Are there too many new observations to identify?

In the UK I’ve noticed that the number of needs id observations in identify has gone up from under 200 thousand a early this year to well over 300 thousand when I started writing this post* and is 450 thousand now. I think the numbers are still going up fairly rapidly. I’m concerned that there are so many new observations that many won’t be getting identified and that this may put people off using the site.

I’m not really sure what if anything can be done though.

* I just discovered it as a draft.

5 Likes

There are too few identifiers.
Also, there are possibly too many unidentifiable observations.

22 Likes

Each day we’re getting more new observers, much more than new identifiers, so it’s logical that the amount of non-RG observations is rising. I don’t see anything too rapid, but it’s probably a geometrical rise and expected.
Probably new graphics with statistics as those that were posted some months ago could help.

7 Likes

This is a big part of it - a lot of users’ observations of a blurry bug (including some of mine!) are never going to make RG. The only thing to do is to mark them as “can’t be IDd further” to remove them from the queue, but ideally only if you’re enough of an expert (or the image is awful enough) to be sure that someone who knows more about the taxa can’t ID it either).

12 Likes

There might be a seasonal component to this too. I, like I suspect many others, are more focused on making observations during the spring and summer months. In the winter, I’ll spend more time identifying. Since we tend to have more observers in the northern hemisphere, more of us are in the “making observations” mode now.

I also make a point every week or so of going through a bunch of observations in my state from accounts that were created in the past 2 weeks. Even if I can’t ID something of theirs, I’ll make a comment welcoming them in an attempt to increase the chances that they’ll continue to use the site.

19 Likes

People are often quick to point out that number of needs ID observations usually does down in northern hemisphere winter, presumably because people stuck indoors by cold weather ID more and/or observe less. That’s relative though and overall I’m inclined to say you’re right, the numbers generally increasing and many newbies are already turned off by the lack of prompt response. There’s usually a bump in the spring because of the CNC and pretty wildflowers in general. For me I notice a bump whenever the local schools assign kids to use the app, as many did in the last few weeks.

6 Likes

I’m right there with you on everything you said. I’ve noticed things have been slower too and hoping it’s because seasonality.

For me I always try to maintain a 1:1 observation:ID ratio (or better), with a focus on my state/county. I’m no master naturalist but between birds and common plants it’s pretty easy to do. For others who know less about nature this may be difficult but it’s a fun little mental game to keep yourself identifying.

8 Likes

It takes almost no expertise at all to take a picture versus identifying an organism with any certainty. That’s just the nature of it. But the more data the better I suppose, not every user will ‘catch’ and become a naturalist (unfortunately).
I feel like the CV actually provides some amount of engagement and a tentative ID for people to follow up on if they wish. Though it has its issues.

12 Likes

but even so, just being interested enough to be on iNat, we can each tackle some part of IDs.
I can see as I chew down from recent, that there are IDs swimming below the cream we skim off the top. People who never answer … is it the bee or the flower?
Since there often is not enough info for an ID in the picture, if the picture is blurry as well - mark as reviewed, and pull up a fresh page.

1 Like

but even so, just being interested enough to be on iNat , we can each tackle some part of IDs.

Very true. I know of other users that weren’t aware that anyone can make identifications. I wonder if there’s any use in a notification from the app along the lines of “Congratulations, you’ve made 20 observations on iNaturalist! We’d love to have your help in making identifications too. Here’s a link to getting started making identifications.”

7 Likes

I would be wary of encouraging that. Rather let people learn as they go, and ask questions for the next step.

5 Likes

Well, they can still be given a coarse ID of arthopod or insect (depending on just how blurry they are). They’re still at “Needs ID”, but it’s better than nothing.

3 Likes

That pretty much captures the issue. There will always be a backlog of records needing attention because the time and effort invested in providing an ID (if the photo isn’t good or the organism is in a difficult taxonomic group) or fixing some problem with a record will usually exceed the time the submitter put into posting their record. Unless the record has an excellent photo and it’s an organism that a lot of iNatters are familiar with, a very rapid ID is unlikely. I suppose new users should be made aware of that fact.

7 Likes

I’m one of those disappointed newbies y’all are talking about. I posted 10 observations to iNaturalist back in June and July. They are all very clear photos, and for 9 of the 10 observations, I already knew what the species was and set it for the observation. I was just hoping for confirmations, or even a “nice photo”. But to this day not a single one of my observations has received a single comment or ID suggestion. So I haven’t posted any observations to iNaturalist since then. I know y’all say that’s just the nature of the platform, but on BugGuide I generally get a comment or ID suggestion within a week of posting. Maybe there are some improvements or incentives that could get more people doing IDs.

8 Likes

True. And many new users don’t care about genus and species. They want to know Beetle, Aphid, Weevil, Orbweaver, Coral vs King snake, Blue Jay, Coneflower.

I was going through the Casual observations because I’ve noticed that the new users snap whatever they see first: in a schoolyard, out the back yard, on an end table, pets. If they properly use the Casual setting, almost no one looks at them and they go away, presumably disappointed. Or they forget to add the date and/or location (presumably using an older camera?) and it’s marked casual. Also, never to be identified. I can at least tell them it’s a frog, sunflower, spider, daddy longlegs, or squirrel. And if it’s a really cool observation, I tag a leaderboard identifier and ask them to look at it. (Sorry, leaderboard identifiers.)

I posted a plea in Nature Forum last week asking for Casual identifiers, but it didn’t get a lot of traction. There are an enormous number of new observations. By the time I get finished with the unknown observations from my area, there’s a new page or two generated.

9 Likes

You have posted some really nice photos and jumping spiders are charismatic enough where I normally would have expected someone to have encountered them by now.

I wonder if having the location marked as ‘Private’ might have something to do with it. I obviously can’t speak for everyone, but I periodically go through observations from South Korea looking for ones from new users to help with identifications and to offer a welcome message. If ‘Private’ locations don’t show up via a search it could be that people who might have seen your observations through a geographic search aren’t getting them as results. If that’s the case, the people most likely to come across your observations are those browsing through all the observations uploaded to iNaturalist, where yours are going to be just ten of ten thousand.y

Sorry to hear you’ve had that experience. Finding someone else in your general area and interacting with them (identifying their observations, commenting, or following them) might be one way to get someone looking at your observations without changing the geoprivacy away from ‘Private’. You might also consider tagging one of the top identifiers for Salticidae in your state for the observation you have that’s not as species level. That does require a little extra work but I hope you’ll find it rewarding and get positive results.

19 Likes

I agree the private setting is likely a big block on getting id’s . It means other users have absolutely no clue where it is, so unless it is completely visually distinctive, it is very hard to assess.

Try using and/or changing to use obscured. People still wont see the exact location, but rather s random spot with a 22km2 box.

Spiders do however tend to have a relatively low ID rate be as many require very detailed at times microscopic photo detail to differentiate.

15 Likes

I have to agree with @whaichi. I can’t speak for spider IDs, but if I can’t see where the record was obtained because it’s Private, I will skip it … that’s assuming I see it at all since it’s not going to show up in those places I focus on. A lot of IDs are made in part based on what we know should be in a particular geographic area. To me, it’s not a record if the reviewer can’t see all the pertinent data that makes it a specimen record and not just a nature photo.

Great photos, by the way.

11 Likes

If you’re just trying to hide the exact location (like your house or a species likely to be poached), you can use Obscured instead of Private. That gives the identifier a general idea where it is without revealing the exact location.

12 Likes

This is the opposite of my experience. Occasionally it does occur that somebody doesn’t know but does care about what type of organism they’re looking at. But the vastly more typical response to a higher-taxon ID is “well I know it’s a spider, but what kind?” And of course the perennial “genus species?” either as the observation notes or as a comment.

And the identotron’s hyper focus on providing species ID’s even if they’re not plausible continuously trains new users to expect the most specific ID’s. The countless subspecies related discussions of late are merely the next iteration of this process.

I’ve said this before, but with the machine taking over, generalist identifiers are no longer needed. You can help sorting the unknowns into kingdoms, and many people do that. Or you can be the borisb’s, k8thegr8’s or brandonwoo’s (*) of the community providing tricky case by case ID’s. Identifying common organisms to family level is completely obsolete.

(*) not @-tagged to not call them out - I can think offhand of maybe 15-20 people who I would count in this group, a few more if I look them up

10 Likes