Exciting experiment with iNaturalist identifications

Props to the folks at iNat for putting together this experiment. Looking forward to seeing some results!

http://www.inaturalist.org/pages/identification_quality_experiment

Posted by sambiology sambiology, March 20, 2017 14:13

Comments

Thumb

Interesting. I entered as expert for North American birds but I wasn't sure about some other regions for birds or other taxa. I probably could have put in a couple more things.

Posted by vermfly over 2 years ago (Flag)
Thumb

Yeah, I'm curious what would be considered a "taxonomic expert..." I sure haven't published too too many papers, so I'm not sure I'd qualify as a true expert on anything. I would assume that these "taxonomic experts" would need to be true academics (various post-docs and multiple papers published and still publishing). Not sure though.

Posted by sambiology over 2 years ago (Flag)
Thumb

Oh, I never considered that they would only want publishing academics. That's pretty funny considering I've met a few of those who couldn't tell a Savannah from a Grasshopper Sparrow.

Posted by vermfly over 2 years ago (Flag)
Thumb

Definitely interesting. This would be very helpful for taking IDs into account based on the person's experience. In general, people only make IDs they're confident in, but it would still be nice.

Posted by ryanandrews over 2 years ago (Flag)
Thumb

Excited to see what they find! I am fascinated by the way iNat crowd sources data, and I've been curious about how accurate it is overall and how it impacts the scientific community.

Posted by tigerbb over 2 years ago (Flag)
Thumb

This is a good move on their part - I'm also interested to learn the results.

Posted by troutlily57 over 2 years ago (Flag)
Thumb

I got invited to work on the birds of the United States and Canada. They strip all the info except the pictures, date, and location. I never realized how much the descriptions of behavior and field marks not shown in particular pictures help to point towards the right ID. It is significantly harder.

Posted by vermfly over 2 years ago (Flag)
Thumb

The "expert" question is a very important one. For most plants, I'd really only consider someone an expert if they've published some taxonomic work or commentary on the group in question. This isn't because I don't think there are some very good experts who haven't, there most certainly are. But in the interest of getting reliable statistics, this would exclude a huge group of people who simply wouldn't know enough. Also, this isn't to say that published taxonomists are always that good at IDing from pictures. I've personally met some that say they hate trying to ID organisms from photographs.

For animals, I'm sure it's a lot different as so much more is available to the layperson, which should lead to a whole lot more people who know what they're talking about and could probably be judged on whether they have given talks about the identification of the animal group.

I would participate, but it might not be too good for the statistics to compare against myself. :)

Posted by nathantaylor over 2 years ago (Flag)
Thumb

I think that's really cool. We are a young homeschool family that loves to walk in nature and see what we find. We use iNat to record some of our findings and, quite honestly, rely on the expertise of others to help us learn more about what we find. The people on here, especially @sambiology, have taught us a great many things about how to record observations, what to look for, and just how to really see what we are looking at. Having this new level of expertise is amazing. We love iNaturalist and are really grateful to all those on here who are helping us to learn. :-)

Posted by aprilbarnett75 over 2 years ago (Flag)
Thumb

This is a really interesting experiment. I usually refrain from ID'ing anything I'm not sure of, either as the initial ID or agreeing with others' IDs. But there are times I am tempted to confirm an ID by someone I consider an expert...a sort of "halo" effect. It looks like this experiment will determine if the halo effect affects crowd identifications. I know some non-academics who are supremely qualified as experts in a certain area, and I know academics who don't know nearly as much as they think they know. Not sure how you pick out the qualified experts, but I'd hate to lose the knowledge of those that aren't academics.

Posted by naturemom over 2 years ago (Flag)
Thumb

Very interesting experiment!

Posted by laurenjansensimpson over 2 years ago (Flag)
Thumb

It is indeed a really great experiment, and I hope it leads to a publication by @loarie @kueda @tiwane and whomever else at iNat.

I do think the "comment" section is underused on iNat. I'm challenging myself to use it more often so others can see how and why I give an ID. If I say the characteristics that are important, perhaps people will be more willing to add in ID's for others based on those crucial defining characteristics.

Also, in my opinion, it's TOTALLY appropriate to use the comment section to say "Wow!" or "neat-o" and "nice shot!" After all, this is what makes iNat fun for me -- it's a social network of nature enthusiasts. It's ok that we're friends with each other. :)

I'm definitively not an expert, and I wonder if I shouldn't participate in this as I toss in loads of ID's already. I think they may be seeking out folks that aren't already on iNat adding in identifications. iNat is quite fortunate to have many experts (at least, people that I consider experts!) that are already active participants.

This will be a fun thing to watch indeed. :)

Posted by sambiology over 2 years ago (Flag)
Thumb

I really like using the comments when the picture is less than 100% clear. I give explanation about habitat, behavior, and characteristics that aren't in the picture to help the reviewers understand why I IDed a species as I did.

Posted by vermfly over 2 years ago (Flag)
Thumb

@vermfly See? Now we've learned something else! It didn't really occur to us to to list the things you're talking about in the comments. Occassionally we add some small detail but usually nothing more than the general location where we found it. We need to be more descriptive in our comments. That makes for better observers! Thank you for teaching us something new today!

Posted by aprilbarnett75 over 2 years ago (Flag)
Thumb

I use the comments both ways that you described, @sambiology. In addition to naming the species if I can, I indicate atmospheric conditions and temperature, number of specimens observed, and the plants on which I found them, as well as any other details that I think might be of interest to someone using the data for research. My project last year was to try as much as possible to make a daily record of the diurnal invertebrates that I observed at our home gardens---although admittedly, I still need to upload my data and comments from the last quarter of 2016 (Life got insane! But all of the data is archived and there). I also like to comment on beautiful photos or interesting observations that others make. I love that iNaturalist is not simply for data collection, but is also a community of folks. When I present about pollinators and wildscaping, I highlight that wildlife conservation is also a community effort: it's something that brings the human community together, as well as something that tends to wildlife.

Posted by laurenjansensimpson over 2 years ago (Flag)
Thumb

@aprilbarnett75 I come from a birding background and all those details are what allow reviewers to determine if someone actually saw what they think they did. I can often make an ID for someone over the phone despite the newer birder having limited knowledge of specific anatomy of the bird as long as they can give some basics on color pattern, habitat, and behavior. Relative shape of the bird can be useful. Size is often a terrible feature to use because new birders often don't have anything to compare. With that limited information I can often point them to the right species in their field guide so that they can confirm their mental image with the book.

Posted by vermfly over 2 years ago (Flag)
Thumb

I'm excited to see the results of this. If anyone belongs to any identification groups or expert groups on Facebook, definitely share this link with them there. I've shared it to a few groups to try and recruit more taxonomic experts who haven't dipped their toes in the iNaturalist water.

Posted by carrieseltzer over 2 years ago (Flag)
Thumb
Posted by sambiology over 2 years ago (Flag)

Add a Comment

Sign In or Sign Up to add comments

Is this inappropriate, spam, or offensive? Add a Flag