Ok, I knew about the sound recordings, but I was paraphrasing. I think my point stands. An observation without a photo/audio/other supporting evidence will be treated as casual. Why should a similar observation that has a poor photo included (ie. one that doesn’t support a species level ID) go to RG just because the photo doesn’t preclude the proposed species level ID? Aren’t the two scenarios roughly equivalent?
Non-expert (and sometimes expert) users add incorrect IDs all the time. Explaining why those IDs are incorrect is a lot of work, and half the time, the explanations are ignored. Enlisting other knowledgeable IDers to outvote the incorrect IDs is a lot of work as well. Yet here we are, worrying about knowledgeable IDers bumping IDs up to higher levels for observations that lack sufficient evidence to support a species level ID. The first is a problem we live with, the second is a non-issue that’s is being used to argue for what I think are unnecessary changes.
In my taxon/region, there are very few people bumping IDs to higher levels. By and large, only a handful of knowledgeable IDers are doing this, and we typically provide explanations when we do so. If there are corners of iNat where non-expert users are bumping IDs inappropriately, then I think the solution is to address the behaviour of those users, rather than taking away the tools that knowledgeable users need to do their work. If some people are running with scissors, the solution isn’t to blunt all the scissors (unless maybe you’re running a kindergarden). Some of us have work to do and we need proper tools to do it.
i’m not sure why you think anyone is going to take away any functionality from the system. please scroll up a few posts to my last post, and look at the link that i referenced there. that’s the latest plan from iNat staff related to all this. please read it, and tell me where there’s any mention of any functionality going away.
I think this argument is running around in circles. I don’t think anyone would want observations that don’t have enough evidence to support the ID go to research grade. But some folks can’t stand leaving them in Needs ID either. Since a frequent concern for unsupported use of the disagree button appears to be to keep observations from reaching research grade and/or making them casual, maybe a DQA item for “insufficient evidence to confirm” or something like that would help. If enough people marked that, it could go casual. Forcing these to a higher level without evidence supporting a disagreement is a poor work-around and I don’t think how iNat is supposed to work.
But that’s beside the point of the discussion in this thread, which started out focused on the poor wording of the ‘orange’ and ‘green’ options and the apparent disconnect between what it says on the button and what clicking it does to the community ID and the disagreement message generated on the observation. However, fixing the wording doesn’t fix the behavior that has developed around the use of these buttons. This is something I was told on another thread: The result is what counts, no matter what the words on the button say.
The argument that it takes too much time/effort to post an explanation for bumping something back to genus always seems a tad on the lazy side for me. iNat is a social media site along with being a repository for nature observations, and for me communicating with other users is a big part of it. So what if it appears the observer is gone and unlikely to come back? Then think of it as communicating with the next ID’er looking at that observation.
I read that. The whole discussion around ancestor disagreements is confusing (I know, I know, they went to a lot of trouble to illustrate things using examples). My understanding is that the proposed solution is to offer a 3rd disagreement option, and we will still be able to bump an ID up to a higher taxonomic level when the evidence does not support a species level ID. If that understanding is correct, then I’m fine with that.
But the impression I’m getting is that some people are unhappy with that solution, and are arguing that iNat should limit the ability to punt IDs to higher levels or perhaps they are arguing that IDers shouldn’t use the option if it is still available. So count me as one vote for iNat to stick with the plan and not cave in to pressure to limit our ability to punt IDs to a higher level.
i don’t like the idea of encouraging people to knock things back to genus or higher if they can’t tell what it is. I think a better idea would be to make the ‘no, it’s good as it can be’ button work differently. Reword it to ‘can not verify this species with available info’ so that it can’t get research grade unless it gets outvoted. And allow it to work with just one ID. Riht now it’s too complicated - if i see something that i think shouldn’t be research grade to species, but i don’t want to knock it back to genus, i have to add a non disagreeing genus level ID AND click the button. It’s confusing.
I disagree. We are “forcing” observations to a higher level “without evidence”, we’re putting the observations at the lowest level for which there is supporting evidence. To my mind, it’s the user who has applied the lower level observation who is doing the forcing - they are trying to force the observation to a lower level than what is supported by the available evidence.
Your argument seems to turn on species level as being the ultimate goal. You suggest that if the evidence doesn’t support species level, then the observation should go to casual level. The evidence provided may be perfectly good for IDing the observation to a higher level (eg. genus), and that hiigher level observation may be useful to someone. In fact, that higher level observation may contribute to the accumulation of information that can eventually lead to us finding ways to better identify “problem group” observations down to species level, at which point it may become possible to refine that ID. Moving the observations into the casual category is a bad idea IMHO, but I suppose it may be appropriate in some cases.
I don’t think I made that argument. Please re-read that paragraph. What I’m arguing is that lots of people suggest incorrect species level IDs ALL THE TIME. It’s a lot of work to make the corrections, and provide explanations. That’s a problem we live with, and nobody is arguing that we need any changes to address this problem. People enter IDs that are DEAD WRONG, and we deal with it. We patiently provide corrections with explanations. We treat it as part of the process.
So why are we having this discussion around punting IDs up to higher levels? I can’t speak for other taxa/regions, but in mine, this doesn’t happen very often, and when it does, it is typically done by knowledgeable users (one might venture to call them experts), who usually provide explanations. If we can live with the first problem (a much bigger one IMHO), why are we unable to live with the second? I’m not sure I get why some folks agonize over:
a) an ID that isn’t at the species level
b) an ID that remains at the “Needs ID” status
I spend a fair amount of time (pretty much daily) explaining why I am punting IDs up to a higher level. I probably spend much more time explaining why species level IDs are incorrect, and a different species level ID is more likely to be correct. To my mind, these are just two different aspects of the ID process. Like I said, the first is pretty much a non-issue as far as I’m concerned. But I imagine there are taxa that are much more problematic, where it’s much more difficult to ID observations down to species level, and it’s probably frustrating to see a lot of observations lingering in the Needs ID state at higher-than-species level.
One way to do this kind of thing now is to use the Favorites feature for this purpose. (Not the “list of favorite taxa” feature, but the “favorited observations” feature. I wish these two features had different names.) I mark observations as favorites if they contain useful discussion that can help me identify other observations in the future. But it would be similarly useful to use it to mark observations that you want to refer back to for other reasons.
I’m here because another thread was merged into this one.
Suggestion
The wording of the “Potential Disagreement” prompt is incredibly misleading!
It’s in dire need of rephrasing into something like:
Thank you for your suggestion!
Do you think the evidence presented in the observation could allow a more specific identification by someone in the future?
Yes
No
Bold formatting for emphasis of prompt elements I consider important. With formatting itself not necessarily being part of my suggestion.
Critique
As it currently stands:
Is the evidence provided enough to confirm this is %PRIOR_TAXON%
I don’t know but I am sure this is %CHOSEN_TAXON%
No, but it is a member of %CHOSEN_TAXON%
The prompt is:
Complicated by verbosity, stuffing giblets of the intended question into the answers.
Inconsistent, since the single subject of an observation would have to be a “member of” a non-species taxon in both cases but only one option uses that wording, leaving the user to wonder if this is the dubious distinction being inquired of them.
Confusing, as the options read like two phrasings of the same exact sentiment if you approach it as a communal scientific effort. The “No, but” of the latter option is easily interpreted as still implied with a degree of uncertainty, especially after being primed by the “I don’t know” wording of the former. See gripe #1.
Furthermore, there are cases where you expressly intend to disagree with a more specific prior identification, are sure of a less specific taxon belonging and do know that the evidence present is enough for further identification to a more specific taxon but aren’t able to provide such yourself.
The technical effect of choosing the former option does what you want but its usage of “I don’t know” comes into conflict with your position.
Neither of the options is phrased to fit this scenario notionally and leave the user (that’s me ) in an irritating stalemate.
I believe that my suggestion at the top addresses all of these issues with proper conveyance of reasonable degree of confidence, while also signalling the technical substance of the decision being made instead of inferring it.
P. S.
I also believe it’s better to deal away with coloring of the buttons altogether, and I get that this is a vibe-based opinion, but it makes the user worry that their suggestion will be rejected by the platform if they display misunderstanding of the rules they suspect having missed due to the sheer mind-boggling confusion of this misleading prompt.
Also, I just changed “down the line” into “by someone in the future” as it conveys what I meant with the former but I now realize that the idiomatic nature of the former wording may require unreasonable linguistic aptitude of some of our multicultural platform’s users.
I agree that the current prompt is confusing, but what you’ve proposed seems to be addressing a separate issue. Your wording sounds more like what the DQA bit about “Based on the evidence, can the Community Taxon still be confirmed or improved?” is trying to ask. What the orange / green disagreement buttons are really asking is not whether a more specific ID is possible, but rather whether the existing more-specific ID should still be considered viable. A person might choose the hard disagreement and not know if a more specific ID is possible or not, they just know that it’s not %CHOSEN_TAXON% and want to kick it up a level.
I think that the one that causes a disagreement should say “The evidence is enough to be sure that this is not %PRIOR_TAXON%.” As it is (and it’s been this way at least since Sep 2020), it’s impossible to tell from the wording which choice causes a disagreement, and I’ve hit the wrong button and had to withdraw the disagreement. I had identified something as a dicot, noting that the flowers look beanish (but they were somewhat different from beans I know). Someone else identified it as a species of indigo. As indigo is a bean, I revised my ID to Faboideae, not intending to disagree since I don’t know indigos.
I agree that the current wording is confusing, but this proposal would not cover cases where there is not enough evidence to confirm that it is %PRIOR_TAXON% and the ID needs to be pushed back to a broader level not because it is definitely not that species, but because there is no way to be sure whether it is or it is not.
For those of us who work on taxa that are difficult to ID from field photos, this is a very common situation (by contrast, if there is enough evidence to know it is definitely not that species, I find that often this means there is enough evidence that I can suggest a different specific taxon rather than a broader one, so the new ID is automatically a disagreement and the message does not come up).
Thought experiment: design and usability aside, a gradation of choices with more explicit (?) wording and hints. A big fat form titled “Give your opinion about the current identification for this observation (Taxonus specificus)”, and at least four options offered:
simple agreement – “I agree, to me this is indeed Taxonus specificus”: [Agree button]
agreement + refinement – “I agree this is Taxonus specificus, and I even think it could be: [dropdown list of enclosed finer taxa e.g. subspecies]”
soft green doubts no matter the reason (limited knowledge or inconclusive evidence) – “I am not totally sure whether this is Taxonus specificus or not, but this is at least: [dropdown list of enclosing coarser ranks e.g. genus, family, phylum]”
hard orange rejection (thanks to contrary evidence) – “Considering all the evidence (media, location, date, comments), I think this is not Taxonus specificus but rather: [input field for any other ID at any rank]”.
Well, the problem in modern day biology is that we can never be sure of anything any more. I used to confidently ID certain plants down to species, which were then split up in recent publications so now I’m not sure any more. What can we still be 100% sure of? Nothing, really. “Being sure” becomes subjective - for some folks it’s maybe 95% probability, others want to be 99.9% sure. I don’t see a “need” for pushing everything back to genus or higher.
The problem I have with the current system is that the wording on the button suggests that you can’t confirm it is that species based on the evidence, while the result is a hard disagreement on the observation page saying that you are SURE it is NOT the suggested species but must be something else instead. “Can’t confirm this” is different from “can confirm it is NOT that.” I always ask when someone does that on my observations and find in many cases the person isn’t actually trying to say the ID is wrong but just wants to say there’s not enough evidence. So there’s a disconnect between intent and result and it’s misleading and confusing. It’s even worse when someone e.g. disagrees on a flowering plant species by bumping it all the way back to Plantae. It effectively counts as saying it can’t be anything in Angiospermae and must be something else in kingdom Plantae.
If there are two species that are both reasonably likely and the details needed for ID are not visible, any species ID is going to be arbitrary. How is it not justified to push such observations back to a higher level? This isn’t simply “we can’t be sure of anything in this global age and x feature that is traditionally used for ID isn’t visible to exclude some species that has never been documented in this area”, but rather, there are known to be multiple similar species in the region that are difficult or impossible to distinguish from the average field photos (the situation for a lot of arthropods).
In many cases, IDing to species without evidence results in a vicious circle of people assuming that there is only one common species because all observations have been ID’d as such, even though there is at least one other species that is not rare, it merely isn’t as well known.
Yes, disagreeing to a high level creates problems, but it has nothing to do with disagreeing with observations that should not be at species level because there is no evidence to support such an ID. The only way to change the ID for observations where there is not enough evidence is to disagree (orange button), regardless of what the text that accompanies the button says. It indicates “I disagree that it can be identified as that species” (not “I believe it is definitely not that species”). Adding a non-disagreeing ID does not accomplish that, because in most cases the observer leaves their unjustified species ID and sooner or later someone agrees with it because they don’t see any reason why it couldn’t be that species.
Yeah, i feel like this has been unclear for a while and causes problems. For instance i had two hemlock observations someone tried to knock back to ‘Pinaceae’ because they couldn’t tell if it was hemlock. The photos were blurry, yes, but it’s an enigmatic species and i though the evidence was more than sufficient and they were overreaching. People seem to use it when they can’t tell, and unless they are the sole expert on that taxa or it’s a very dramatically impossible situation like something that says ‘hemlock’ and is a literally smooth blur with no shape visible, i don’t think people should be knocking back things so readily. Data quality is important but people seem to also not understand that othres are good at blurry photos of distinct species. Sometimes it’s even used as a comment on photo quality, like the ‘frassing’ in bugguide. That’s not what iNat is about and I hope it never is, I don’t use bugguide for a reason.
… to their taste (and eyes, and knowledge, and grasp of probabilities).
I mean… sure, it might be [insert some similar-looking species from another distant continent, never observed at or near that location over five centuries, and with a 0.0001% chance of ever landing and thriving in that place if eventually carried by a disoriented African swallow]". But hard-disagreeing, seriously?
Because what I see with arthropods (where there are many perfectly realistic scenarios in which there are two or more similar species that are sympatric and they can only be reliably distinguished some portion of the time, e.g., only in one sex, or only adults, or only if the head/legs/butt etc. is visible) is that 99% of the time IDers are not arbitrarily pushing observations back to genus/subgenus/complex in cases where there is a vanishingly remote possibility that it could be some other species, or because they think the photos are too poor, but because they know that there is not enough evidence to support a species ID based on the current scientific consensus about how to distinguish that group of organisms.
In fact, most of the time what I see is that arthropod IDers are often reluctant to push observations back to a higher level because they find it discouraging to do so and because it makes observers unhappy. As a result, there are any number of genera where the species level ID means essentially nothing, because a large percentage of the IDs have been chosen arbitrarily without any compelling reason; as a consequence, the distribution and frequency data provide misleading information about the individual species. Explain to me how a meaningless species ID is better than a genus level ID that realistically reflects our level of knowledge?
Too often to my taste, but I deal mostly with the (peculiar, impoverished) flora of a location most iNat identifiers are unfamiliar with. I understand they push back IDs in good faith, because that is the reasonable thing to do in their part of the world, and/or they don’t know other literature and id keys beyond theirs. It’s fine, and can be discussed - which not all IDers are ready to do unfortunately, for lack of… time? notifications? anything else?
Opting out of Community ID can help in some annoying cases; and if you ask me, hard-disagreeing should be made less trivial, perhaps through stronger incentives to offer at least a reason no matter how poor (“unknown in the country”, “petals color”) when hard-disagreeing.
) in an irritating stalemate.