Bioblitzes and observation quality…or quantity over quality

General
1

Going to start this with a question….is there an ongoing effort to help bioblitz organizers both achieve their goals while not resulting in a landslide of observations to iNat that are (1) evidence of people trying to get a top spot for most observations (2) lacking in basic info and/or (3) burdensome to identifiers becuase the photos aren’t cropped, initially identified by anyone associated with the bioblitz, etc. etc.?

I ask this because I dug into the cultivated plant “unknowns” and found a bunch of pollinator challenge observations that were badly cropped, but had bees, but were ID’d (correctly) as cultivated plants because the photo wasn’t uploaded in a way that anyone would easily pick the bee. Plus, multiple photos in short periods of time suggesting the same bee but an attempt to game the leaderboard. I looked at the rules for this particular bioblitz and there really weren’t any. Worse, it said that people on iNat would identify the observations. Notably, the current bioblitz guide tells bio blitz organizers to recruit ID’rs. Not going to link the bioblitz here, but it was connected with a much larger national organization.

I know there’s the ambassador program, but is there any effort at iNat to address these issues? I’m newer so I apologize if this is old hat. Some of these problems sound like what may have plagued the CNC, but I don’t know for sure.

Thanks….

11 Likes
2

as an aside, the unfortunate reality is that no matter what iNat staff do to help address these issues, there will always still be bioblitzes which frustratingly produce lots of low quality data like you describe. As one example:

A few years ago I was asked to help out with a bioblitz. I won’t reveal specific details, but basically it involved 2 or 3 main organisers of the event broadly, and then numerous local organisers each responsible for their own group/location; basically similar to the CNC, but on a much more localised scale. In the several months leading up to the event, which ran across a week or so, I did 3 separate hour long workshops over Zoom for the local organisers, going through a very basic how-to for iNat, both desktop and the app(s), and also doing an extensive run-through of all the dos and don’ts for bioblitzes and iNat broadly from a data quality perspective, touching on all of the things you mentioned, and the stuff that regularly gets discussed in other similar threads. These events were quite well-attended. In addition to me speaking and presenting (where I shared my screen and used iNat live in front of them), there was also lots of question time for the organisers. These local organisers would then ostensibly pass on the information they’d learned to their respective participants.

I then also provided supplementary resources, including written documents and videos (incl. existing material I had made, and custom material I had tailored to this event), to the main organisers who then distributed them to the local organisers, both to cater for people who missed one of my Zoom workshops, and also just provide extra info on top of what I presented.

And yet, when the bioblitz started, there was a flood of low quality records, with many examples of every single thing I told them to avoid/not to do (multiple pics of same individual split across records, multiple species in one record, unmarked cultivated plants, blindly trusting CV suggestions, numerous people uploading pics of the same blurry pot plant, etc etc). I don’t think I could have done much more to prepare them for this event, and yet it seemingly didn’t really make a difference. It was quite a frustrating experience and I felt like I’d just totally wasted my time (and then also felt somewhat guilty for all the dodgy records that were now on iNat).

Now this certainly isn’t a universal experience for me, and is in fact the exception rather than the norm; I’ve directly organised or helped out with many bioblitzes in Australia over the last 5 years, and there have been plenty that ran very smoothly, with diligent local organisers who did a great job at absorbing and passing on all the recommendations and info, and great participants who paid attention and followed them all. But my point is that it’s an inevitability that some bioblitzes will result in these landslides of low quality records.

(note that I don’t disagree at all with the importance of site-derived actions/initiatives/resources, and I’m not saying that just because some bioblitzes turn out poorly despite efforts otherwise that we should just give up and not do anything; I definitely think they’re really important and should be developed and implemented. But just emphasising that the successes + usefulness of these solutions are, frustratingly, sometimes out of the control of iNat staff, plus ambassadors etc, and it’s important to keep this mind)

20 Likes
3

Thank you - this is a really helpful perspective.

1 Like
4

If that’s one single individual bee uploaded multiple times, that’s bad. However, sometimes I’ve stood at a big patch of flowers, e.g. Lathyrus latifolia, and photo’d bee after bee after bee. Some photos might be duplicates of an individual bee because they look all alike, but several of each species are present and I’m trying to post each individual once. What I’m trying to say is that a group of observations from one place at nearly the same time might be an honest effort to post different bees.

Of course, it might not, but we try to assume good intentions. Though I must say that we see some pretty bad intentions from the big bioblitzes sometimes, especially the CNC.

The problems of bad photos, the subject of interest too small to ID, lack of cropping where it would help, ambiguity about the subject of the photo – they’re all familiar. Having the organizers link to a page with suggestions for good photos can help somewhat. Conversations with individual observers can help – or sometimes not. iNaturalist is citizen science, though, and tries to welcome everyone. Citizen science is always messy. Best to focus on the usable data you did get and not the observations that are for one reason or another not so useful. (We do laugh or groan at the 10-pixel bird on a wire, for example. Sigh.) These problem observations do waste identifier time, but often not much – reviewed, move on.

10 Likes
5

For organisers who repeat their blitz with a new bunch of people each year ? Should iNat require them to show good faith in their use of iNat again for each new project ? The organiser can tidy up last year’s mess, before they may dump a fresh batch. Wording this carefully - one of Cape Town’s universities sets their entomology students loose on iNat each year. The students are on iNat for weeks for that assignment. And gone. The admin sets up the project. And gone. The obs are mostly junk. Frankly not a good advertisement for the entomology department.

10 Likes
6

I wish these events had different metrics by which they were assessed (if they must be competitive). If the goal is really to have people connect with and learn about nature, why not have “judges” of some sort who look for interesting or exceptionally detailed observations. For example, someone who observes a few mushrooms, gets good photos of the gills and stem, records/describes the habitat, takes a spore print, and provides a well-reasoned initial ID after reading up a bit is going to learn a lot more than someone who photographs every mushroom they see on their speed-walk through the park and posts blurry photos of the caps with the CV-suggested IDs.

Of course this is far more subjective and time-consuming to assess, but so are school science fairs, and the purpose is similar. Seriously, imagine if science fairs we measured by some impersonal measure of quantity. Make these events something that provides and rewards a bit of depth rather than superficial breadth. I understand that not every observation will be amazing (mine certainly aren’t), but it’s the effort that counts, and prioritizing this aspect would get people started down the right path. This produces a higher value dataset for IDers and researchers and a more meaningful and rewarding experience for observers, so seems like a win-win.

Edit: Or imagine an art show where it was a contest to see who could make the most paintings. That’s obviously absurd, and for a kid who actually cared about art, it might even be a frustrating experience, giving them an early lesson on misaligned incentive structures and the modern fixation with metrics over substance/purpose, which is presumably not the goal.

16 Likes
7

Or inexperienced users who don’t know how to upload multiple photos in one observation or who are not sure whether all the photos are the same bee.

I don’t think it is necessarily exclusively a bioblitz problem. I’ve encountered numerous cases of non-duress and non-bioblitz users who are working on some pollinator project and are using iNat to record their data even though they lack any prior experience using iNat and do not have the necessary skills/equipment to take decent photos or identify the pollinators themselves.

In other words, it is also about not knowing what is involved in biodiversity recording and that collecting good data requires a bit of thought and work – it isn’t a matter of just uploading an observation and iNat magically does the rest; iNat is merely a tool for storing and organizing the data.

The “plug and play” nature of the app (point your phone at something and get an ID) and iNat’s open scope (few requirements about what to observer or how) are strengths of iNat with respect to encouraging user engagement and involvement, but they are simultaneously also major weaknesses with respect to data quality.

Many citizen science projects deal with this by being much narrower in scope and providing a certain amount of structure and instructions for participants from the outset – e.g., count as many bugs as you see at one place in an hour, or sample the water using this protocol. This probably results in somewhat lower participation and less data overall, and certainly less opportunistic data, but the data that is collected is probably of better quality.

Perhaps the lesson here for both iNat and bioblitz organizers is that more structure would be useful. For iNat, I am going to be a broken record and suggest that user onboarding should be integrated into the interface (not in the form of optional tutorials). For bioblitzes, perhaps reframing the purpose of the event would be helpful – not getting as many people recording as much as possible, but specific goals not connected with the numbers of observations.

13 Likes
8

I’m pretty convinced that, no matter what iNat or local organizers or anybody does or could do in the future to try to teach observers how to use iNat “properly,” some observers will screw things up, just because they are human. I’ve certainly done it and I’m a very experienced iNat user and a retired professional biologist. Sigh…. And, of course, iNat itself, the global organizers of the City Nature Challenge, and many others are trying their best.

So, while I think it’s worth trying to improve bioblitzes or any other effort using iNat, I think we just have to accept that some observers will screw things up and then not respond to comments. Don’t let that stop you from contributing to iNaturalist. Just mark such observations as reviewed or mute certain observers, if their actions really annoy you. And move on.

13 Likes
9

Good to hear a broken record sometimes!

1 Like
10

While it seems inevitable there will be these issues after a Bioblitz, I think the organizers should be more responsible for going in and fixing the issues on their Bioblitz - IDing unknowns, adding comments on duplicates, marking DQA’s for cultivated or multiple species, etc instead of that cleanup work falling on the general iNat population. It seems like there have been several projects lately that have hundreds of unknowns added that are still that way months later. You’d think the creator of the project would care more about actually making the data in the project useful and do some of that work themselves.

6 Likes
11

What about an “Preselection” before you can join a Bioblitz?

  • Needs 10 previous RG observations with full annotation before an Obs. is accepted during the period of the Bioblitz.
3 Likes
12

Unfortunately, some organizers are just as flighty as the observers - they are on iNat for the period of the event, and then they too disappear.

The event I was looking at had this problem. And overall, I think some events are driven by a well meaning but misguided intent to have a community event that is also nature related. Example - people want to save the bees. They have a bioblitz using iNat. Folks involved may think it’s a success because people ran around looking for bees and other pollinators. But was it actually a success? Not on the iNat end at least.

5 Likes
13

Not sure how this fixes the overall problem, unless your thought is that observers have some pre-event familiarity with iNat.

14

Very good point. And that is also what i saw in the event I was looking at. An express statement that “oh, iNat will identify your stuff”.

Aside from failing to grasp that iNat isn’t like eBird or has a bunch of paid staff just for ID’g, it doesn’t engage with the issues like you said.

3 Likes
15

I think this might be a result of conflicting priorities between the Inat community and the hosts of many Bio-blitzes. I am guessing that most long term regular users of Inaturalist would rank high quality scientific data pretty near the top of our priorities when using the site. On the other hand many Bio-blitz hosts (not all) are primarily trying to hit one of two targets:

  1. Increase awareness and/or visitation to the site of the blitz.
    In this case the moment the blitz is over their goals are met, or missed, and thus there is no incentive to clean up the data or even promote good data in the first place.

  2. Trying to get a “species list” for the site.
    This one seems to align itself better than the first with good data collection, but unfortunately what many hosts are actually trying to collect is a list of species from the larger vertebrates or flowering flora, tying it back into target 1.

That being said, I think INaturalist is taking a huge step in the right direction with their new(ish) Ambassador program. By promoting regular Inat users to get directly involved with their local communities they are opening a resource to Bioblitz attendees and hosts alike to get on the same page. As noted in many of the comments it isn’t going to fix the overall problem since there will always be people hosting or participating in blitzes that just don’t have the interest or knowledge to get it right, but it is a step in the right direction.

As part of that program I am going to be organizing Bio-blitzes myself in the coming year for each of the County owned parks in my area and am hoping to avoid this issue by first leading free introduction to INaturalist programs at at the Library locations near those parks.

6 Likes
16

Agreed on what you said.

And I too am hoping the Ambassador program might alleviate some of these issues.

17

There is an unavoidable tension between iNat’s two main goals of (paraphrasing) connecting people with nature and producing high quality biodiversity data. Bioblitzes are perhaps the clearest examples of this tension, with it being really hard to both include everyone and make sure that no participant is ruining the data.

9 Likes
18

I think if you host an event, a party, a gathering, etc. You hold some responsibility over the impacts of that event both positive and negative. If it’s an awesome success, great! You hosting that event are partially responsible for that success. If it causes many issues and requires a cleanup, you also hold some responsibility for that.

I think hosts of events that use iNaturalist that do not engage or help at all if their events cause large issues should be suspended / blocked from creating future projects if all attempts to communicate with them after months fail.

Again, when you host an event, say a gathering at a house. You take partial responsibility for the event and even actions of the people you invite. While the analogy isn’t 1 to 1. I fully believe that hosts are partially responsible for the content participants of their events upload.

Does this mean the burden fully rests on the host? No and it shouldn’t, but if there are issues, especially widespread major issues with the contents being uploaded from your event. You need to show some responsibility for that.

5 Likes
19

That is exactly what I am suggesting: Get familiar with iNat, make some useful contribution where you prove that you have understood the working of iNat, and then you are welcome to fun and games with it.

1 Like
20

This is why I would like to see some “preselection”, a proof that the “Bioblitzer” has understood what a community science project is: A project where the community helps the community to appreciate science

1 Like