Best practices for instructors using iNat participation for extra credit

I have used iNaturalist for college classes, and for credit in assignments for nearly 10 years now, and these kinds of open-ended assignments are often rather problematic. I am not doing those anymore, especially not for beginners.

Instead we do a species bingo where they have to find species of a certain category, I provide a clear manual with rules to the project, a report form for the students, and lecture slides that explain how to take good and bad photos of species (especially plants) with an associated assignments. Some of this is up on the website BotanyDepot.com.

Open-ended projects work great for people that voluntarily decide to join (like Personal Bioblitzes and similar bioblitzes or projects), or when you have an event with a leader that shows everyone how to get started. Many of my class projects are built into the curriculum but also take a LOT of work from me and other people that check all ID and cultivated/captive categories.
But for students that want to get extra credit with rather minimum effort (lets be realistic), I don’t think iNaturalist is the best place.

Some links:
Personal Bioblitz 2023 Project Webpage: https://tinyurl.com/PBinfo2023
About the Personal Bioblitz: http://tinyurl.com/RUPBabout
iNaturalist Project, Personal Bioblitz 2023: https://tinyurl.com/PersonalBioblitz2023
How to photograph plants - https://botanydepot.com/2020/07/27/presentation-how-to-photograph-plants-and-more/
Species Bingo: https://botanydepot.com/2020/03/13/activity-inaturalist-species-bingo/

16 Likes

Thanks all for the many replies and ideas! I’ll see what I can come up with to share with our faculty to hopefully provide some guidance and help address some of the issues that tend to come up with first time users like students participating in projects.

Just to clarify: This is not a class project by nature but a cross-campus challenge, sort of City Nature Challenge style (it will happen the week before CNC), with some faculty deciding to jump onto the bandwagon and send their classes out to participate. Several of our field courses went out on trips under faculty guidance and those were great and need little input since the faculty involved is also active on iNat. Other classes didn’t lend themselves to this sort of activity, so faculty just offered extra credit for students going out on their own. This is the target for trying to improve communication.

The first time we did this, bio faculty commented that this was the most engaged they’ve seen their students in a long while. So as far as getting students out into nature and involved in making observations it worked like a charm and there is an interest from the faculty side to keep the momentum going. I’m sharing some of the concerns, in particular regarding the excess of cultivated plants being observed and since the next iteration will be so darn close to final exams and at that point students tend to be more desperate for any sort of extra credit opportunity being offered.

I will also try to find some time to have a look at the user stats from our previous project to see which of the concerns that were voiced apply to us and may need some additional attention. It has been around four months now and I’m particularly curious to see how many students who joined iNat because their instructors told them about it are still active this semester.

6 Likes

Thank you for sharing; I’m very glad that you had a good experience when you were a student, and that you have continued to use the site regularly. :grin:

Now, not speaking to you (as I have no doubt you did a great job completing the assignment correctly and with care)… but speaking, in general, to instructors — The scenario described above by this former student is not a good way, in my opinion, to structure how and when students receive credit. It leads to some students deleting their observation (when one or more disagreeing IDs have been added by the community) …then adding a new observation with the same photo, the corrected community ID (as their own initial ID this second time around), and a new note describing the (other) organism.

Here’s a quote from my previous comment on this same thread, speaking to instructors in general:

1 Like

It is worth noting that your project is currently set up such that no casual observations show up. So you wouldn’t easily see how many students had stuff flagged for copyright infringement or submitted mostly cultivated plants since those are casual grade observations. Some of the students have 6 observations in the project, but 100 observations total, with most of them cultivated plants. There isn’t a good way yet to see how many students of a project were flagged.

7 Likes

Your comment made me curious, and was an interesting point. Well, here’s a link (not to the project’s observations, for the reason you mentioned), but filtered for:
(a) —the 7 days of the 2022 BioBlitz
&
(b) —Captive
&
(c) —with a box around the university on the map

https://www.inaturalist.org/observations?captive=true&d1=2022-10-24&d2=2022-10-30&nelat=36.21879518221118&nelng=-81.6734113016706&place_id=any&swlat=36.20711576405168&swlng=-81.70659736773537&verifiable=any

This won’t be an exact number, but gives a general idea.

1 Like

Yes, that’s the way to do it. We had to create a place for the project,so using Explore with filters for the same time frame and place as the project and unchecking “verifiable” will get you all observations including casual ones. It’s then just a matter of playing with the filters a bit to sort things out. The casuals are almost entirely made up of cultivated plants, and based on the observer list I’ve been able to identify the class/instructor and type of assignment (you probably guessed it: offering extra credit for students to go out and observe on their own) that caused the bulk of this. I’m hopeful that this can be improved for the next round.

3 Likes

Looking over the data, a couple of trends I’ve noticed so far:

  • Casuals: Cultivated plants are the elephant in the room - the vast majority of these (~95%) were observed by student participants. Other casuals include a few dogs and ducks and observations without media (technical issues with uploading pictures?). “Joke observations” were a non-issue and basically did not happen. We had one human and not a single observation marked “no evidence” because it was rocks or trash or whatever. So we need to focus on better instructions regarding wild vs. captive/cultivated.

  • Research Grade: The group of participants with more iNat experience was smaller than the student group and overall made fewer observations but made the majority of RG observations. This may be a proxy for better image quality etc. Students who followed instructions and joined the project did better on this than their peers who participated without joining. Maybe this is hinting at the value of receiving instruction and being willing to follow directions.

  • Identifications: These appear to separate into three chunks, each roughly making up a third of the IDs provided: 1) Participants who also made observations, 2) IDers who joined the project but did not add observations themselves, and 3) the iNat community.

Several things learned:

There are people on iNat who are happy and eager to help with student projects. Some folks joined the project and only added IDs, no observations. The average length of time these folks have been on iNat is 3.5 years, so most of these were experienced users. How can we find/recruit more of these? We were fortunate to have one Super-IDer who was not associated with the university in any way but joined the project apparently for the purpose of helping with IDs and quality control. Our Superhero single-handedly added 25% of all IDs (mostly plants) and marked a lot of stuff cultivated where needed. The help was noted and much appreciated!

Several of the instructors and graduate student assistants involved also pulled their weight on IDs, while others joined the project but did nothing. This latter group may need some guidance on how to follow up on their students. On average, the folks in the project who had prior iNat experience and added IDs accumulated ~80 IDs per person compared to a ~5 IDs per person average for the general iNat community.

The number of IDs made by students for others is negligible and there was no evidence of students adding nonsense IDs or much of any IDs really to get obs to RG. This was not part of their assignments (Obs counted, IDs/RG did not) so it didn’t happen.

The number of students who joined for the project and are still around (active in 2023 and having added observations or IDs after the project was done) is quite small (4%), but they do exist and at least one of them has started identifying for others.

9 Likes

One way to help students make quality observations would be to get them to focus on only a few species. Helping the students get up to speed on recording essential characteristics would help them to learn the process of photographing observations so that others can identify them.

For example, conifers. Knowing that they need to get needles, cones, bark, and the whole tree would help people ID observations.

I remember a friend telling me that I needed to include leaves in tree observations, bark wasn’t enough. Right. I always have cut trees in the winter, and bark and tree shape is how I see trees. Now I do get leaves, needles and seeds whenever I can.

6 Likes

One suggestion I have given to other teachers/group leaders is that the student’s first 3/5/10 (you decide how many) must be reviewed and verified by the teacher before any credit is given. The student is required to add the teacher’s inaturalist account in the first comment of their first observations of the project.

Is it a lot to ask a teacher? Yes, but no more than the grading of the final assignment, which will be easier knowing the quality will be higher.

7 Likes

Ahh…

One idea: make completing a tutorial with a quiz part of the extra credit requirement. The tutorial could focus on understanding data quality assessment with an “iNat or iNot” vibe. It wouldn’t have to be long or hard, but just enough to twig participants to the idea that observations have DQA requirements.

I know that if you do create such a thing, a lot of CNC organizers would love to copy it for their audiences.

6 Likes

Why? Wouldn’t there be anyone to attend the bioblitz without the aim of gaining these extra credits?

Consider the possibility to make them use iNat for at least some weeks before being enrolled in the bioblitz.

Good!

Frankly, maybe, should the whole system observation=extra credit be rethought?

There were a good number of folks (including students) who participated for fun and without this being associated with any sort of class credit. I think the faculty who offered extra credit were trying to help collect as many observations as possible without creating extra work for themselves (we’re all busy). This being a competition between universities for who can find the most species on their campus may ultimately be at the root of this.

From my perspective, extra credit for verifiable observations may just be the motivation students need to invest the mental effort to learn the difference between wild and captive/cultivated because otherwise they’ll miss those points. I’ll have to look through the data day by day again to see if there is evidence of student behavior changing after the first day or two. For me, that would be a positive outcome and one possible reason to encourage extra credit again. It will definitely require follow-up in the form of marking all cultivated plants captive as soon as possible so students will notice them dropping out of the project while the event is still going on.

In any case, making people do stuff has one problem for me in this situation: These are not my students, I hold no power over their assignments and grades. I’m just trying to provide guidance to colleagues less experienced with iNat to help steer this away from causing issues. Once observations get posted on iNat, I can keep an eye on them and follow up.

3 Likes

Well, a quick look day by day reveals there are days with more or less casual observations but no overall trend. One thing to look at I think would be by specific location (main campus vs. other areas), another would be finding out which classes went out on which days, and taking a closer look at percent casual per day for observers who made observations across multiple days.

The top graph shows number of observations total, while the bottom one splits up the percentages by category for each day. Our BioBlitz kick-off on Monday (10/24/22) was not on main campus and consequently there weren’t that many casual observations. Tuesday and Thursday had the most - I’ll have to check with our instructors to see which classes/assignments might have contributed to that. It seems whichever class contributed on Thursday would be a prime target for more guidance. Friday (10/28/2022) we had a couple of field courses go off-campus again for instructor-guided observations and casuals dropped down to normal iNat averages. Saturday was a football game day and students probably focused on tailgating, with some of the extra-credit stragglers presumably trying to catch a few more points on Sunday.

1 Like

Using the specific map box I marked out (link above), I got:

Total observations: 4,079

Breakdown:
Filter = Verifiable: 2,532
Filter = Captive: 1,528
Neither: 19

It’s interesting (and thorough) that you took the time to analyze and look for daily differences or trends.

It’s too bad that you don’t have control over the students’ assignments or grades, or else you could require a post-blitz class/session where students were required to (and instructed how to) “fix”/review their own observations (that need it)… such as:

  • review and research disagreeing IDs added,

  • add IDs to observations that they accidentally forgot to ID before uploading,

  • add or edit any missing or incorrect dates/locations (check for screenshots),

  • reduce unnecessarily large location accuracy circles,

  • delete observations that do not have an organism in the uploaded photo,

  • correctly mark DQA for organisms they accidentally missed as being captive/cultivated during initial upload,

  • separate multiple species mixed into one observation,

  • delete exact duplicates (same photo & same observer & only 1 individual/organism depicted in the photo, but multiple observations),

  • combine near duplicates,

  • delete casual observations with no photo and also no ID (there are 13 in the data set above),

  • add a cropped photo (if doing so might make an unidentifiable organism identifiable — note that iNat observers are not required to provide cropped photos),

  • review valid and relevant requests or questions from IDers

  • etc.

Speaking in general, not to you specifically:
These things are a normal, positive and common part of learning iNat, which is a learning curve for us all (myself included). But it’s also the part that’s missing (as several people have mentioned in this thread) in the case of a student who uploads observations, gets class credit, and quits iNat (without reviewing, curating and keeping up with their uploaded observations). I’m not saying that all students fit this scenario; thankfully some students do keep iNatting, and enjoy it a lot. And thankfully some instructors do monitor their students’ observations, as well (which is a lot of work).

6 Likes

annkatrinrose

I’m particularly happy that I’ve retired and can choose where I put my energy.

students did not pay attention to that and only took note when their observations were excluded …

Do you give them a rubric, in which requirements are laid out, and marking structure is explained?

Some instructors will not have prior iNat experience themselves and may not even have an account.

And somehow this will be your responsibility.

I’m picking up that your concern is: You’ve been made responsible for a ‘Bio-blitz’, whos main purpose is not to gather biological data, but to fill in time, tick a few administrative boxes, using instructors that don’t know what they’re doing.

Using iNat for this is like (metaphor alert) doing a trip to a food factory and allowing the kids chuck junk into all the inputs and machines.

In a past life I have been put into your position, ethically and professionally risky. Although I spose abusing iNat won’t cost you your job. My solution was to refuse to take on this role, but that has costs too.

You and others have identified something that worries me. An exact quote from a student “I’m not here to learn, I’m here to get marks” - and similar from many others.

geragotelis - sounds like you got a well planned, well recourced, well communicated experience. Lovely to read.

2 Likes

Just an update on our spring BioBlitz experience. We had a lot fewer captive/cultivated observations this time around - I interpret this as evidence that additional instruction helps! Here’s a comparison:
SpringFallBioBlitzComparison
We did see a good number of students who had also participated in the fall BioBlitz already, so that probably helped as well since they came to this with some experience. I have to dig a little deeper into the stats to see how many returning participants we had. Some have already asked me when the next one is going to be. Overall, we had fewer participants this time, so probably a more dedicated participant group.

Based on the feedback I’ve received, our top contributors did it because they really enjoyed it, not because it was some sort of class assignment or because there were prizes to win. (One of them was actually surprised to have won because they were completely unaware there were prizes, haha!)

While there were a lot fewer cultivated plants this time around, we had more students getting confident enough to add IDs. Not all of these were correct, of course, but I don’t see that as a big problem. I think we did a fairly good job recruiting faculty familiar with iNat to check IDs at least for their fields of expertise and prevent wrong IDs from reaching or staying in RG, and some students are responsive to such corrections. There is still a bit of an issue with wrong IDs that are not withdrawn once corrected by someone more knowledgeable, even if the student is still active. That might be the next thing for us to focus on trying to get across to students. They may just not realize yet that withdrawing/correcting their IDs helps with getting observations more quickly to RG.

8 Likes

Brava - good progress - that is encouraging to read about!

Wow! That’s impressive!

What kind of additional instruction did you offer?

That’s great! I also don’t see any issue with honestly incorrect IDs - that’s part of the learning process. I agree that teaching people about withdrawing or bumping back IDs is useful, but also trickier. One key learning objective that can be tied in there is understanding the nature of science as an iterative process where we correct mistakes and improve understanding over time.

3 Likes

One thing we did is pointing participants to your awesome iNat or iNot quiz! Thanks so much for putting that together.

I think another thing that happened was that the largest class participating this time around had already been exposed to iNaturalist for a class project before the BioBlitz. (I have to check but expect a lot fewer new accounts created right at the start of the BioBlitz.) We tried to tie this to an actual research project to make it more relevant and not just a class exercise. This was an in-class project where students made annotations and generated phenology graphs on iNaturalist under instructor guidance. So students out of that class who chose to do the BioBlitz for extra credit had worked with the website before (not just the app on their phones) and were more familiar with the concept of generating or cleaning up data for use in actual research. I think overall there was a stronger understanding that iNaturalist observations represent raw data for research and thus quality is important. Some of our top contributors were bio students involved in research projects.

In summary, I think framing it as a research project and not just a competition, as well as giving students prior exposure and a behind-the-app look at iNaturalist and what else can be done with the data beyond asking the questions “what plant is that?” has helped.

7 Likes