This might be very similar to a request that was made before (https://forum.inaturalist.org/t/optimize-the-user-interface-for-annotating/7662), but I’m not entirely sure exactly what was suggested before. What I’M suggesting is this:
Instead of having a drop-down menu to choose a value, and then allowing people to agree or disagree, just have a table with yes/no checkboxes (or thumbs up/down icons). Here’s what it would look like, with the O’s representing a checkbox:
Yes No Cannot be determined
Flowering O O O
Fruiting O O O
Flower budding O O O
Male O O O
Female O O O
I think this should be easy to implement, as the checkboxes would work exactly the same as the checkboxes for the Research Grade Qualification table. There are a few advantages to this:
Fewer steps- You just need to click once. Right now you need to click twice. It doesn’t seem like a big deal, but every click saved makes the experience a lot better. Besides, it’s only two clicks if you actually click the right thing. I can’t tell you the number of times I’ve meant to click “adult” and accidentally clicked “egg” or whatever.
Data- Wrong annotations are very annoying right now. For example, let’s hypothetically say that this observation actually showed a luna moth egg, not an adult. All I can do right now is to disagree that the life stage is an adult, I can’t actually label it as an egg. People might not make that particular mistake very often, but for other things they might (like sex, for example, which is sometimes a lot less clear). Having the checkbox system makes it clear what the strength of the consensus is for each trait.
It’s easier to add multiple annotations- I often add multiple annotations, such as life history, dead or alive, and then sex. The problem is that the current layout changes when you add or take away some annotations. For example, adding “flowering” for “plant phenology” adds another row of “plant phenology”. That is EXTREMELY annoying when you forget to account for that and you were about to click the next annotation when suddenly you’re clicking the wrong thing because the whole table shifted down one row
also, please move the location of the annotations table (in the desktop version of the site). One issue is having to constantly scroll up/down, which makes me dizzy (no, I will not reduce the scrolling speed! I refuse! lol…), but the most important issue is that adding an ID makes the community taxon area change size. I often add an ID, and then the whole annotation table area moves just as I’m about to click on an annotation. It is extremely annoying. Personally, I’d put the community taxon box right above the ID’s, or at least move it below the annotations, but adding a placeholder image/element before there is a community ID would work too. If it’s the same size as the community ID element, updating the graphic wouldn’t move everything else on the page around.
Sorry for the obscenely long post… now you know why I’m not active in the forums.
I’m intrigued by the checkbox system, but as far as efficiency I think using the Identify page to quickly annotate observations is a good method. Although that usually means on is only adding annotations, and not IDs as well.
If you’re using the Identify interface then much of the benefit of advantages #1 and #3 can already be obtained by using the keyboard shortcuts to apply annotations:
Shortcut
Effect
p then l
Add “Plant Phenology: Flowering” annotation
p then r
Add “Plant Phenology: Fruiting” annotation
p then u
Add “Plant Phenology: Flower Budding” annotation
l then a
Add “Life Stage: Adult” annotation
l then e
Add “Life Stage: Egg” annotation
l then j
Add “Life Stage: Juvenile” annotation
a then a
Add “Alive or Dead: Alive” annotation
a then d
Add “Alive or Dead: Dead” annotation
a then c
Add “Alive or Dead: Cannot Be Determined” annotation
s then f
Add “Sex: Female” annotation
s then m
Add “Sex: Male” annotation
The shortcuts are quicker to use than multiple clicks and you can apply several in succession if the observation justifies this. You may see the interface response lag somewhat, but so long as each combination was transmitted to iNat, they will be recorded. With practice, you can be pretty confident about when to double-check and when you’re safe to move on.
Your benefit #2 raises an issue about how disagreement is handled for annotations that only allow one value. Where an annotation allows multiple values (e.g. plant phenology) other iNat users are free to downvote my assessment and add their own.
EDIT: And in Identify mode you can also use keyboard shortcuts to navigate between tabs, between images and from one observation to the next.
Yeah, I can see how using the identify page can sometimes be a good option. I should probably start using it more. However, I like adding things as I go, I don’t particularly want to JUST add annotations. I think providing people with flexibility is important, and the checkbox system wouldn’t affect those who like using the identify page (is there anyone who prefers the drop-down menu??)
Anyway, many of the observations I identify/annotate are ones that showed up in my feed, so going to the identify page doesn’t make sense.
I’m sure this is an issue with my browser settings or something, but for some reason the keyboard shortcuts don’t work well for me. I can add the observation as a favorite, but when I try to add annotations it just opens up a “find” search.
The drop-down menu sucks (there, I said it). In general, a drop-down menu is meant to conserve precious screen real estate, at the expense of being tedious. That is the exact opposite of what’s required to annotate observations.
Yes, but the bolded part doesn’t seem to be working in the few cases where I wanted to do that. All one’s able to do is to disagree. Maybe I’m missing a step though.
Fair enough, I didn’t notice that caveat. It still leaves life stages for animals. Maybe that should allow multiple values. Maybe there should be a feature request for that…
People have been arguing for and against multiple annotation values for Sex and Life Stage for as long as I’ve been on the forum. Staff has made it clear that would go against “an observation is for one organism”, so they will not allow it. Personally, I think that rule is outdated and doesn’t reflect how people actually use the site, but I’ve given up arguing that point.
I think iNat has a built-in source of inconsistency in that annotations are applied to observations and not to photographs:
An iNat observation is intended to be of a single organism, seen at a specific place and time.
Annotations are structured as if they apply only to that single, observed organism.
iNat observations allow multiple photographs to be associated with an observation.
In practice the photos associated with an observation may each contain between zero and an uncountable number of organisms. If we’re lucky, there is a single, target organism that appears in all the photos, but there are lots of common scenarios where that doesn’t happen (flocks of birds, plants with habitat).
Annotations are surfaced in a couple of obvious places: phenology graphs and photo gallery filters, but because annotations are applied at the observation level and photos may show different individuals, there is no reliable way to approach applying annotations that won’t fail some of the time.
The result is that we get a bunch of photos with misleading annotations (flower buds labeled as fruiting, males labelled as females) because these were applied based on the overall observation. And we get phenology graphs that include some observations with annotations reflecting multiple organisms (less of an issue).
There is a dead feature request to add annotations to individual photos. As it is, it can be difficult to find the actual picture of … fruiting, among the flowers and leaves and landscapes.
Show annotation values as buttons [ ] without any dropdown menus.
Only takes one click per annotation.
This is not a change in data structure, just accessibility.
Example with Larva and Organism selected and confirmed:
Alive or Dead:
[Alive] [Dead]
Evidence of Presence:
👨🏻Organism [Agree👍(1)] [Disagree👎]
Evidence of Presence:
[Gall] [Molt] [Scat] [Track]
Life Stage:
👩🏾Larva [Agree👍(2)] [Disagree👎]
Sex:
[Male] [Female] [Cannot Be Determined]
Or alternatively:
Resume
Alive or Dead:
[Alive] [Dead]
Evidence of Presence:
👨🏻Organism
[Agree👍(1)] [Disagree👎]
Evidence of Presence:
[Gall] [Molt] [Scat] [Track]
Life Stage:
👩🏾Larva
[Agree👍(2)] [Disagree👎]
Sex:
[Male] [Female] [Cannot Be Determined]
(With representing user icons.)
(Unselected cells, such as the “[Alive] [Dead]” cell, should have a grey background.)
Other QOL features:
Resume
@dep’s other points:
2. Taxon is determined by voting. Tags are determined by observer. But Annotations and Observation Fields are determined by whoever comes first - that part doesn’t make sense.
3. Currently when you add an Evidence of presence, there’s this one annoying second before a new row appears. Solution: Just add the new row immediately, but greyed out until the server confirms it.
On the Android app, please add an Annotation menu on the bottom of the observation-creation page.
Please add bulk annotation of the Observations of others. We can already bulk annotate our own observations.
• If I bulk annotate Larva, and some of the observations are already marked as Adult, just ignore my Larva anno for those, or count it as a downvote for the Adult anno instead.
• On the Android app, let us press and hold an observation to start selecting multiple observations (the same way you select multiple objects in most apps) to bulk annotate.
I also have a number of ideas for making Annotations and Observation Fields more useful, rather than easier to assign, but I’ll post those later in a different topic.
Organism 
Male 
representing user icons.)