'Needs ID' pile, and identifications

General
524

I have been working on getting Verbascum at species or genus level to the correct ID. In many areas, there is nothing else it could be, large parts of North America, most of South America, Indian sub continent, Japan, Korea,China. So I think a mass correct would be a mistake. I do look in areas that have other species they could be mistaken for and correct errors. And even in areas that do have other species, V. thapsus is often the most common by far, so many of the guesses get it right.

But I can’t keep up with all the rosettes and dead stalks identified as V. thapsus. I don’t know what can be done with them.

I also do a daily quick scan of all new Verbascum each day.

2 Likes
525

If you ID to Verbascum, I will look at it eventually. Verbascum

1 Like
526

Sounds like me. I go with plant if I don’t see a flower. I remember the diagram cross section of a monocot and a dicot from junior high school, but I don’t know how to tell which a plant is by looking at it.

527

So the other day I tried doing no specific place or taxon, just set the filter to ascending to take care of the ones that have been waiting the longest. I could only identify one or two per page, but I noticed that in the time it took me to do that and mark the rest as reviewed, the page count might stay the same or even go up. After many pages of this, I got discouraged because there was a bigger backlog than when I started.

We often discuss the value of iNaturalist data; but that value is reduced when a data point (observation) goes unidentified or stays at Needs ID. I think that it is a mistake to assume that uploading more and more observations is necessarily adding a lot of value.

Maybe not everyone can reach a ratio like mine – 428 observations uploaded to 20,324 identifications made for others, for a ratio of 0.02 uploaded observation per identification. But even if everyone just went with a ratio of 0.5, that is, make two identifications for every observation uploaded, we would very quickly reduce the backlog. This huge, seemingly hopeless backlog is because we have a huge number of users who only ever upload and do not identify. If you are interested enough in nature to be uploading, then you are interested enough to learn to identify something.

I wonder if some encouragement along these lines in the onboarding would help?

As it is, my feelings about this are such that, until further notice, whatever identifying I do, I will do in ascending order.

5 Likes
528

At the beginning of the year, I came to the same conclusion, that identifying twice as many observations as I was adding was a good goal. It didn’t take me long to catch up in terms of IDs compared to my observations (but note that I live in New England in the US, where there are lots of iNat users, many making observations of easy-to-ID species).

Sometimes I do IDs in descending order, just looking at everything people uploaded recently, and sometimes I’ll concentrate on one species, IDing everything back into the past, as much as I can. I am aiming for doing 60 IDs a day, every day. Maybe I’m deluding myself, but I think it’s making a difference, at least for those species I can ID.

As for value, I think there are two kinds of value here: First is the scientific value of a large dataset like iNat. For this kind of value, you are right that it’s not terribly valuable to have many observations of the same species from the same small area (a local park, say) within a few days in the same year. I keep daydreaming about a digital atlas on iNat of the flora of my state, Massachusetts, to a finer level of spatial detail than traditional herbaria-based county records. For that, people would need to go to new places more often to make observations. (Please talk me out of this crazy idea!)

The second kind of value is what iNat states as its goal - to connect people with nature. In this case, the more people are out looking at bugs and salamanders and wildflowers, the better, even if all they are observing are the same very common, well-documented species, year after year. But still, even for this second kind of value, making IDs is valuable, as it gives people validation (or maybe a correction) of their own ID and probably prods them gently to go make more observations the next day, the next weekend, the next month.

For both kinds of value, identifying is vital. I try to encourage people to make IDs, but yeah, we all need to figure out how to cultivate more identifiers.

12 Likes
529

And, by the way, thank you so much for all the identifying you do!

3 Likes
530

I feel your pain! I often work through things from the oldest first, and if a taxon has been done quite thoroughly before it means you’re wading through the hard stuff. (Though some taxa there’s loads of old IDable stuff). If you know the taxon well enough to make this judgement perhaps it is an opportunity to deploy the ‘no it can’t be improved’ function.

Usually when I do this and find it hard I also do a page or two of the most recent ones. I think this is actually of more immediate value to the community - especially encouraging new users to stay - if they get quick responses. The difference between an ID being made after 2 or 3 years makes less difference. It also just makes you feel better if the older ones have been a bit of a slog!

4 Likes
531

With 100 million records submitted to iNat and growing, there will always be a lot that will remain unIDed. Some new submitters will be frustrated their pics don’t get attention fast enough or at all, but that’s life. I try to ID about 5 records for every one I submit but that’s a small drop in a very big bucket.

3 Likes
532

Not sure how this is or should be communicated to new users but the chances of getting an ID are improved if you only submit pics that are reasonably in focus and preferably cropped to better show the subject. I’ll pass over any photo that requires me to do a lot of work trying to decipher it . I think this is where the app fails as it makes it too quick and easy to snap a pic and upload it immediately without the user being careful that it’s in focus or not. I spend a fair amount of time cleaning up my pics before submitting and if I think they won’t be worth someone else’s time trying to ID I don’t submit.

10 Likes
533

That is skewing what you see to scraping the bottom of the barrel, either too blurry. Or too difficult for most of us. Whatever you can clear is a major achievement.
Page count will creep up as the fresh stuff lands at the other end.

And yes iNat should encourage people who rack up obs, to give back by IDing for others. Identifiers are not automagical machines.

6 Likes
534

Sadly it will not move from 1% rule, both for observers and iders, and just seeing how we get millions of new obs, but iders number changes by dozens only is very sad.

2 Likes
535

It’s always been interesting to me that the people who help staff keep iNat running, who do the bulk of the IDing and problem solving, are not the target group according to the mission of the website. They are already dedicated and knowledgeable naturalists. The new and casual users are the target but how many of those will move into the dedicated group to help deal with the growth of the site? Probably not enough.

5 Likes
536

For the last month I had lots of thoughts on another thing, there’s a constant focus on getting new audience and never on power users, if you’re already an established user it’s expected that you will stay whatever happens and ready to wait years to get something back, to you think about ECE of ids: iding a thousand of observations from last-week created profiles will generate maybe 5 users who will then upload 100 observations if we’re optimistic, if you id a thousand observations of even one older user this user alone will be motivated to observe more than all those users combined, with higher chances of rarer species and less “defects” in those observations. Personally it’s especially frustrating when locally it’s advertised to visit new places to gather data, data that then is not looked at by the same people who wanted it.

6 Likes
537

I think one of the powers of iNat if you really are interested, is the community and getting to see ‘familiar faces’ in your region. I now know some great butterfly and mushroom folk in my area; and am able to contribute to their research. The county I do a lot in for caves, also has very little done above ground. Its very back-country, to the point of much of it being the kind of area you need to know what hollows NOT to go into or be met with shotgun to face from protective locals; and what hollows folk are friendly if you are friendly (or even which folk are friendly to ask and which to avoid).

Being on here posting stuff from that county has lead to more than one researcher contact me to look for things; and given me folk to report back to, some of who run state-level programs. So even though not all my observations come here (some these folk don’t need photos so long as they trust you can ID correctly, and half the time photos are hard to get, so those observations just get reported directly to those who asked) I am still able to contribute directly to those who do need it. But you could well say that data transfer, even if it doesn’t go through iNat, is only happening because of iNat.

8 Likes
538

@jasonhernandez74 – I find that the observations still “Needs ID” from the early days of iNaturalist have problems that probably prevent identification (not showing the right features, cryptic species, out of focus, etc.). So when I want to work on the old things I prefer to search for observations with the current day of the year but back two or three years. May make the process slightly less frustrating.

2 Likes
539

As a new user who is learning one genus (with a backlog of 10,000 observations without id that I have not reviewed), I think new users can be encouraged, start with something widely distributed and easily identified like Purple Crown Vetch and just continue.

3 Likes
540

If you want to go through a taxon or place you can also sort by “Random”, just to mix things up a bit.

6 Likes
541

Well, it’s possible. In my case, iNat Forum pages and pop-Ups are often truncated. Similarly, the website upload pages get “stuck” in one spot so that much maybe out of view and inaccessible. I rarely use an actual computer.

542

For plant phenology it is easier to ID Unknowns by staying in the same month at least.

3 Likes
543

Monocots have parallel veins (spreading from the base of the leaf, never splitting or intersecting) while dicots have branched veins. That’s the main feature I use. There’s also a difference in the amount of flower petals, but I don’t remember what it is right now. Pretty much anything that looks like a normal plant is a flowering plant, though, even grasses and deciduous trees. Vascular plants are even more common, being everything except mosses and liverworts (I think).

6 Likes