Recruiting more identifiers

i’m not sure what the disconnect is here, i am not saying we (not they - i am one) are primadonas. I am saying from my experience we don’t want more ID weight (decided by who?) or flashy badges.

Ok, here is what i think we want: A free, powerful communal data management tool. Put some energy into adding the features on iNat that those of us who monitor ecology will use! transect,plot, bucket trap methodologies compatible with the app. Not just the species list, allow some ecological data in with the plots. Negative data and survey tracking functionality. Ability to make secret observations (if that’s a burden, charge people a little bit, whatever). Ability to upload and save data and maps on the app. Basic natural community/ecosystem/wetland mapping functionality. Revele datasheets. And hey… is your taxa full of bad data on iNat? How about better personalized filters to CHOOSE whos data and what taxa to display on the newly upgraded maps? Don’t like Bob’s observations or Sue’s IDs? Turn them off? Most of my ideas come from the botanical world so if your specialty is another taxa i’m sure you will have more ideas too. And none of this is super hard, it just hasn’t been done because the goal was to grow the ‘citizen/community’ aspect of lots of ‘amateurs’. Do what BIOTICS doesn’t do, what Arcmap can’t do without freezing, what you’d do with R if you thought it was worth setting aside enough field time to learn it really well.

Doesn’t that sound better than weighted IDs? You’re a field guy right? Wouldn’t you use that? Or maybe you have another idea that is better. I don’t know. But again i am not arguing for forced equality, just for avoiding having credentials forced on iNat from some external source.

Well, if you want to know what I feel like are the real problems here, all you’ve got to do is read the post you responded to. :-) And I didn’t recommend badges, so I have no idea where you’re getting that!

For what it’s worth, filtering out uninteresting observations is one of the places where I feel like egalitarianism really comes into play. Wouldn’t want to let people ignore the uninteresting users… of course, we’re still going to do our best to ignore the uninteresting users, it’s just going to be more work and iNaturalist is going to be less rewarding as a result.

sigh. i did read your post, you are assuming if i read it i will automatically agree with you? On that note I think i’m done with this part of this conversation.

People constantly suggest badges.

we emailed about 400 users last week who met certain criteria (mainly English locale & iNat.org site affiliation) and fell into two main categories:
-2000+ observations and fewer than 200 identifications for others
-500-2000 observations and 0 identifications for others

Huh… I have ~2,000 observations on iNaturalist but have provided ~11,000 IDs for others (as viewable on my profile). Is that not the norm? I don’t always have time to go outside and rigorously look for organisms, but identification is always accessible and helps me hone my skills.

4 Likes

That’s the kind of testimonial that we need for recruiting more of the same! Thanks for all those IDs

4 Likes

An attempt to summarize a few of the issues and existing and potential solutions for attracting accurate identifications.

New, uninitiated-to-iNat, expert identifier perspectives:

  • iNaturalist is just a toy, “Instagram for nature”, and not useful to me
    • A general “iNaturalist is useful” page? Expand/improve this one?
    • Using iNat as your data platform guide
    • Feature research or conservation wins that came from iNat observations (Tony does this on the blog/social media)
  • iNaturalist is full of bad data and a waste of my time
    • Show how they can help!
    • Provide public-facing stats on accuracy based on blind ID experiment with experts
    • Address data quality issues w/ computer vision (in progress to my understanding, such as only providing higher taxonomic ranks as suggestions for some cases)
    • Address data quality issues with too-quick agreement (…reputation system?)
  • I don’t have time to volunteer here
    • Show how iNat can be used for work
  • Not having 100% control over taxonomy
    • I don’t think iNat is likely to develop a free-for-all where people can enter whatever they want as an ID, though I would like to be able to at least sort observations by my ID, and
    • Have an easier way to export observations with my IDs as the labels
    • As observers, they can opt out of community ID
    • iNaturalist can deviate from the standard taxonomic authorities if needed
    • Users can use a project to export “project curator ID”

Existing iNatters who observe but don't ID:

  • Prompt to help ID
  • Not sure how to use Identify page
  • Lack of confidence
    • Cultivate a welcoming community where anyone can make a mistake without fear of being yelled at
  • Lack of knowledge
  • Providing ID tools
    • Easy way to ID from mobile
  • iNat website is slow
  • Feedback on helpful IDs
    • Highlight helpful IDs
    • Tony does frequently highlight identifiers in his Observation of the Week blog posts
    • Emphasize Improving and Leading IDs
    • Some very modest badges or automated acknowledgment of helpfulness could be useful for spurring additional activity, e.g. “Thanks! You just added an Improving ID for the first time!” or, when the community confirms it, “That’s the first time you’ve IDed SpeciesX!” (and not anything like “Congrats on adding 1000 IDs, can you get to 2000???”)

Preventing ID burnout for existing identifiers:

People who want to invite new identifiers:

  • Who to ask?
    • Smarter breakdowns of expertise by taxon/area
  • What to say?
22 Likes

Oh wow great summary, thanks. Were my other ideas about expanding inat function for surveys and such too off topic to make it in?

1 Like

If you write up a feature request I can link to it. ;) There is the Trips feature (still in development I think) to capture search effort and absence data.

2 Likes

it was just my meandering post above. I will try to distill it into some feature requests when i get a chance… someday? :)

1 Like

Non-video tutorials is a great point, and aside from personal preference would probably be very helpful for many of us who also have the “iNat website is slow” issue- I don’t really know if it’s the website that’s slow or my connection, but I do know watching videos bites sizeable chunks out of my monthly data block.

6 Likes

OK Charlie you have proposed some ecological data types that you’d like to see iNat capable of collecting and analyzing. Neat thought experiment, but I see a lot of incompatibilities. Some examples:

Transect data: The user travels a line and collects data at given points; users are trained to do spot IDs, so they don’t typically take photos of everything. But iNat deals only with evidence-based data like photos. Much slower, and who wants to be second-guessed by a viewer years in the future? Also transect scale can be anything, millimeters to kilometers etc, and iNat’s geo-precision is not good enough for that.

Quadrat data: In one common form, user estimates the percent cover of each of several species in a quadrat frame placed on the ground. Picture having seven or eight species, each of which requires a separate entry in iNat, by duplicating the photo and describing which species this entry is for. Not happening.

Secret data: I can see why some users would want to collect secret iNat data and might even pay to do it, but then iNaturalist gets into legal issues of assuring secrecy, and what if the secret data is something illegal, maybe not even biological? Let’s stay transparent.

Wetland delineation: I could almost see this based on veg, but hydric soils and hydrology data also are part of a delineation, and iNat doesn’t do that. And the veg data wouldn’t be useful without wetland categories, which vary by region for each species and are updated periodically.

I do agree with you that more ways to filter data with all kinds of Boolean searches would be good. I’d like that kind of expansion more than jumps off into specialized realms.

I’m envisioning iNat supporting these data types not being the end-all. With transects, it would not be hard to program a transect location especially if a straight line between two points. The idea would be to photograph some or all of the plants, which si pretty fast and could be valuable for a lot of reasons. In terms of those photoless observations one might have they will be casual grade but the observer knows the data and its uncertainties. In fact i’ve already entered lots of data if that sort to iNat. I’m not sure at all what you mean by geo-precision either but plant locations are typically not recorded within the existing transect, which is usually 100 meters long or something like that, but it depends on the methodology. I don’t see why iNat couldn’t be used for a transect, and i’ve done similar things.

It isn’t? I’ve done it a bunch of times (not always with the same photo but there’s no need). More than a few others have done it here too. It would just be nicer if there were more built in functionality.

Obviously anything can be hacked and it shouldn’t be used for anything illegal or super secret. That’s true for all internet sites. And if it’s data that isn’t appropriate for iNat (rocks or whatever) but also not visible to anyone else, and the person is paying for server space, it seems an odd thing to worry about. But… i am envisioning combining iNat’s functionality with something like NatureServe BIOTICS which again, already does this stuff, but is just harder to use and as of a few years ago at least didn’t have a good field portal.

yeah, the whole point was to add new features. Wetland mapping is different than delineation though. I wouldn’t envision iNat used for formal delineations, for reasons you mentioned and others, though it isn’t impossible, the GPS on a smartphone may not be good enough.

So yeah in short i am already doing this stuff i just think myself and others would really find a lot of power in iNat if these features were more supported instead of kind of being an informal, back-end use. It sounds like you aren’t interested or wouldn’t use it, but that’s fine. Lots of things on the site I don’t use too. I am interested if other people would use it though. Obviously if i were the only one interested, it would nto be a good use of dev time! But i think you’re really selling something short that’s already happening and could even more so. I personally think it would get a lot more ‘professionals’ involved.

3 Likes

I would make frequent use of surveys if they were supported in iNat. There are a lot of non-iNat projects that could be easily persuaded to switch to using iNat if such things were supported, and they would bring with them experts who would probably also lend some of their time to IDing observations of interest that were not submitted as part of their specific project. An example I’m involved with is the phytoplankton monitoring network. Right now you have to fill out a form and email photos to be verified, it’s kinda clunky. But what’s even more clunky is trying to add the data in those forms to iNat, which I tried doing with fields but then how do you group all the observations in the same sample or update the fields for all observations at once when something needs to be changed? It seems like the kind of thing that iNat is almost good enough to support, but not quite.

Expanding the capability of fields so you can do things directly in iNat like creating a plot showing the average measured length of leaves of some species of plant for each month of the year in North America would be really neat and also make it more likely that other projects would migrate to iNat. You can already do such things by downloading the data and using it in another program of course, but I doubt this gets done very often and isn’t something most people think about when they talk about what iNat can do.

It doesn’t seem like such things are going to be added to iNat anytime soon, but I have been thinking about making my own program to record my data but tie the IDs directly to iNat. Little third party things like that probably still wouldn’t be much of a selling point.

1 Like

yeah, if i had more time and more programming skill i’d definitely try that. It is open source after all. you might have to link to an external database too though which would be a pain, or else add a ton of various indexing fields that other people could change, etc

For plants, iNaturalist is not taxonomist-friendly at all because we are dependent on the observers to tag the cultivated plants to not see them, which they don’t bother to tag for most of the time.

Also there is no way to sort regional native plant list excluding non-native plants. A regional native plant list is something that plant taxonomists would want to work on and iNat does not support this.

1 Like

I’m unclear what you mean by this, any regional checklist can be filtered (if not sorted) by if the species is native or introduced. Of course the data has to be populated, but the capacity is there.

Here is a screencap of the Ontario checklist page.

OOPS - just realized that’s in Danish, and I forgot to switch over to English, but its the 3rd one down ‘Oprindelse’ is a translation of Establishment Means, and the options are Endemic, Native, Introduced and Alll

1 Like

What about in the observation page?

Use the URL qualifier introduced=false, e.g. native plants documented in Illinois on iNat:
https://www.inaturalist.org/observations?introduced=false&place_id=35&view=species&iconic_taxa=Plantae

1 Like

Absolutely.

All observations of species of introduced plants in Ontario (please note I’m not guaranteeing all species are properly entered yet)
https://www.inaturalist.org/observations?introduced&place_id=6883&subview=grid&taxon_id=47126

All in Ontario that are not introduced (which oddly is done by saying introduced = false)
https://www.inaturalist.org/observations?place_id=6883&subview=grid&taxon_id=47126&introduced=false

Looks like we were writing at the same time, one minor note / correction to what Cassi wrote, Introduced = false will get both cases where it is marked as Native or Endemic, but also if it is not filled in at all. So it is possible that an introduced species that has not been indicated as such will be there. For example the first record as I write this is a Butternut, but the status is actually not populated there, it is not explictly marked as Native

1 Like

In the observation page you can only filter with *wild.

But thanks for the URL. With the native AND wild filters, still… I’ll be seeing selected cultivars, because nobody tag them cultivated.