There are a lot of plant and probably other observations with a single photo that has multiple species in it and the observer does not say which species the observation belongs to. This seems like the most obvious reason someone would choose to check “Evidence related to a single subject”. It’s the same as having multiple photos of multiple species in the same observation.
I disagree that these situations are equivalent and believe this statement is incorrect. Staff have made it clear that the single subject DQA field should not be used in this scenario:
The Help page for this DQA notes:
"Please do not vote “No” to the “Evidence related to a single subject” DQA condition in these scenarios:
- When there are multiple species in a photo and it’s unclear what the subject is (e.g. a single photo with both a duck and a pigeon in it)."
9 Likes
Good to know. That is very counterintuitive and most people would probably never guess that nor would they seek out the instructions in this situation. I certainly wouldn’t have. I guess maybe it is good that I mostly ignore the DQA. This feature request would, I guess, help in the single photo situation then but, of course, many people will probably be confused by why it is greyed out. “Evidence related to a single subject” as written gives absolutely no hint that it doesn’t refer to multiple subjects in the same photo. If that is what they want it to mean, they should reword it so that it is accurate for only the situations they want it to be accurate for, especially as there are so many situations that fit “Evidence related to a single subject” that apparently the DQA should not be checked for. If it is specific to multiple photos/recordings, it should be something like “Observation doesn’t have different taxa in different photos/recordings”.
while I agree that if the “single subject” option’s wording would be somewhat ambiguous if it were given with zero context, in practice this is exactly what the help pages are for – to clarify and give context. I’m not sure about other people, but for me, if I need help looking at something complicated that I don’t fully understand (such as all the switches and checkboxes of the DQA panel)… then I read the manual. everyone should be regularly encouraged to consult the help pages, in my opinion.
but to the main point, the case of a single photo with multiple organisms isn’t irredeemable because identifiers (even if the observer has left iNaturalist) can agree to focus on one given organism in the observation – whereas multiple photos, each with (usually one each of) a different organism, are not possible among which to pick only one species.
4 Likes
Just today I saw an observation marked as casual using the “single subject” DQA for a duplicate observation. Obviously not what it is for, and the observation in question only had a single photo.
2 Likes
I disagree. I only go to help pages when things are ambiguous or confusing and I would guess most other people do the same. “Evidence related to a single subject” unambiguously refers to more than the staff want it to refer to. Because “Evidence related to a single subject” is unambiguous, there is zero context for someone to guess that it means only a subset of “Evidence related to a single subject”. That said, I see no good reason not to tag single photos as “Evidence related to a single subject”. Why make a tool that only tags a select set of observations with this issue? If they only want a subset of “Evidence related to a single subject”, they should change the name to make it accurate for what they want.
Almost every photo on iNaturalist contains evidence of multiple species/organisms. By this logic, I would think that any photo showing an organism present on another organism, an organism eating another organism, in anyway interacting with another organism, or two or more organisms near to each other could be flagged.
This statement
however, is not correct. Staff have explicitly asked users not to, as highlighted above, which is a good reason not to do so.
However, if there are users that will continue to use this DQA field to downvote single photo observations, despite appropriate guidance not to do this, it is a good reason to grey out/make this behavior impossible in my opinion.
8 Likes
Especially considering other DQA fields, especially “location is accurate” and “organism is wild” are routinely abused.
Because. It is a solution to this problem.
Obs with pictures of - a bird, a flower, a tree, your dog. We cannot ID because the observer needs to split it into separate obs. The new DQA is to manage multiple sp, each with their own photo.
“It is not clear what the subject of your photo is” is a different problem and must fight for its own new solution. Again the observer has to tell us - what are we looking at? Bug or flower? But if the observer doesn’t respond, identifiers can pick - I will take that bug.
4 Likes
A different problem if “Evidence related to a single subject” is reworded to accurately reflect what it is looking for. A facet of the same problem if not reworded. And, of course, it is kind of silly to make a DQA for one and not the other if treated separately.
My only real issue is that “Evidence related to a single subject” can easily be misinterpreted and, by not using more accurate wording, it is asking for people to misinterpret it. I put in a feature request this morning to see how many other people felt this may be an issue and if there could be a discussion on getting the wording of “Evidence related to a single subject” changed to more accurately reflect what it is asking for and thus reduce the number of people that will inevitably misinterpret it and check the box when they shouldn’t. That request hasn’t popped up yet, so maybe it is in the queue or maybe was deleted. Either way, I’ve offered a simple solution to what may turn out to be a major problem. People can do what they’d like with it. I’ve done my duty and it is out of my hands.
The ambiguous wording of the name was brought up by multiple people when the feature was introduced; see:
https://forum.inaturalist.org/t/new-dqa-does-evidence-related-to-a-single-subject-mean-the-same-individual-is-in-all-photos/49220/
https://forum.inaturalist.org/t/easy-way-to-mark-multiple-species-observations/278/194
So staff is aware of the issue.
I see that the German translator of the DQA has rendered it as “all evidence relates to same organism”…
4 Likes
You are welcome to find it ‘kind of silly’ but I fought for a solution to an issue I have battled for the 6 years I have been on iNat. Just dump it in the broadest ID that fits everyone - was never a useful or effective solution. Now we can take it out of the Needs ID queue, so that identifiers can concentrate on obs where their ID effort is useful.
2 Likes
Yes thank you - that makes better sense. (Subject does not even need to be ‘evidence of life’)
And that is great but it is a two pronged issue that could easily be solved by two DQAs rather than just one. It’s great to solve one of the issues but silly to not solve the other related issue at the same time, which would also address the potential for confusion I bring up. Perhaps someone will have to battle for another 6 years to fix something that should have been included in this fix.
Why do you feel that a DQA is necessary for observations where it is not clear what the subject is? What would such a DQA do? Would it make the observation casual?
Technically speaking, there is nothing defective about such observations (in contrast to the other DQA items, which all indicate some issue with the observation or the data connected with it). An observation where there are multiple identifiable organisms in one photo but no clear focus is a perfectly legitimate observation in terms of fulfilling the basic requirements for data and evidence; there is no rule that the focus of the observation must be obvious. In such cases, we merely haven’t been informed about the observer’s preference.
In practice, such observations tend to get resolved: either by asking the observer what they want ID’d, or because an IDer chooses some organism in the photo to ID and hopes that it is in accordance with the observer’s wishes.
This is not really all that different than an observation entered without an ID where there is an organism prominently featured in the media that is likely to be the subject of interest. Here, too, IDers might make a guess about what the observer wishes to have ID’d – they may guess correctly, or the observer may indicate they were interested in something else, or the IDers may disagree about what the focus is (e.g. plant people vs. insect people).
But this is a communication problem, not a data problem.
8 Likes
Fair arguments. I guess I’ve run into a lot of observations with an ambiguous subject where the observer has not responded with clarification years after being asked, many more than I have run into of observations with different subjects in different photos. So, essentially they clutter things up in the same way and keep appearing over and over again to people trying to work their way through old observations needing an ID. Maybe these are not similar problems in some ways though they are very similar in others.
There is still a good argument for having the DQA text being less easy to misinterpret. If I’m misinterpreting things, lots of other people are or will as well. Why choose vagueness over clarity? All text should be as intuitive as possible and as unlikely to cause misinterpretation as possible.
iNat has its own jargon. And translating it must sometimes be … interesting. Much more difficult for people whose first language is not available on iNat. Even English as first language speakers cannot agree.
Casual is not versus formal clothing. It means Not Wild or broken obs missing location or date.
Research Grade means 2 people agree. That is ‘research’.
iNat jargon means what they say it does, and we must work within that.
I completely agree. a solution is not silly just because it’s a solution to one thing or not another. I wouldn’t throw my house key to the ground out of irritation that it’s not also my car key.
Exactly as you go on to say, such observations are perfectly redeemable.
If, in an observation:
… then once again, it’s “free real estate”. Identifier’s prerogative, as far as I’m concerned, to identify whatever makes the most sense.
The disputed DQA option is about whether a single subject can or cannot be chosen, not about whether multiple subjects could be chosen. I agree, it would be hypothetically nice if it were clearer on that. But after a certain amount of prevarication, it would take an entire paragraph to explain. That sounds like what the help pages are for, not the necessarily short summary description down in the DQA box.
3 Likes
This discussion was getting pretty far afield from the original one so I moved it to a new thread.
2 Likes
More evidence that the wording needs to be clarified, e.g. “Same species shown in all photos.”
We shouldn’t be having this kind of problem but we are, and it could be prevented.
2 Likes