I have learnt to check what effect, if any, my ID has on the CID.
Did it do what I want?
Does it make no difference - right so - skip that in future.
Have I made the ID broader, instead of finer?
That CID algorithm is convoluted. Very.
I hugely doubt that iNatters rack up IDs for the numbers. There is no ‘reward’. They are better served on other social media with RewardS R Us.
I have seen several people talk on the forum (although I can’t remember where) about being motivated to ID specifically in the hope climbing up the leaderboard or competing with a friend. I will admit that I have done this to some extent myself so it is definitely a thing. I personally have no interest in facebook and the kind so this is really the closest thing to social media I have.
Over my time IDing on iNaturalist, my personal rules (which need not be yours!) have changed. My current rules:
Of course, I violate my own rules if I feel like it, sometimes with the excuse that I’m trying to help students in a botany class. And now that I’ve learned that “Reviewed” hides the observation from me but has no other effect, I use “Reviewed” a lot.
Thanks @sedgequeen for this perspective! It sounds like we have to take it on a case-to-case basis sometimes!
Hi zzravizz,
I’m one of the iNat users whose observations you have been identifying. I think it’s great that you are so enthusiastic about learning about nature. But, if I may be honest, I don’t feel comfortable with you IDing every single observation I make. How can one know so much about fungi after only a couple weeks?
I look forward to seeing what experts (or at least someone with a bit more experience under their belt) say about my observations: whether I am correct or incorrect with the ID I have suggested.
I have been in your situation and made IDs on observations before, which I wish I hadn’t! I’ve made incorrect IDs based on visual information. I suggest viewing people’s observations, which will help you learn, but maybe not making IDs on every single observation you come across. Every time I try to ID a mushroom I’ve found, I realize how little I actually know. And the world of fungi is ever-changing: experts are discovering new things all the time.
I hope this helps you with your question.
Yeah, fungi a lot of times can be hard, and sometimes getting down to species level can involve knowing what tree species they grow with, what they taste like, or what the spores look like under a microscope LOL
I’d suggest searching up some good online information guides, or even buying some ID books, if you really want to get in to fungi - for online resources, Mushroomexpert.com is fantastic, as is the bolete filter (for boletes.)
Bookwise, the bible is Mushrooms Demystified by David Arora, and then generally it can be good to pick up field guides for your area.
Also, generally, you only want to add an identification if you can independently confirm on your own. Sometimes I’ll personally only ID my own observations to genus with a species in mind but not wanting to influence other IDers and then add my species level confirmation later, but I try to only do this with stuff I have a pretty good idea on.
if I feel the CV (computer vision) is not good in a certain area, I’ll also push the ID into a more general taxon before posting, but a lot of times I feel like the ID represents the CV’s opinion. As long as someone doesn’t come along and blindly agree, I will leave the CV’s ID.
Yeah sometimes the CV is super obvious and then I’ll just put it at species.
Its hard with fungi though, there can be so much individual variation within a species, and that’s not even counting the look alikes - and CV isn’t going to go know if that Trichaptum species is growing on a hardwood or a conifer.
@mpearlstein and @lothlin (and @bonesigh) Thank you for your advice! It does help me to see how my IDs affect others, how I don’t have to ID everything and I can still learn from observations, and how to best be helpful and keep learning. I probably created a little extra work for myself, but I went back and fixed most of my Fungi IDs so I apologize for the mass amount of notifications just now. When I learn more, I can always go back and get more specific. It helped me to review also on all the different things I’ve seen.
I am still figuring out when its appropriate to be able to use the CV suggestions to help ID, but I think getting some more experience with fungi and plants will help me figure that out later.
Well, most insect people focus on specific taxa, so it is generally more effective to ID insects at least to order when possible. The bug and beetle experts, as a rule, aren’t going to be looking for observations in Pterygota.
There are people who specifically work on sorting observations that have ended up in Pterygota for one reason or another – larvae tend to cause confusion, for example, as do less familiar orders like Trichoptera – however, they’re not necessarily the experts who can narrow the ID further. So if you (generic “you”) do have a basic grasp of insect orders, applying that knowledge will help get the observation seen by experts more quickly.
Being aware of what you don’t know is a good skill to have, though. I’ve found that the times where I reach too far beyond my area of minimum competence are generally the ones I end up regretting. IDing to the lowest level you are confident about is a good rule of thumb to minimize errors, though it of course won’t eliminate all mistakes. I’ll reiterate that it’s also fine to skip observations where you don’t feel like you can help/contribute anything useful.
If this is a response to my comments in the other thread, honest mistakes are not an issue; they are part of the learning process – as long as people read their notifications and withdraw if needed. I put an appropriate ID on the bee mimics and hand them over to the fly people, tagging them when needed, and they do the reverse for the bees that end up in their pile.
I do get somewhat annoyed, if you will, by IDs that seem to have been made without thought – e.g., where it is obvious that the user just accepted one of the CV suggestions without even looking to see whether the suggestion made sense or looked like the organism they observed. This annoyance does not directly correlate with the correctness of the ID or rather lack thereof. I am equally unenthusiastic about rapid agreements where it seems unlikely that the user spent any time considering what they were agreeing to.
There are a few things I find are helpful when evaluating CV suggestions. One is the “similar species” tab on the taxon pages, which will provide an idea of how often the organism is misidentified and what some of those species are (and whether you can see the differences). Another is to go up a level or two – to genus or tribe or family – and look at a couple of things: First, what percentage of the observations are “research grade” in your area; if this is quite low, it suggests that the taxon in question is likely difficult and you should be cautious when adding IDs. Second, on the explore page for your region, click on the “species” tab to see other species in the genus/family and get a sense of whether they look similar and what the distinguishing features might be.
If you’re serious about IDing, it is definitely also important to find and use external resources (keys, field guides) for your taxa of interest. But experience helps, too, and judicious use of the material on iNat can be a good starting point to orient yourself.
Something I’ll do for difficult taxa, sometimes, is follow the taxon trees back to the highest intersection that I’m confident with.
Like if its plant that’s not in flower, and I think its probably in genus Solidago, but it could also maybe be in Aster, I’d probably just put it in Asterales or even asteraceae… not that I’m generally in the habit of trying to identify things in Asteraceae when they’re not in flower, because that is a futile road to madness.
In my experience, the couple of people who refine Pterygota IDs can do it with more accuracy, often higher precision than order, and less effort than it would take me to do so. But in order to do it they need those observations placed in pterygota in the first place, so I think it is helpful to tell beginning IDers that it is ok to just place things in pterygota unless they are pretty sure about a finer ID.
This is kind of what I mean; in plants I would pretty much never get annoyed at someone for clicking a genus-level CV suggestion. I also wouldn’t get annoyed at observers for picking species-level CV suggestions. I would however generally suggest that new IDers don’t pick species-level CV suggestions, because sometimes the observer will then just blinding ‘agree’ as a thank you, taking it to RG without any real thoughtful human intervention.
I think I could probably write down a way to mathematically quantify the fact that above some threshold % accuracy for the CV, boldness to genus is information-theoretically better, at least in plants. I guess the difference with invertebrates is that you are often expecting that the observation will never get to genus, whereas for most observations genus will be possible in plants.
Asteraceae is its own special beast!
I wish. I have a nightmare stored in URLs I never get to. It is … a Plant!
You can filter by taxonomy or location in the identify tab. I started by picking a few organisms I know and identifying those or places I grew up in and was familiar with what people might be seeing. Dont hesitate to identify to even something as broad as beetles or anatidae if the identification is even more broad, every little bit helps to get it in front of an expert doing what I suggested in the first part of the comment
Well, 92% of plant observations are already at genus or better, so the statement ‘most’ is justified on the whole over all observations. It would almost surely not be true that 90% of currently ID’d to class or above plants can be identified to genus or better. I don’t know what the percentage there would be.
Are you including cultivated plants in that 92%?
What % for wild plants?
Hi Ravi! Welcome to iNat!
I have been doing the same, honestly, and I am not as new. For example, if I see a skull obs, for instance (am in a skull and bones project) and someone IDed it as a ringtail lemur, I might agree that it is a lemur, but since I don’t have as much knowledge on discerning exact species, I will ID it as lemur sp. A pop-up will appear, asking me if I think it is a lemur but not ringtailed, or if I am not sure but I do know it is a kind of lemur. I’ll choose whichever I think is appropriate.
In my opinion, any contributions are helpful, and in my experience, a more knowledgeable person will add their ID and explain specific identifying features in a photo to me.
All-in-all, I don’t think it’s distracting, and I think that even a more broad ID can be helpful too. As you use iNat more and more, you’ll find ways to ID certain organisms better.
Hopefully I explained this all in a way that makes sense, lol.
Have a great day, and have fun with iNat! :)
Verifiable plants: 60,085,444
Verifiable plants genus or better: 55,394,223 (92%)
Captive plants: 7,290,948
Captive plants genus or better: 6,358,479 (87%)
South Africa specifically is at 93% for verifiable plants being ID’d to genus or better.
I think we spend a disproportionate amount of cognitive effort thinking about the hard ones, so it can give us the psychological impression that the hard ones are in the majority, when they are not. Basically the availability heuristic https://en.wikipedia.org/wiki/Availability_heuristic
Thanks for reminding me that I see the problem children, because I choose to focus on them. I will keep reminding myself of the 92% !!