- “verifiable” should be unchecked by default. if anyone disagrees, let’s use donations to decide. tug-of-war is another option but there are a few logistical obstacles. voting should never be used for anything even vaguely important.
- inat should remember our filter preferences. and all of our other preferences. i can’t imagine anyone disagreeing. surprise me?
I disagree. I like good observations, at least in the Explore page.
9 Likes
maybe it’s different with your specific nature interests, but i just finished looking at all the observations of ficus auriculata in america. there were numerous useful observations that hadn’t been verified yet. some of these trees even had labels on them since they were in botanic gardens. i really don’t see the benefit of throwing the baby out with the bath water.
1 Like
Verifiable means has all relevant metadata, is wild, and has not been made casual for other reasons, not that the observation is necessarily not verifiable.
I work with wild plants both professionally and for fun. I also cultivate plants. When I am looking up where plants grow I have no interest in what people grow in their gardens (botanic or otherwise) it’s more of a descriptor of what nurseries and landscapers do vs any pattern of biogeography or habitat preference, and not valuable to understanding plants on their own terms. I think that sorting out plants that are naturalized vs cultivated is one of if not the biggest problem with quickly using iNaturalist to glean useful distribution info. I do not want to see captive plants by default, and really don’t want to mess with messed up metadata plants. As long as I can sort plants as captive and filter with that I’ll annotate and accept that garden plants will be a major part of iNaturalist.
I still think there is use in these captive observations though. If your Ficus example there are wild growing populations noted in California but not Florida (quick search of one source so others may differ). Most of the Florida plants here are certainly cultivated (some not marked as such) but there is at least one that looks like it may be naturalized. Is this a future invasive? Should Floridians be watching out to see if it is spreading on the landscape? Should someone write a distribution paper? The messy data makes it difficult to even see that this may be something to investigate further, but having the captive plant data lets people know where to look for those plants that may be on the verge of escaping into native landscapes.
People use this site differently and fortunately the site is robust enough to handle a lot of users who use it for different things.
6 Likes
It depends on what you mean by this. I often have multiple Identify tabs open at once for doing different things, and I often have different setting on each page. If iNat tried to remember one particular set of filter preferences across all my usage, it would probably go insane, and I would be highly frustrated. So, remembering for a given tab? I guess so - but it does that already. Remembering overall? No thanks!!
3 Likes
‘Verifiable’ isn’t a term I’ve ever really thought much about in the filters before, so it’s interesting to know what it means. But it leaves me a little confused - how does it differ from widening a search to include ‘Casual’ observations?
Verifiable is checked by default because it encompasses observations that iNaturalist as a platform prefers. That maybe be a bad wording, but @ddubois2 seemed to cover that point in their reply.
As far as point two, there are other feature requests on this forum for “sticky” preferences, and that conversation could be continued there.
3 Likes
Verifiable means eligible for research grade, whether at research grade or a needs id state.
Casual is everything else, “un-verifiable”, could be due to being captive, or missing a date or location or a photo or any of the other requirements to be verifiable.
3 Likes
google’s default is to filter out “unsafe” search results, but it tells you this. so you know that your results have been filtered. and then it’s up to you to decide whether you leave the filter or remove it. this is the only example i could think of where a filter criteria is applied as the default. all other search systems, that i know of, show you all the results by default.
here on inat, i wasn’t informed that captive plants weren’t included in the results. and why would i assume this to be the case? the very point of filtering is to decide for yourself how you narrow the results. it really isn’t intuitive to assume that by default, you have to click on the filter to broaden the results. so i had to repeat several recent searches and it was a big waste of time.
my very 1st searches were for ficus racemosa and ficus auriculata in the himalayas where they are native to but also cultivated. in most cases there isn’t a clear distinction. was a tree near the village planted or did it grow from seed? i don’t think that even most villagers know the answer. i already published my results, and now i have egg on my face, because i failed to assume that inat was filtering the results for me even though i really didn’t ask it to do this for me.
iNat makes it clear that the focus is on wild. You can upload Not Wild, but varieties and cultivars are not the focus here.
Why would you assume, without checking the facts?
May we have a link - Ficus in the Himalayas sounds interesting.
2 Likes
I would say it is the responsibility of people using data to make sure that they know what the data they are using represents. It is easy enough to see what your search includes if you open up the search filters in Explore and see what boxes are checked. A more efficient way to download data for lots of observations from iNat would probably be the API, which as far as I know requires you to specify exactly what results you want to include.
Google performs all sorts of filters on your search results without asking that it does not always explicitly tell you about unless you check your settings. It may prioritize results based on language, location, your search history, sponsoring, etc.
8 Likes
just now i searched for ficus dammaropsis within the united states.
No results found
there’s zero indication that a filter is being applied. there isn’t a red 1 on the filter button.
when i click on the filter button, and uncheck “verifiable” then there are 10 results found.
there are so many problems with this entire thing that it’s ridiculous.
- instead of saying No results found it should say “No VERIFIABLE results found because i’ve taken the liberty of assuming you’re only interested in seeing VERIFIABLE results. if i’ve assumed incorrectly, then please click on the filters button and uncheck the VERIFIABLE check box.”
- here’s the 1st unverified result. unless i’m missing it, nowhere on that page does it say anything about needing verification, or unverified. instead it says “The Community ID requires at least two identifications.” it should say something like, “Please help verify this observation by providing an identification.”
- is “verifying” about providing 2 ids or is it about an organism being wild? because right now it seems like it’s about both, which is problematic. the word “verified” makes sense for 2 community ids, but it really doesn’t make sense for wild.
- instead of saying No results found it should say “No VERIFIABLE and WILD results found because i’ve taken the liberty of assuming you’re only interested in seeing VERIFIABLE and WILD results. if i’ve assumed incorrectly, then please click on the filters button and uncheck the VERIFIABLE and WILD boxes.”
- in this case there should be a red 2 on the filters button. at a glance it should be obvious that 2 filters are currently being applied… verifiable and wild.
the thing is, i’m not a noob to search or filters. i’ve programmed filters before. what i’m saying is really basic, so i know for a fact that this is not a case of user error.
and since i’m a smart guy, i know for a fact that i can’t be the only user who has been tricked at least once by inat. plenty of other people must also be interested in seeing all the relevant results regardless of verification or wild. but when numerous relevant results were excluded, this was essentially inat lying to these users. except, most of these users never realized that inat had lied to them, since most people typically only conduct one or two searches every blue moon. just like how most users only submit a few observations. it took me several searches before i started to suspect that inat wasn’t being perfectly honest with me.
a big aspect of user friendliness is being honest. if inat is applying a filter by default, it should be upfront, direct, and obvious about this. it should be crystal clear to even the dumbest user that inat is excluding relevant results.
The observer used DQA to say Not Wild, this is a cultivated plant.
The information is there, if you scroll down to look for it.
3 Likes
Relevance is in the eye of the beholder. I have zero interest in someone’s house plants or garden, but if something has escaped into the wild, now that could be interesting.
8 Likes
Labels are not always correct.
1 Like
I will agree with one of your points that when on the main explore page and verifiable is selected a number 1 should be shown that observations are being filtered.
currently, this is empty until any other filter is added or selected
But again, as many of your points this should be a feature request and not a complaint.
I expect the reason this is not the case is because verifiable is the default and all other filters affect that. Someone else could explain better but coding reasons.
1 Like
Your first suggestion seems reasonable–at the very least, it would be nice if the “no results found” text indicated that the “verifiable” filter could be unchecked to find unverifiable results. This seems like a low-friction change that would improve the site.
Your second suggestion does not seem reasonable. The wording of that particular phrase has been carefully analyzed and is the result of more than a decade of thought. “Please help verify this observation by providing an identification,” or anything along those lines, is likely to prompt new or inexperienced users to add IDs using the computer vision or poor guesses, one of the major pain points of iNaturalist. There is a bit of a learning curve on purpose.
Your third question misses the point. “verifying” is about getting at least a 2/3 majority agreement in user IDs, but “verifiable” is used to described observations that are not casual. Casual observations are observations missing important metadata or observations of captive/cultivated organisms, which the iNaturalist organization has decided, with over a decade of consideration, fall outside of the primary scope of its mission.
Your fourth point seems like a rephrasing of your first point. I will go ahead and say the more language you put about taking “the liberty of assuming,” the less likely it is to be adopted by iNaturalist, 1) because words take up space and attention on the screen and because 2) that phrase sounds like it was written by someone exasperated by the site.
Your fourth point is reasonable but is unlikely to be adopted. Of the almost 400,000 people who have offered identifications on this site, almost no one has been upset by this. Overwhelmingly, we’re looking to identify wild organisms (as wild organisms are the stated priority of iNaturalist), and we’re wanting to identify things with appropriate data (photos, sounds, dates, locations, etc.). Automatically filtering out observations that don’t meet those requirements is frankly awesome and saves a lot of time for, I reiterate, hundreds of thousands of people. I would be mildly annoyed if there was a little red 1 or 2 next to the filter icon every time I opened up identify, and changing the GUI on the identify page to fit the preferences of a very few identifiers is not likely to happen.
3 Likes
- …if anyone disagrees, let’s use donations to decide. tug-of-war is another option but there are a few logistical obstacles. voting should never be used for anything even vaguely important
A few thoughts:
First, when donations are used to decide, that undeniably falls under the definition of a vote (e.g., “a usually formal expression of opinion or will in response to a proposed decision”–Merriam-Webster). Except now, it’s an inherently biased and unfair vote. This seems to be a straightforward self-contradiction with your claim that “voting should never be used for anything even vaguely important.”
Second, Boaty McBoatface is not at all evidence to your claim that “voting should never be used for anything even vaguely important.” For instance, voting is an incredible tool when you’re in a tightknit community of people acting in good faith, which iNaturalist is (in general). Boaty McBoatface’s problem was that the poll was open to the entire internet, who were incentivized to be humorous and were completely divorced from the consequences of the naming. Voting in iNaturalist works more like if the team of scientists, sailors, and manufacturers voted to decide the name of the boat.
Finally, in iNaturalist’s current model, voting is not used to decide anything. Rather, feature requests that receive a lot of votes are given attention by the iNaturalist team, who then internally decide what to do. The iNaturalist team is never beholden by votes, and they can choose to answer a feature request that has not received many votes. Votes are an extremely efficient and convenient way to gauge broad support for a particular feature request. Donations are an extremely inefficient way to gauge broad support. For instance, a few users could potentially outvote the tens, if not hundreds, of thousands of identifiers who disagree but don’t have any money to compete in a donation war. That would be incredibly inefficient and counterproductive to iNaturalist’s mission.
8 Likes
I change what I identify frequently. Keeping my latest search would not be helpful to me at all. For the same reason, I don’t much care what the default is; I’ll be changing the filters often anyway. So this isn’t a feature I’d vote for. If the current system were changed, I wouldn’t be too upset about that, either.
4 Likes
I disagree. And I don’t understand where this sentiment that programs have to be intuitive and learning how to use them should take no time at all comes from when we don’t hold physical tools to the same standards.
Just like everything else, getting familiar with iNat takes practice.
For physical products, you have user manuals and usually, you read them (at least in part). You don’t expect every button to come with a sticker describing exactly what it does.
In the same way, iNat has a lot of wikis, guides, faq pages and other things where information is available.
6 Likes