College courses and "iNaturalist for credit"

[Will Van Hemessen]

Hi,

I've noticed a growing trend of college courses requiring students to submit iNaturalist observations for credit. In general, I think this is a great way to recruit new users and collect large amounts of data. Unfortunately, I find that the quality of these observations is often low, and especially starts to decline towards the end of the term. I can speculate on a number of reasons for this: new users still learning the functions of the site; students scrambling to submit the requisite number of observations; students putting in low effort because submitting observations becomes "homework"; etc

The outcome of these courses is something that I imagine will be contentious, but which I think needs to be addressed: they end up flooding the site with hundreds of poor quality observations. A community college course in my area has 49 students who are required to submit a minimum of 10 observations for credit. This means that at least 500 observations are submitted to the site by this course each term and all of them need to be ID'd and vetted by regular users like me. When the quality of observations inevitably declines, it gets frustrating. I regularly see students submitting photographs with no clear subject and making no effort to identify a particular organism. This is, in my opinion, the result of low effort on the part of the students and is detrimental to the overall data quality on this site.

I'm not sure what the solution to this could be, so I'm opening this up to frank discussion. Let me say that I am very open minded when it comes to the quality of observations on this site: if there is an organism visible, I will attempt to identify it. I also believe that criticizing the quality of observations - be it the quality of the photo, the location description, or any other attribute - is discouraging and not consistent with the spirit of this community. But observations which are clearly low effort need to be addressed, especially when they come in large doses like what happens with these courses.

Thanks for hearing my frustrations and happy naturalizing.

Will V

[Charlie Hohn]

In short, as you know (since we discussed this too) i agree it is a problem. And it seems to be getting exponentially worse as iNat expands My two ideas:

-ramp up even further the effort to get teachers to curate. Maybe even delete projects if the teachers will not do so.

-harder but maybe more effective: create student accounts (based on using the site under 'duress' rather than on age or experience). student accounts would not be able to give each other research grade (to reduce them agreeing with each other when they don't really know) and would be filterable.

I'm sure others have other ideas. And i know it was discussed in 2014 too so there is another thread out there (which I just posted in so it's at the top of the groups now too). Thanks for bringing this up again though. It's definitely becoming worse over time.

[Scott Loarie]

Hi Will,

As long as the observations are of living things and are 'verifiable'
(e.g. have dates/locations/aren't of captive stuff and if there are
issues with these they have proper data quality flags) then its fine
for observations not to have good IDs. But if there are data quality
or suitability issues, definitely alert the teacher.

If its possible to locate the teacher's account (check for projects
associated with the students or their obs and see if you can connect
the dots via the project admin), I find that public pressure on the
teacher to do a better job following the 'Be the Steward of Your
Class's Data' tips here
https://www.inaturalist.org/pages/teacher's+guide is often effective.
If the teacher isn't responsive, feel free to mention me or other
curators in the conversation so we can pile-on the community
pressure.

Scott

> --
> You received this message because you are subscribed to the Google Groups
> "iNaturalist" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to inaturalist...@googlegroups.com.
> To post to this group, send email to inatu...@googlegroups.com.
> Visit this group at https://groups.google.com/group/inaturalist.
> For more options, visit https://groups.google.com/d/optout.



--
--------------------------------------------------
Scott R. Loarie, Ph.D.
Co-director, iNaturalist.org
California Academy of Sciences
55 Music Concourse Dr
San Francisco, CA 94118
--------------------------------------------------

[Tom Norton]

I agree with the comments of both Will and Charlie on this topic, and I agree that this issue needs to be given higher priority from now on.  Charlie's two suggested fixes make sense to me.  This past fall, I got involved in helping to review observations submitted by a college biology class.  I enjoyed the process, to the extent that I felt I was helping students to learn about the natural world.  Where it got frustrating was when the students realized that they could agree with each other's identifications, and thus achieve RG.  They were graded not on how many observations they submitted, but on how many of their observations became RG.  I eventually figured out who the teacher was, and sent some comments plus a link to the iNaturalist Teachers' Guide, but I couldn't tell how much good that did.

- Tom

On Friday, February 17, 2017 at 9:35:25 AM UTC-5, Will Van Hemessen wrote:

[Sam Kieschnick]

Good conversation on this.

This is something that I watch closely.  I've written comments on hundreds of observations that need some guidance.  In some cases, they are cultivated plants around campus -- sometimes, they are just poor images.   I've created a template of standard responses to these based on different circumstances -- feel free to use any of these: 
http://www.inaturalist.org/journal/sambiology/8013-standard-responses-to-observations-that-need-guidance

Most of the time, I just plead with the observer to ask their professor to be a steward of the data.  If I can find the professor's contact information, I send him/her a gentle email of encouragement and guidance (with links to the 'teacher's guide').

A couple examples of projects that I've tried to help out with...
TCU (Texas Christian University) has an urban biology project for critters around campus.  This one is properly curated by the professor:
http://www.inaturalist.org/projects/tcu-wildlife

UTA (University of Texas at Arlington) had a broader project for any and all observations, and it was not as properly curated by the professor:
http://www.inaturalist.org/projects/biology-of-texas-at-ut-arlington

However, even the one that wasn't properly curated by the professor was still quite successful.  Some of the students continue to make observations outside of the class project, and some of them are extremely high quality! 

All in all, I think the more that we comment on the observations and give guidance, the better the data gets.  Early on, my observations were quite poor, but folks continued to give me guidance, and I improved the quality of my observations.  I think this can happen with students (and with professors that use this tool!) as well.  Just my two cents.

~Sam

[Will Van Hemessen]

Sam, these standard responses are fantastic. Thanks for sharing these. I will use them from now on.

Will V

[Jon Sullivan]

Interesting thread.

I'm one of the university staff guilty of incorporating iNaturalist into student assignments. It generally works really well, as long as I invest the time during the semester in curating the observations. As you've said, class observations generate a wide range in observation quality. I'm very aware that it is my job to curate these observations as quickly as I can (and ID everything I can) to not become a burden on the iNat community. That takes time with large classes. On the flip side, using iNat in classes is a great opportunity to promote iNat with naturalists in training and to get lots of useful (plus some not so useful) observations.

It's been clear that only a minority of my students have gone on to be regular users of iNat after the class. A few of those have become really big contributors. I'm not sure if my recruitment rate is much different from running a BioBlitz event that encourages lots of people to sign up and post their first observations.

I certainly don't do anything that connects "research grade" to credit. That sounds dangerous. Rather, I'm interested in students' observations being of high quality, meaning that they have photos identifiable to species, correct date, time, and location, and correctly filled in project fields. Observing and identifying are different skills and identifiable is more important than identified for my assessments.

I am interested in the idea of a student type of account that gets connected to another iNat account (the teacher's account). There could then be an interface specifically for the teacher to curate all the observations coming in from their students. This also has the advantage of making it easy for the teacher to curate these observations after the class has ended when many of the students never log back in. Perhaps, on request to the teacher, some student accounts could be promoted to full iNat accounts at the end of the class.

I don't see a student account feature as essential though, as most of this can be managed within class Projects. Still, there is certainly room for polishing this experience for teachers and a student account could help with that.

I disagree that student accounts should be unable to make other observations research grade. Some of my students have been really knowledgeable about certain taxa. I don't see their IDs as being different from any other random iNat user. The trick is not to link assessments to IDs.

I agree with Scott that we need to keep pressure on teachers to curate their students' observations. That's got to be the social contract for classes using iNat.

Cheers,

Jon

[Tom Norton]

Kudos to you Jon, for incorporating iNaturalist into your courses in such a thoughtful way.  If all other teachers took a similar approach, there would be no issues at all.  I agree with your comment that the process of curating class observations isn't really much different than curating observations from a public bioblitz.  Both types of projects need some ground rules, some education (and encouragement) for beginners, and a lot of careful follow-up by experts.  And it is wonderful that some of your students have continued on to become active naturalists.  That makes it all worthwhile!

Keep up the great work.
- Tom

[Charlie Hohn]

yeah just lest anyone misunderstand i think it's a great way to use the site as long as teachers/professors/TAs CURATE which is what Jon isdoing. And i agree that if research grade is not linked to credit, there is less of a need for students to not be able to give each other research grade. The underlying problem is people get excited and create a school project without researching how the site and our community works.

I am very opposed to giving inat users less ability to use the site because they are young, or less experienced, anything like that. It's only in the specific case where it's part of an assignment and they cheat. It's pretty much the only case where cheating occurs on iNat and is the cause for maybe 90% of the copyright infringement as well. If we do create student accounts, it is important that those students who want to keep using the site after the class is done are able to transfer their account. while recruitment rate is quite low, maybe as low as 1%, some great users came here that way