Records on iNat that are of "no value" because they're not identifiable

Maybe forcing / encouraging people to give some idea of what they are observing. That the Unknown or Life fuzzy image instead starts at animal or plant or fungus.

1 Like

If someone does not know what it is they have documented, there is no point in forcing them to guess. There are already too many complaints about inaccurate computer vision ID’s, clearly wrong guesses (lookalikes from different places) etc. That will just drive those numbers through the roof.

Speaking as someone who does a good number of coarse ID’s, in all honesty, unless you can get it tio family or genus, adding an ID of plants or fungus has low success rates. The occasional one comes through (and I still do them in hope they will), but rates remain low. There are just too many records and too few eyes to spend the time. In other areas (birds, reptiles, fish, mammals, some insects etc) it has good success.

3 Likes

I’m thinking of a a few obs where the ID goes to the insect, then the observer says no it’s that plant off to the left I want an ID for.
And the nice green pictures full of plants, but, which one needs an ID?

1 Like

as a clarifying point i believe this is meant inclusively not exclusively. It doesn’t mean ‘it isn’t for mapping any possible things’ but rather ‘it isn’t for mapping all possible things’, meaning it isn’t there to map litter, rocks, weather events, selfies, or whatever. I took it as meaning that though it could use re-wording.

There’s a huge range of options between ‘data is a byproduct’ and ‘delete people’s observations’. I don’t think either end of that is the right way to go.

3 Likes

There’s some discussion about best practices for the tricky ones here: https://forum.inaturalist.org/t/identification-etiquette-on-inaturalist-wiki/1503

6 Likes

Not presuming to speak for @kueda, I think it’s clear from the rest of his posts, and from all the development activity that happens within iNaturalist, that he and the whole staff value the data generated by iNaturalist at least as much as the rest of us data users. The word “byproduct” carries some pejorative baggage in other contexts. What I think is true in this context is that the “data of iNaturalist” are completely dependent on the success of the mission, connecting people with nature. No connection, no data. The more connection, the more data. And of course, the more noisy data too, for which the iNat developers spend a lot of thought, time, and resources to help the rest of us filter as needed. But they can’t do that at the expense of the primary mission, or it all implodes.

19 Likes

Well, as it was already pointed out - you could always just skip the obs you think of no value from YOUR point of view, but someone else might find some value in them. Also there’s neat feature (see screenshot) i use occasionally in obs in my geographical area, if it’s just a torn leaf or a flower in someone’s hand and ONLY if i’m VERY confident it can not be identified finer than Family level, let’s say. I’m aware that it can bring obs ID’ed at Genus lvl to RG. But then again - ID is a community thing and someone can downvote my decision.

PS/ funny it’s got hummer icon :joy:

4 Likes

i think it’s a gavel? I am trying to see a hummingbird but i can’t.

2 Likes

Oh dear, new english word for me, thank you! In kazak it’s the same word for gavel and hummer, so i was quick to…judge :sweat_smile:

7 Likes

no worries, but you caused me to stare at it for several minutes trying to imagine a hummingbird :)

3 Likes

my bad, i misspelled hAmmer)

5 Likes

It is funny that they chose a hammer. Very hard to ID after it has been hammered flat.

Apologies - a gavel - the judge’s decision is final.

5 Likes

Interesting philosophical discussion. The problem stated is “no-value records”. The staff informs there are none.

The disconnect for me is that I also contribute observations because I think there might be valuable information in it (for science and everybody else). The stated idea that inat shouldn’t restrict any contribution so that it truly is all-inclusive in my opinion doesn’t necessarily cause a problem. If inat doesn’t mind the extra data, I don’t either. While helping with identifications I can easily enough ignore observations that don’t make sense to me. But if inat wants to have useful data there has to be a way to separate those from the rest. That’s apparently intended to be done by IDers. I think defining a procedure on how to deal with “no-value records” would help. What happens when everybody who finds a specific observation meaningless simply ignores it and marks it as reviewed?

1 Like

I think it is much simpler than that. We make observations. Scientist wants to use the data, they can search by taxa, and then add fields to “mark useful records” for inclusion in their projects. They can scour what they know to be similar taxa looking for misidentifications. They can contact observers and ask for more detail.

To me (as in from where I sit in the observer spectrum) it comes across as the ultimate laziness on the part of the scientists, in that they seem to want the observers doing all the work! We are not your field technicians that you can order around… we are volunteers bringing you something for nothing! The gift horse should not be looked at in the mouth, as they say!

By all means give advice on what would be “more useful”, but getting pissy because someone calls the data a “by-product” is just ridiculous… pedestal just got a lot shorter!

11 Likes

This is perfect in regards to how I feel the mission works. My observations are fairly linked to what I’m interested in learning about and if I feel I can get good feedback from the iNat community.

The knowledge I learn here extends way past iNat and into other platforms and real world encounters where I spread the knowledge and interest. And I hope others do the same with the knowledge I share that I’ve spent years working to collect. That is what, IMO, is the most valuable contribution iNat offers to the community overall, and it will far outlive and outreach iNat itself in the long run.

8 Likes

good point! that’s cool that we have same proverb in kazak, but specifically mention horses’ teeth)

3 Likes

That’s a good point- as I have learned (at https://forum.inaturalist.org/t/using-the-field-similar-observation-set-for-linking-observations-of-lepidoptera-when-raising-on/1018 thanks!), anyone can mark almost any record with a new field.

3 Likes

this is true in a broader sense, ordering field technicians around is a good way to get garbage data because they get mad or apathetic but most field efforts like that don’t have photo vouchers so you never even know. definitely seen that one in action, even worse if they are unpaid.

Any scientist doing it right will either review every observation of interest as you say or else will use such a huge pool of data that the known errors don’t matter too much. but this is all off topic of course. There are truly ‘useless’ observations in the case of duress users photographing their friend (or enemy) and classifying as a baboon or whatever, but those are rare and easily filtered out. Meanwhile i just encountered someone harassing a user for a ‘too blurry’ photo when it was actually really easy to ID. It was of an invasive species and i don’t know why anyone would care more about a blurry photo than documenting an invasive species.

Maybe it needs to be more clear somewhere on the site that while offering advice on diagnostic features to photograph is welcome, harassing people about ‘blurry photos’ or ‘not getting out of their car’ (this one was on an interstate) is not appropriate.

12 Likes

In regards to kiwifergus’s comment.

I used to work in a research museum where my main job was handling and processing animal specimens and doing QA/QC on the data associated with the specimens. It was the physical equivalent in many respects of iNat. All manner of material was submitted to the museum for curation and cataloging, from beautifully preserved specimens with complete and accurate collection data to barely identifiable carcasses that had sat in a freezer for years with incomplete or ambiguous data. The former were a pleasure to curate, the latter were a pain and sometimes just had to be discarded*. So I’m well aware that the quality of a specimen – whether physical or photographic – can vary a lot.

I once heard that about 10% of all cataloged specimens in a well-curated research museum are misidentified to species. I don’t know if that is an accurate figure, I’m sure it varies a lot across collections, but it would not surprise me. Where I worked, it was always made clear by museum staff to research biologists using the collection that the burden of identification rested with the user and not the curators. I think that’s a given in all such museums and the same principle should apply to researchers using iNat. Mistakes happen and some specimens may not have had their ID determined by an expert.

Bottom line: researchers who want to use specimens and specimen data are responsible for reviewing them, confirming or determining the IDs, and assessing their value for themselves.

  • – Specimens that were discarded were judged to be of “no value” which was sometimes a tough call since you didn’t want to throw away something (even in poor shape) that might be of use someday to someone. But research museums are limited by storage space, which is not really an issue for virtual collections like iNat.
13 Likes

I have encountered the exact same comment/attitude, and the worst part of it was they thought they were doing “good”! No amount of careful dialog seemed to have any effect (at the time), but the attitude did change after I backed away from the discussion.

1 Like