How to correct thousands of misidentified observations uploaded by a class?

[sedgequeen]

There is an e-mail behind each account. Therefore, it’s likely that staff can contact the project leader, even if he/she is not responding to messaging through iNaturalist. (Probably not worthwhile in this case.)

[sedgequeen]

I wonder if we should brainstorm a recommended message to project leaders for accounts with problems like this. Probably not on this thread. Maybe even in Educators? A message that would have to include, maybe start with, “Please do not delete this project or the associated student accounts!”

[sedgequeen]

@jurga_li

As it is on the same subject, I wanted to point out another similar project with similar problems

That project seems focused on South African organisms, mainly plants. The few I can ID are mostly right. Most of their Needs ID plants are mostly South African endemics ID’d to species, though, and I can’t judge whether they should have stopped at genus. They have at least used a local field guide. I added “Usnea” and “No, it can’t be improved” to some of the lichens.

[paul_dennehy]

Might it be useful to have an option when creating an account to specifically create an “educator account”?
This could be something like an account that, when created, funnels the user into a sort of onboarding process to place the information in the Educator’s Guide front-and-center as they start to use the platform. Maybe even automatically add the new user to a “sample project” so they can see what a well-designed project looks like and get connected to other educators in the project immediately. Or have an option to fill out a “classroom project template” to start their class project, that looks different from the typical project-creation page and tailors the project more to what we’d like to see from a classroom project- i.e. automatically include a copypasta set of best practices for creating observations in the project description, or perhaps have a built-in link for the educator to follow that says “click here to review your students’ work!” that takes them to the Identify tab filtered to the project’s observations. If they see how things look when they’re being done right, it might be more beneficial than only being contacted once they’ve started their assignment to hear what they’re doing wrong.

[upupa-epops]

I’m not going to link all the previous topics with similar titles “help! a big new student project!” but if you search for them you’ll find them and all the associated discussions parallel to this one.

@AdamWargon

And iNat might develop a sandbox version for students

There is a feature request for this: Add a ‘sandboxed’ project category for educators having students use iNat - Feature Requests

@ciafre

Could there be a popup during the project creation process to the tune of “hey if you’re an educator and this is the first time you’ve made a project for a class, here’s a resource for setting up students for success and getting the most out of a class project” with a link to the educators guide. Or if not a popup, then something else that would be difficult to scroll past without acknowledging while making the project. This wouldn’t help the educators who don’t us a project to track their students’ observations, but it might have helped here.

And for this: Provide more guidance when creating new projects - Feature Requests

Also, a feature request for a separate kind of student account was turned down, but see Tony’s review of the conversation: Create Separate Accounts for Students Assigned to Use iNaturalist - Feature Requests

Also somewhat related: Addition of Teacher Friendly Features To Traditional Style Projects - Feature Requests

[DianaStuder]

Cape Town to Antarctica - I get the full benefit of the first wave. The Geomodel anomaly is picking up the confused ones.

Then there will be sea birds and such.
Then Antarctic sp.

[madisong]

I also had to download iNat for a class

[jnstuart]

@sambiology

I see that lots of the students have indeed deleted their accounts (and hence, deleted the ID’s that loads of folks had added to their sometimes good observations).

This would seem to be a “flag” for any identifiers who would devote time and effort to helping with a school project that the records along with the project itself might just get deleted at the end of the semester. Does not encourage me and perhaps others to participate in IDing and cleaning up such efforts.

[sedgequeen]

A lot of the student observations from this project are still on iNaturalist. They can be found by using the Identify feature to find observations in Norman, Oklahoma. Nearly all of the most recent ones from this area are from this class project, and they’re not hard to recognize once you’ve seen a few of them.

9 species of grass (so far) identified as Indiangrass (Sorghastrum nutans). A minority of them actually are Indiangrass.

[GothHobbit]

@paul_dennehy

Might it be useful to have an option when creating an account to specifically create an “educator account”?
(snip)
If they see how things look when they’re being done right, it might be more beneficial than only being contacted once they’ve started their assignment to hear what they’re doing wrong.

@sedgequeen

I wonder if we should brainstorm a recommended message to project leaders for accounts with problems like this. Probably not on this thread. Maybe even in Educators? A message that would have to include, maybe start with, “Please do not delete this project or the associated student accounts!”

I would be happy to assist with both of these. I have a great deal of experience as a Speaker to Educators. Although my native dialect is Elderly Southern Academic, I’m pretty fluent in everything from Head Start to Principle Investigator.

[AdamWargon]

@upupa-epops

There is a feature request for this

Thanks, Caleb! I need to spend more time studying feature requests . . .

[Italopithecus]

@ciafre

fewer than 1 in 10 seem to be correctly identified

When I observe similar cases in my country I wonder what educators/teachers think the students will learn from the experience with iNat.

@ciafre

One of my biggest concerns is that the creator of the project (presumably the professor or a TA) has just 2 observations

This is possibly the real cause of this issue. Users who clearly are not aware of what iNat is can involve a rather large number of students who can upload a huge number of observations that can be barely curated by few other users. I think that iNat should be useful for users but not in this way.
I think I won’t be popular but, in my opinion, only “expert” users should be allowed to create projects of whichever type. 500 observations could be a reasonable threshold.

@ciafre

Has anyone proposed a way to flag school projects to temporarily block additions until the educator/project admin has a better understanding of how iNaturalist works and commits to curate the project better?

This could be ideally right but could be frustrating for the students to see that they cannot add observations to the project and, after all, will not prevent from uploading observations.

@ciafre

Over 150 students have contributed to the project, so I don’t know how many more there might be before winter break.

As regards, I have developed the idea that using iNat with students in a proper way is feasible but only with a small group of students (not more than five), with an expert user escorting them and without giving the experience the nature of a competition.

[spiphany]

@neylon

It would also be nice if people couldn’t make projects without a minimum amount of observations and ID’s, say, 100 observations and 500 ID’s.

@Italopithecus

only “expert” users should be allowed to create projects of whichever type. 500 observations could be a reasonable threshold.

The number of observations and IDs is not a very good measure of how well a user understands iNat. There are plenty of users with hundreds of observations who still seem to struggle with some of the basic functions, or who have explored iNat very little beyond uploading observations using the app. There are users who have only made a handful of contributions but have spent a lot of time exploring iNat and how it works.

At the risk of repeating myself, I think this is a problem that needs to be addressed by better onboarding offerings. I would consider it reasonable to restrict the ability to create projects until a user has completed a relevant tutorial.

[DianaStuder]

A little quiz, with a few trick questions, to be sure they understand how to iNat. Before they bring us grudging students.

[sedgequeen]

@ciafre

so far 525 observations have been identified as indiangrass (Sorghastrum nutans); fewer than 1 in 10 seem to be correctly identified

True. Really bad – for the grass. At least 9 species, with the name applied one of the less commonly posted. However, it’s possible that all the hundreds of Solanum eleagnifolium are correct, as well as most of the Maclura. Of course, one could debate the usefulness of hundreds of observations of the four species in one place in one season. But the purpose of this project was probably to get the students out there in nature, and that’s useful(!) though it certainly could have been handled better.

[jasonhernandez74]

Although I don’t know much about Oklahoma’s biodiversity, I actually could help with a few. Where someone had corrected the “Indiangrass” to Common Reed, for example, or Cattails, or Genus Andropogon, I could add my agreement. Similarly, where there were Osage-orange fruits visible, that was a straighforward agree.

What ticked me off is that mixed in with all this nonsense were observations by a sincere user with a filled-out profile. If I was that user, it would frustrate me so much to have my sincere observations buried under this avalanche.

[belgianferns]

@spiphany

The number of observations and IDs is not a very good measure of how well a user understands iNat. There are plenty of users with hundreds of observations who still seem to struggle with some of the basic functions

I only discovered the “Identify tab” after about 4000 IDs… before that I just subscibed to topics and clicked on them on my feed.

[sedgequeen]

I contacted the sincere user. He’s now helping clear up the mess. We could use your help on these observations because it will take 3 identifications to overcome the initial wrong ID.

Also, do you happen to know Solanum eleagnifolium? There are lots of observations of it and I don’t know it.

[taylorse]

@sedgequeen

it’s possible that all the hundreds of Solanum eleagnifolium are correct

They are not, I corrected one which contained no visible nightshades at all. (…but that was just one. One mistake out of hundreds is pretty good.)

Alas, I can tell that it’s a nightshade, but I can’t get them to research grade because I don’t know that type of nightshade.