As an observer, I really want to have my observations correctly and I am happy with any input from people who know what they are doing. I don’t do plants :) but please don’t hesitate put my observations in Animals if you are sure it’s wrong. Better “animals” than wrong! Obviously, finer ID is always appreciated - but something is better than nothing.
8 Likes
When I’m going through Araucarias, sometimes I see a conifer that’s clearly not an araucaria. I have no knowledge of other conifers, so I put in a ‘conifer’ ID.
11 Likes
Moderator Note:
I removed the links to observations in the initial post here as they could be perceived to be calling observers/identifiers out for their work on iNat. Please avoid posting links to specific instances of mistakes/behavior on iNat itself. Thanks.
17 Likes
If you disagree iNat select the highest common taxon of the suggestions. This might be life if disagree eg an animal or plant with a fungus. But if you disagree plant with another the result will be some plant taxon eg at family level
When I have to disagree with an id, I always explain why I disagree. I usually have to go up a level or two to give an id I am sue of. I feel the explanation is very important for several reasons:
(i) to help the previous id’er understand; and
(ii) to educate other people who look at that observation.
I have learnt a lot from such comments!
4 Likes
Many people have already replied with proper explanations, but I’ll just chime in as an iNat staff and say that the choice to not have a “Disagree” button is intentional and is almost certainly not going to ever change. ID contributions should be constructive, which is why you need to add something to the community taxon rather than just disagree with it.
38 Likes
There are two ways I see that you can do this. One is to simply identify it as a Plant. This will require you to type in “Plant” (or, like, “Pla” probably) and then select it from the drop-down, which is more than one click, but not a lot. The other is to automate this procedure with the metadata tool extension, which you can search for in these forums, and which will allow you to create a button that identifies things to “Plants” with only one click.
Actually, no, there’s a third way: You can click “reviewed” and move on. (This may not suit your purposes, but it’s an option, and it takes one click. Or type “r” if you are in identify mode.)
While it is not wrong, I would very politely request that people not choose “life” unless that is really the best that they can do. If you know it’s a plant, at least call it a plant.
5 Likes
I know this isn’t one click but less mouse moving, if you are in identify mode or an a observation you can disagree with zero mouse clicks by pressing “i” on your keyboard and then typing in a higher taxon, pressing “down arrow” to highlight the taxa you want then press “enter” to make the identification. You can then press the “left or right arrow” to proceed to the next observation.
Here is the list of the other keyboard shortcuts
I often find them useful, for my workflow it is more efficient than a trackpad or mouse
As said by others, you don’t need to identify to species to disagree. If you are as knowledgeable of a botanist as you claim you are I expect that you would know common orders and families are, simply add an id to the limits of your knowledge. As already said, putting dicot or angiosperm is good enough and better than plantae or tracheophyta. I suspect you can tell when something is a dicot or monocot.
If you are concerned about time, I timed myself with the keyboard shortcuts and disagreeing at a leisurely pace took 5 seconds
4 Likes
Disagree
I disagree with this take, and by writing this response I’m also providing a concrete example of why I disagree with it.
The goal of iNat is to get people to engage and to do so in a positive manner, both with nature, and in general with each other. While the ‘agree’ button is, in my opinion, a bit too easy to click at least it provides some positive engagement, and if it’s used responsible is a benefit to identification. A ‘disagree’ should, in my opinion, should require a far greater degree of engagement than simply a click. This is part of why when you provide an ID that is of a higher level classification there is a question that pops up to ask more about your decision.
Just having a single option ‘disagree’ button opens iNat up to a lot of misbehavior and makes it more like Reddit than it ever should be.
I vehemently disagree with the idea of adding a ‘disagree’ button, and explaining why here provides a clear-cut example of what sort of engagement should be involved if you want to disagree with someone’s identifications.
18 Likes
let’s imagine a simple quiz with a pic of a pyrus tree.
- what kind of tree is this?
a. orange
b. elm
c. ficus
d. pyrus
e. other
someone answers “c”. and for 6 years nobody grades their quiz. for 6 years. like, who’s the teacher? where did they go? did they get fired? are they taking a sabbatical? where are all the other teachers? how tall is the stack of ungraded quizzes?
i come along and see the quiz. i’m not a teacher. i’m a researcher. my area of research is ficus, so i know for a fact that the answer definitely isn’t c. is this information useful? well, it’s more useful than 6 years of absolutely no feedback or engagement of any sort. a little help is better than absolutely no help. a little hint is better than absolutely no hint. a little engagement is better than absolutely no engagement. a little feedback is better than absolutely no feedback.
i lied when i said that i’m not a teacher. here’s today’s lesson, always make it easier, rather than harder, for people to share any of their knowledge.
as someone pointed out in my thread on optimal recruitment, nobody here is getting paid. if we’re all teachers, then we are all unpaid teachers. we’re all volunteers. everyone here is intrinsically motivated to share their knowledge with others. a disagree button wouldn’t change this.
the issue is, if there are too many completely ungraded quizzes, then maybe intrinsic motivation isn’t cutting it. maybe there’s a reason why real teachers have to be paid to grade quizzes all day. i’m not saying that we should be paid, but it’s entirely possible that inat is greatly underestimating the usefulness of extrinsic motivation.
1 Like
If it’s too time-consuming to figure out a higher taxon, perhaps just comment “I don’t think it’s a ficus” or similar?
13 Likes
This is why research grade exists, proportionally very few observations make it past and are incorrect. If they are incorrect most often due to wrong species but correct genus.
You complain about the quality of the data but it is a data set that is freely available, researchers are responsible for their data and should vet it, if you have problems with it you can fix it yourself.
In your example the correct thing would be to select option e meaning none of the above. Even if it was actually and elm, maybe it takes another 6 years to get to species but at least it isn’t misidentified. Although I really do not see the value in this example as that is not how “identifying” (labeling an observation with a taxon) on iNat works.
you say always make it easier but the current method is easy
You are correct, a little engagement is better than no engagent, I don’t understand your issue with leaving a disagreeing id for a higher taxon eg: “tracheophyta”/“angiosperma” etc.
As far as I can tell it accomplishes your goal of letting the observer know it definitely isn’t what they originally id’d as, and counts as a little help/hint/engagement/feedback. I would like to hear your argument against doing so if you are still conflicted.
If you are still against that, leaving a comment saying this is not ficus, or tagging someone who has the time and willingness to identify it are other solutions.
I am even sacrificing myself, please tag me in any observation that you come across that has the issue you are complaining about and I will try to fix it.
As already said by iNat staff ID contributions should be constructive disagreeing without providing an alternative goes against the spirit of the platform.
One last note, @epiphyte78 to find observations you might be interested in but don’t have an id but are potentially ficus you can use the unknown projects (more info) for moraceae family which includes ficus:
https://www.inaturalist.org/observations?project_id=unknown-moraceae
You might also like:
you can use the url search parameter ident_taxon_id= shows observations to which a particular ID has been added (even if that is not the consensus ID) (more info) an example with ficus sycomorus:
https://www.inaturalist.org/observations/identify?quality_grade=needs_id%2Ccasual&ident_taxon_id=340164&lrank=complex
7 Likes
If you hard disagree at plant
iNat’s CID algorithm quietly insists that you actively disagree with any future ID within the plant kingdom. That is not expected behaviour. And absolutely never intended by the ‘plant’ identifier. I have battled Ancestor Disagreement since I fell over it for the first time.
https://help.inaturalist.org/en/support/solutions/articles/151000173076-what-is-the-community-taxon-
3 Likes
P
will give you
Plantae
as the first choice. It is about 2/3 of iNat’s obs after all.
And a final PS to the original question.
Use a text expander to leave a comment.
Useful info to the observer, and more importantly, to all future identifiers.
This is not a Ficus
4 Likes
I appreciate a good process observation. Kudos!
4 Likes
As others have pointed out, bumping the ID back to a higher taxon is the way to note that it definitely isn’t c. It takes perhaps 2 seconds longer than a single click, and in the time it’s taken you to write a convoluted forum post, you could have reviewed hundreds of ficus observations.
You have many prolific iNat identifiers telling you they aren’t in favor of a disagree button (and they’ve all granted you the service of providing an explanation for why they disagree), but…
You are just one of many other researchers here on iNat, myself included. I am grateful to the people that take the extra 2 seconds of their time to help me come to the correct ID; why would I not want to provided that same service in areas that I am knowledgeable in.
13 Likes
That’s why teachers don’t just say, “you’re wrong on this”, but they also provide the correct answer.
18 Likes
25 posts were split to a new topic: Questions about how ID disagreement works
Interesting, I was thinking along these lines the other day when someone made an ID on one of my observations which I thought was incorrect. I disagree, I shouted to myself! (not exactly what I said to myself), we need a disagree button! I looked at the person’s bio and understood where they were coming from. Then I studied the species which they had named and, yes, if I was in that species range I may have made the same ID, so I learned something. I then wrote a note to the ID’er and said why I believe it was wrong. A little later they came back and said they had not looked at the map and they corrected their ID.
However, along these same lines, I was wondering if there is some way to use Bayesian statistics to further improve IDs. I know that iNat does use some location features in suggesting ID’s, could these be weighted and query the person making the ID when the species is out of range or there is some other obvious error? Also, I do appreciate people without a lot of experience making ID’s - that is where we all started. It took me a long time to learn to not use iNat’s species level suggestions on some species and to use other information to assist me and I thank those who helped me learn this for being considerate. Is there a way for iNat to also use experience or expertise in defining whether an ID is valid - maybe ranking the quality of an ID? I always appreciate when someone steps in who has an abundance of ID’s in the family or genus. I also very often check the bio of people I don’t know and greatly appreciate ID from those working on a post-doc in the area - said somewhat facetiously - I also appreciate the many many amateurs with niche areas of expertise.
A long winded answer saying no to the disagree button.
3 Likes
Other people have had similar suggestions in the past. They are unlikely to be adopted for several reasons.
-
the computational requirements for asking identifiers “are you sure” before submitting an out of range observations are significant. (It would also have the potential to be annoying, especially with poorly documented taxa or taxa in places without many iNaturalist observations—if I was a local identification expert getting started on iNaturalist, I would be annoyed by an “are you sure” popup on many observations because they’re “out of range. And sure, iNaturalist could invest time and money into making the model better, but “improving” a model always comes with tradeoffs, and iNaturalist already has a solution to the problem you’ve noticed).
-
the newly released geomodel tool already addresses this issue (identifications are “scored” on a geomodel, and observations outside of the predicted range for a species can be filtered to and corrected, if needed).
-
interestingly, sometimes the computer vision is ahead of the scientists. The CV consistently suggested Potentilla hebiichigo as an identification in North America, where it is considered absent by all authoritative sources. This lead some iNat identifiers, myself included, to “correct” these “incorrect” identifications, which turned out to be correct upon further study. You can read more about it here: https://forum.inaturalist.org/t/potentilla-indica-vs-potentila-hebiichigo/57037. But the point is, a popup like the one you’re suggesting would have made this less likely to happen.
As to your suggestion about ranking identifications based on the identifier’s expertise, it has been brought up on this forum several times before, and has been strongly rejected for several reasons. 1, there isn’t actually a good way to measure expertise. Many top iNaturalist identifiers of a taxon have no expertise and may have many misidentifications. Many distinguished taxonomists have rusty field identification skills. Even if we did find a metric that worked, it goes against the fundamental mission of iNaturalist. The iNaturalist organization has made it clear on multiple occasions that everyone is on an equal playing field, from a platform standard—no one’s vote counts more. This may be annoying to some, but I think it’s brilliant. Experts are too quick to pull the “I’m the expert” card, and forcing them to provide evidence like the normies is certainly a good thing (and probably keeps iNaturalist one of the nicest places on the internet).
5 Likes