For fungi/lichens, in most cases I presume this would be photos of the microscopy or the outcome of the DNA sequencing, or possible the results of chemical tests. I am willing to consider written notes as evidence, but these notes need to convincingly tell me something about what the observer saw; thus, a note “confirmed by microscopy” would not be sufficient, but I would treat a description of the structures seen under the microscope as relevant evidence, provided there is reason to believe that the user is accurately describing what they actually saw, not merely repeating what the standard key says.
If the user knows that this evidence is needed for ID and is not providing it, this might be because: a) the observation is for their own records and they do not care if it gets confirmed; b) they intended to upload additional evidence but it hasn’t arrived yet or they are busy and haven’t gotten around to it; c) they don’t think they need to because they confirmed it in the lab or some other expert confirmed it.
In cases b) and c) I might remind them that on iNat we are strongly discouraged from making IDs based on someone’s expertise that cannot be confirmed by the evidence and would encourage them to make the effort to add this additional information, which will help make iNat’s data stronger.
If the observer is not interested in providing additional evidence, I think I would have to figure out whether I am comfortable leaving the observations with unconfirmable IDs or whether to push the IDs back to a level that does seem to be justified by the evidence. This decision might depend somewhat on how convinced I am that the observer in fact did their due diligence and is skilled enough to make the IDs in question, rather than merely guessing. It would probably also depend on how notable the record is – e.g., if it is a common species and the most likely one at the location of the observation, I might be more likely to leave the ID but not confirm it. If it is a species that is commonly misID’d or ID’d without sufficient evidence, I might leave a comment that the ID is not confirmable as information for others who might look at the observation. If the ID is for a taxon that would be rare or unexpected at that location, it would probably feel more important to make sure that it is supported by evidence and I would actively disagree to push it back to a higher level.
It might also depend on how the user feels about me disagreeing. Very occasionally users have told me that they identified something in the field but can’t provide better evidence and they understand if this doesn’t meet the criteria for ID on iNat. Or if they are merely uploading for their own records, don’t plan to provide additional evidence, and don’t mind if the observation becomes casual, it might be feasible to come to an agreement with them that I will push back the ID and if they want the observation listed under their ID they will opt out of community taxon and one of us will use the DQA “ID cannot be improved” to make it casual and take it out of “needs ID” and observation maps for that taxon. I don’t know how many users would be amenable to such a solution, however.
7 Likes
True, but most of the time I’m not CERTAIN it can’t be identified from the picture. I just suspect so or suspect the observer has not been able to do that.
1 Like
I understand treating someone differently if you know them personally, but at the scale we’re working with on iNat I don’t think it’s realistic to expect identifiers to vet every observer they identify for. Especially with difficult taxa that have an overwhelming number of needs ID observations, I wouldn’t want to burden identifiers with extra tasks like that. It’s much more efficient and manageable to treat all observations by anyone consistently, and let observers respond as they see fit.
9 Likes
In that case, of course yes - do not bump them back. But for a lot of cases the current literature is clear rnough on that matter.
I simply don’t care if the submitter is a so-called superuser or a newbie.
I have seen “superusers” uploading tons of ugly and mostly unidentifiable observations and much more frugal ones creating beautiful observations with plenty of details.
So, if the organism is identifiable to the species I provide my own ID. If it is identifiable to an above-the-species rank, I lower the precision of the ID.
8 Likes
My approach to ID is that my ID stands on it’s own. If an observation is more specific than I can ID, I put in what I can. I won’t disagree with the previous ID unless I can actually see something that shows otherwise. Especially if the ID comes from the observer. They were there and have more to work from than just the photo. Uncertainty is not evidence. If you told me you have a pen in your pocket, I’d be stupid to tell you otherwise because the clip I can see could be part of a knife.
I sometimes ask questions and/or point out how difficult it is to ID those to species without dissecting the specimen. Sometimes I get a response or a change to their ID. Sometimes not. I just truck along adding IDs as I can.
4 Likes
It’s true that when we observe we look at many features that don’t always appear on the photos. Specially when some photos end up a bit blurry or the animal just moved. I think this is why an observer comments should be sufficient evidence if s/he is specific and objective about what s/he saw.
But in case of doubt, it’s not a big deal if it’s not RG. For the sake of data integrity, I rather have a casual ID that I trust than a RG that I can’t.
The priority for me it’s to be in the moment when I’m out there and observe truly, instead of shooting [pictures] at everything that moves.
2 Likes
Your approach is great for me!. When I get an ID that disagrees with me, I take the onus to go back and figure it out. When someone takes the time to point out what I missed or what characteristics would help me make them more informed identification I’m over the moon happy about it, and more likely to ask them in the future for help, but I think the onus is on the observer
5 Likes
How would you know whether someone was objective?
I run into those cases very frequently - AND have also experienced them myself on the observer side - where people swear dead sure a certain feature of a characteristic (often it is about size in the arthropod world). Often those observers are sure it HAS to be the other species, because the observed organism was huuumongous… and they are not aware that there are other visual cues that very clearly without any doubt indicate the other species.
I totally get how that can be very frustrating. I have been there. So in those cases I usually respond exactly along that line, in the hopes the observer does not feel silly afterwards realizing that everybody can make such a mistake… but I will stick with my disgreement when I am sure.
2 Likes
Oh, so many different thoughts. As for myself as observer, I know I make mistakes. I hate to be wrong but I’m kind of used to it. Go ahead and tell me. I’ve learned a lot from corrections.
If I can take it, I figure you can. I try to phrase my challenges kinds or neutrally, but I will change an ID to a different species or a higher taxonomic level if I’m pretty sure the ID is wrong. I consider it kind of irresponsible not to.
How I respond to other people’s observations and identifications results from a swirling mass of factors including what’s in the photo, photo quality, ease of identification, notes, abundance of that subject in that area at that season, and the skill of the observer/identifier (if I know about it).
So, if the observation labeled American Robin is an out-of-focus gray bird of a shape that doesn’t rule out American Robin, photo’d in western Oregon, I’ll agree. There are behavioral and auditory clues to robin identification that the observer probably saw/heard. The bird is common to abundant all year. It’s easy to ID. So it’s very probably correct, no matter who posted it.
I welcome notes and take them seriously. That doesn’t mean I always believe them. For example, if the notes say the grass leaf sheath is open, I’ll believe that. Although a note that the sheath is closed would be helpful if true, I’m more cautious about accepting it because people so often mistake the overlapping margins of an open sheath for a closed sheath. I’ll believe it if I know the observer is experienced enough to tell the difference, though.
Hard situations are those where I know the observer is good at this kind of identification but I disagree. I think those over. I may just think, “I’m not competent to challenge this one” and move on. Otherwise, I’ll ask questions. If I’m sure it’s an error, I just add the correct ID and leave some explanation.
And then there are the cases where a person is a world expert or at least an expert in my regions and the plants are ones they know very well. I’ll believe them. If they provide some explanation, I’ll probably think I can see what they’re talking about and I’ll agree. I think of it like consulting a book.
Don’t think that means I just accept whatever an expert says. For example, years ago, my colleagues and I asked the one person in the world with the best understanding of a difficult plant group to write an identification key for the species in our area. He kindly did that and sent it to us. We tested it and didn’t like parts of it, so we rewrote it and sent it back for him to review. He was very nice about it but I think he was shocked. :-)
13 Likes
True… actually. I’ve been looking at some examples recently and It’s quite tricky to assess that. And the level of assertively of some observers can be quite high being backed more by pride than data.
I was initially thinking about comments like “the bird had a yellow stripe in the crest” for instance, despite that it doesn’t show clearly on the photo. For those comments we can trust the observer.
But “, as observers, we should also not be afraid of receiving disagreeing IDs. I’ve seen birders taking it like challenging their own ID skills which in reality is nothing more than: “with the photo you’ve shown, it doesn’t look like that species, rather more something like that”.
2 Likes
This is what I meant. Notes should be considered in the assessment. After consideration it may result still in a disagreeing ID and that’s fine, with a comment explaining it the observer should understand why. If not, then he can opt out of Community ID.
2 Likes
“it doesn’t matter to some observers if their observations don’t all reach RG” That’s me too. For what I do (Texas Lycosids) I’d rather not have anyone agree with me unless they really know their business. Disagreements are always fine because hopefully either of us may learn something when the other supports their position. Some species are relatively easy to ID. Others are a real struggle. For many, the existing scientific literature is inadequate for absolute IDs or flawed in some other way. For Lycosids, getting observations sorted in any rational way is a good thing. Developing field identification guidance is still in the early stages but when that is further along it will be an advantage to have some preliminary sorting done because the errors will stand out better.
5 Likes
I don’t factor in expertise or number of observations when adding identifications. Even experts are human and occasionally an ID has to be corrected or pushed back to a higher taxonomic level. If you are unsure you can always leave a message on their post asking for an explanation for their ID.
5 Likes
The communities of prolific observers and active identifiers in my areas of interest are quite small (except birds). I recognise the user names of experts who assist when asked and observers with good photos or rare finds. I’ve corresponded with most of them and I have an inkling how they use iNat.
If I can conform or identify some observations I will do so. If not, I won’t.
Are you speaking about the original IDer always adding all known info? I guess I’ll stick with the comment that it ain’t the original observer’s job to make INat easier for YOU to use. If you want more info, YOU should ask for it. After all, this thread is about what superusers could do as best practice, not simply to criticize what all observers do or don’t do, or should or shouldn’t do.
Ok, I can see bumping to genus, but I think best practice would be the super-duper user to bump it up to genus, but always tell the original observer why they did that. Bumping to genus with no comment could be a either (1) a super-duper specialist or (2) a user with too little experience with the genus to figure it out to species. The original observer nor anyone else who wants to throw in a suggested ID can NOT tell which kind of user is bumping up to genus without a comment.
So you bump to genus with a comment, not just bumping to genus. That sounds like the right approach to me.
1 Like
Safe bets are only statistically right most of the time. Maybe they are a user you don’t know who has a photo graphic memory, so stating you couldn’t ID from the information given challenges them to explain.