What happens to incorrect ID’s? Should I take them down if I think they’re wrong? Will they disappear if others don’t verify them (or if they “correct” them)? Is there some kind of correction function?
3 Likes
Let’s say you have an ID on something, but you now think the ID is incorrect. If you have a new idea for a more accurate ID, add that new one and your old one will automatically withdraw. If you don’t have a new one to add but you still want the original to go away, it is also possible to withdrawn the old ID by hand (on www.inaturalist.org or the Android app, but not the iPhone app.) *
If you do nothing, but your ID is indeed wrong, it will be difficult for your observation to reach research grade because at least three other people providing the correct ID will be required to outvote your incorrect ID.
If someone else proposes a different ID, using the “agree” button will automatically add that ID as your new ID and withdraw your old one. However the community would prefer you not use the agree button unless you’ve done some research into the taxa involved and come up with reasons why you agree; reasons beyond just “because the other person said so.” If you haven’t done research and don’t have any personal knowledge of the taxon, but you’re pretty sure the other person is more likely to be correct, than it is better to withdraw your ID.
*Edit: as the next comment below explains, it’s also perfectly fine to make your ID something broad like “Plantae” (plant) if that’s the best you can do.
11 Likes
Incorrect IDs should eventually become outvoted by correct IDs. If you put an incorrect ID and two other people come along and put the correct ID, the system will automatically update the ID and accept it, as long as their is a consensus of opinion of more than two-thirds. If there is disagreement, then the system will automatically ID the observation to a less specific level such as genus or family. For example, if there’s a photo where half the people say it’s a black bear and half say it’s a brown bear, the system will ID it as a bear (pending further votes one way or the other). If half say it’s a brown bear and half say it’s a rabbit (admittedly a very unlikely scenario!), the system will ID it as a mammal, as that is the most specific consensus.
If you have observations where you think your ID is incorrect, do not take down the observation. You should change the ID to the correct ID (your previous ID will automatically be withdrawn when you do that) or, if you don’t know the correct ID, just ID it to the best level you can. It is absolutely fine to ID as “plant” or “animal” if you aren’t sure more specifically. In fact this is absolutely what you must do if you don’t know more specifically, as wild guesses are not helpful. Hopefully others will come along later and help narrow it down.
6 Likes
Just in case the concern is over an incorrect ID supplied by iNaturalist, you can always override its suggestion. E.g., just replace the suggested ID with one you’re comfortable is correct (e.g., plant, animal, or even Life).
2 Likes
If you create an ID but you think it is wrong it is good practice to either change it or withdraw it.
To change it just scroll down and “Suggest an Identification” again. No need to touch your previous identification because it automatically gets withdrawn once you add a new ID. If you think your previous ID is the wrong species, make your new ID at the genus level. Or any level higher up. Any information you add might attract the attention of someone else who can help ID your organism. For example, when I’m completely confused by a plant I frequently make my id Dicots (Class Magnoliopsida).
To withdraw it, just look at the box with your identification. At the top right corner of the box is a “v” shape. Click on it then it will give you an option to edit or withdraw your id. I often use “edit” to add a comment about the observation.
Once someone else adds a new ID that is different from yours, iNaturalist changes the description of your observation to be where your two observations diverge. For example if you start with an observation of “Valley Oak” or Quercus lobata and I come along and add the ID of “California Black Oak” or Quercus kelloggii, then iNaturalist will change the description to be Genus Quercus. At that point, if you think there is still a chance you might be right just leave it alone. If you are fairly sure you might be wrong then withdraw your id. If you are sure the other person is right then change your id.
Assuming you left the observation alone and someone else adds another ID of California Black Oak, then iNaturalist will see 2 votes versus 1 vote and change the description to be California Black Oak (Quercus kelloggii) but leave the status of “Needs ID”. At that point if you think the others are right then withdraw your id and iNaturalist will recognize there are now two consistent ids and change the observation to “Research Grade”. If you are not sure the others are right then leave your id in place to make it more likely someone else comes along to help.
If a third user comes along and adds a third ID of California Black Oak, iNaturalist will now see 3 votes versus 1 vote and change the observation to “Research Grade”. At this point I don’t know if it matters if you withdraw your id or not. I don’t know if this affects the statistics on iNaturalist.
My comments above where I describe “how to” apply mainly to using the main website: iNaturalist.org. I don’t use the phone apps for identifications.
2 Likes
I’ve, unfortunately, seen a lot of incorrect ID’s that have persisted for months or even a year or more. It’s problematic when it’s on an ID that gets two agreeing incorrect ID’s, as those records often escape attention and then get marked as “research grade” and contribute to bad or low-quality data.
When I’m engaged in a discussion and there is an incorrect ID that is persisting in spite of my input, I usually respond by using the @-message system to draw in the attention of one or more experts to lend their voice and vote, and usually this fixes it (whether or not I am correct, if I’m wrong, I learn exactly why, in some cases, all of us are wrong and the expert corrects us all! In other cases, the expert will explain how and why ID is tricky for the particular taxon, and advocate restraint in ID, moving it up to a more general taxon, like species → genus → family → tribe → etc.)
Figuring out who is an “expert” though is tricky…you can use the leaderboards for ID’ing a particular taxon, but you really need to read people’s profiles, as many people have regional focuses. For example I am high on some leaderboards but people message me about a species in Europe, Africa, or Central America and I really can’t say anything about it at all because I really only know North American plants and even there I mostly know ones from the east.
In the long-run, though, I try to do my part to correcting the ID’s when I spot them.
One way I like to fix ID’s is to do a search for plants out-of-range, for species I know very well (i.e. well enough to confidently distinguish from other things they might possibly be confused with.) This inevitably turns up a ton of wrong ID’s, but because it’s a species I know well, I can actually check them, so that I’m not actually missing a correct report of a species out-of-range, as these do happen, and these reports are critically important as they give useful information about species expanding their range and/or colonizing new areas.
Another thing that I do is to just go through records of research-grade observations from time to time. I mainly do this by browsing photos, and looking for photos that stand out as looking different or off or potentially wrong. Some of these are not actually wrong, they’re just aberrant individuals, but other times I encounter obviously or egregiously wrong things that slipped through, like a sweetgum or plane tree mis-ID’ed as a maple, or more subtle things that are trickier to ID like red oaks identified as the wrong species of red oak. Some of these are so tough though that I just have to say “I don’t know.” but in those cases I add my vote to go up to a more general level, often disagreeing with the more specific ID if I feel reasonably sure it’s incorrect.
One thing that I find helpful for these “uncertain ID’s”, which can be really great for using iNat as a tool to further your own education, is to familiarize yourself with the different levels of taxa. For example, in plants, often I feel certain something is in the Asteraceae family, but I don’t know the family. Over time, I’ve learned to ID some more specific sub-groupings of this, like the Eupatorieae tribe, or the Cichorioideae subfamily…and then I can look at something and put it in that grouping even if I have no clue what genus it’s in. A lot of these groupings, I was actually unfamiliar with until I started ID’ing stuff on iNaturalist.
4 Likes
I really have nothing new to add to what has been said, but welcome to the Forum! The place where all your questions will get some sort of response.
This is also a useful method for cultivated plants. Since there’s no such category as cultivated and needs ID, scrolling the taxon photo page and scanning for anything that stands out can be less tedious than clicking through them in Identify mode. I order the taxon photos by “newest first” and “quality grade: any” and then scroll until I get tired. Only drawback is that it’s difficult to check all the photos of a taxon when there’s too many to do in one sitting, but no way to mark your place.
3 Likes
I need to make a minor correction to my previous post.
If the vote is 2 to 1 then iNaturalist would have still left the description at the genus level.
This is why I always appreciate an explanation or rationale for disagreeing votes. If mine is the “incorrect” one, I am much more likely to withdraw it if someone explains why theirs is correct instead. If you just disagree with me, with no explanation, well, I may not be convinced, and my “incorrect” ID may remain.
2 Likes
That’s reasonable! As somebody who makes a lot of IDs, I encourage you to ask for clarification when somebody else make an ID that you aren’t sure you agree with. It’s not a good use of my time to write up an explanation for every ID I make, but I am always happy to have a conversation with the observer or other identifiers!
3 Likes
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.