Option to Automatically Withdraw IDs on Observations

I propose a feature that you could toggle on to automatically withdraw your ID if someone disagrees. Obviously not everyone would like to use this, especially experts in the field, but for non-experts with a lot of observations like me I think it would be useful.

Personally I upload nearly all of my observations off my android phone which is quite laggy and doesn’t have a mass upload feature (at least, not that I know of) so sometimes it’s just faster to select whatever the AI suggests and I don’t always have time to double check if it’s something accurate. Plus, especially with things like fungi and invertebrates, many basically go forever without having another person ever contribute. Or, again, as a non-expert, it can be hard to distinguish between certain organisms.

I usually go out of my way to withdraw an ID when someone disagrees but I go long periods without internet and these messages often get buried. It would be great to have a way to just automatically opt out of going through all my old observations and removing every single ID.

This is an interesting idea, and I could see wanting to apply it for some of my observations. I often just write something like “ID based on CV, not my expertise” in the description. My real goal of using the CV is to get the observation findable by someone who has expertise, but I don’t necessarily want the CV to “count” for an ID and do sometimes withdraw that identification later.

6 Likes

I would not use this, mostly because I’ve had scenarios where people add purposefully erroneous IDs, or are just wrong, and I want to be able to question them.

I can see that it may be helpful for some people, but not I.

8 Likes

I don’t mind this idea necessarily but my feeling is that the real issue here is just the ongoing need for the notifications revamp. I doubt many people would really mind taking a look at every disagreeing ID, but they can just easily get buried in hundreds of refining ID notifications, especially if the initial ID was coarse. And there isn’t really a good way to search for them if you do miss the notification, other than waiting until the ID goes ‘maverick’.

It also would need to be implemented in such a way that if, say, I accidentally misclick an autocomplete suggestion the original ID would get reinstated if I withdrew/deleted my completely unintended ID, which could lead to some weird edge cases for the behavior.

8 Likes

I think it’s a feature, not a bug, that we can’t do this. I often ID things tentatively, or going by the cv, and try to declare as much. But it’s a great learning tool, when someone corrects me, to review the obs id while withdrawing my incorrect ID.

11 Likes

I agree… I also use this as learning curve.
But I agree on the fact that it could be easier to detect those disagreements.
There is one URL around to find your mavericks easier, which I stumpled upon at some point, which is helpful, but these kind of tricks are not always easy to find…

https://www.inaturalist.org/identifications?user_id=ajott&category=maverick (just replace username with your own)

4 Likes

Better management of notifications one day will be good.

I am IDing for our City Nature Challenge.
Have 39 notifications since an hour ago. I will check each one. Especially where my ID was tentative - do you agree, are your IDs informed and reliable? Are you kind enough to explain why it is (or isn’t) - what you can see in the photos that makes your decision. (Today’s lesson Mimetes leaves have 3 teeth / nectaries)

Agree or disagree I learn a little more each time. For me that is a very good part of iNat!

8 Likes

I agree, sometimes I’ll write thst in the notes, but it would be nice to have a way that just automatically cancels my ID when someone else adds to it so I wouldn’t need to go and do it myself.

1 Like

Yes, for most of the organisms I document veeeeery few, if any, ever go maverick.

1 Like

I would agree with dianastuder^
I don’t think having this feature wouldn’t be the solution you think it could be. Yes, if you make mistakes and don’t check your notifications often you might have incorrect IDs out there, but it could also give some credibility and weight to incorrect disagreeing IDs if yours automatically withdraw when you were actually correct. I also think it is more helpful for learning if people read actually look at others’ disagreeing IDs and hopefully their explanations why they disagreed. That chance to learn is lost when you trust in that type of function to do your fixing for you.

I think the solution is to have a better way to check notifications that separates disagreeing IDs to yours… Or as I proposed to the iNat team a while ago, a tab for notifications about your observations and IDs that need review. It is easy for those types of notifications to get lost in the sea of disagreeing or refining IDs on top of yours if you ID a lot. Using URL strings is also a bit of a hassle for most users, and I am all in favor of an easier-to-use feature such as a tab or button to sort through disagreeing IDs.

9 Likes

I’m puzzled (and probably naive). Why would someone purposely add erroneous ID’s?

On a larger note, I would not mind the option to do this (it would have to be my choice, for each observation). With some Unknowns I will add a rough ID that may or may not be accurate. Although I check iNat 2-3 times per day, I do miss some notifications. If I roughly ID a beetle as an earwig (as I did recently) I would prefer the ID be automatically withdrawn in case I missed it. Rather than have numerous people having to add ID’s.
I can see both sides, however, so am not attached to the idea.

I wouldn’t say erroneous, but, when I’m looking at ‘unknowns’, I’ll put an educated guess at the lowest level I can. Sometimes it’s wrong, say wolf spider vs running crab spider. I’d be fine with an auto-withdraw for ones like that or when I use the CV when I’m not sure what to choose and someone more knowledgeable comes along. This is not a common scenario, but I understand the intention of the request.

1 Like

Children can add anything that comes to their mind as an id to others’ observations, they’re not much better with own too, e.g. iding photo of a classmate girl as a snake.

6 Likes

But, if this ‘I was wrong’ is automatically withdrawn - you won’t see how a scientist has patiently explained WHY you are wrong. The scientists get weary of people who ‘ignore’ their comments. And you will continue to make the same uncorrected mistake in future. That’s not a win win for anybody.

6 Likes

Yep! There is an observation of a kid in body paint IDed as a tiger. I occasionally get notifications on that one still.

4 Likes

That’s super helpful. Thank you! I had four I needed to clear, some where I was inappropriately blocking community ID.

1 Like

i sort of think this could encourage people to submit uninformed identifications more freely. if you’re not comfortable with your ID, then my take is that you shouldn’t add that ID. if it’s too many buttons to click to make an alternative ID, then i think a better request (for the app only) would be to add a secondary list after the computer vision suggestion list that includes just the standard set of iconic taxa (birds, mammals, insects, plants, etc.) for folks to select from. i think something like that would encourage people to select high-level taxa (and educate them that it’s okay to select high-level taxa) rather than unknown or an uninformed low-level ID.

7 Likes

That is saving you time. While you would like to save even more time by letting iNat tidy up when others - who spend their time adding a slow ID - disagree with you.

That means other people have to invest more time to sort out IDs. Remember the iNat rule. If we disagree about an ID iNat requires MORE than two thirds to agree. Think about the identifiers. 2 people agree = lovely. 2 disagree - now we need 3, against that wrong one. Someone agrees with the first wrong one, 2 against … now we need …

Identifiers can better spend their time on fresh obs than the cleanup on aisle 3. We all value our time. Lots of us battle with slow internet, and I am in Cape Town where we deal with loadshedding slots as well.

Maybe try a different workflow. Upload as Unknown and come back later, when you have time and internet, to add slow informed IDs?

7 Likes

And @marina_gorbunova - thanks. I had forgotten about that aspect.
Do either of you (or anyone else) know of instances where legitimate observations and identifications are deliberately erroneously identified by the same person? For example, if I post photos of pileated woodpeckers and the same person keeps adding Bald eagle to all my observations.

Yes, yes, yes! Cleanup on on aisle 3 absorbs far too much of my time, especially if you leave reasons.
I only add rough ID’s to ‘Unknowns’, and only when I’m reasonably sure about what it is. I do (gulp) make mistakes. Rarely. Occasionally. OK, more than that!

1 Like