Using Identify to Annotate Observations

If I’m in the mood of not wanting to think too hard but still contribute to iNat, I’ll use the Identify page to annotate observations. Many, many observations lack any annotations and for some taxa it’s quite easy to go through them and add a life stage, sex, or plant phenology annotation without much specific knowledge. If enough observations of a taxon are annotated, we can start to get some cool data on them, eg for Monarchs (although look at how many still need a life stage annotation!):

Here’s my approach:

  1. On the Identify page, here are the basic settings I’ll use choose for life stage annotations:

The URL for this is: https://www.inaturalist.org/observations/identify?quality_grade=needs_id%2Cresearch&order_by=random&without_term_id=1&captive=false

This should show me all Needs ID and RG observations lacking a life stage annotation, sorted randomly.

  1. I’ll choose a taxon compatible with life stage, for example Lepidoptera.

  2. Then, when I click on the first observation, I’ll make sure to go to the Annotations tab. There is a handy list of keyboard commands on the bottom left-hand corner.

  1. It’s pretty easy to rapidly go through the observations and tell which ones are of eggs, larvae, pupae, or adults and add annotations for those.

Things to remember:

  • If you are unsure, it’s better to not add the annotation, feel free to move on to the next observation.

  • Don’t go too quickly, sometimes you’ll add the wrong annotation! Be sure to correct any errors you make

  • Some taxa can be tricky. For example, there are larviform female adult beetles and other insects, which are difficult to distinguish from adults. Again, if you’re unsure, don’t add an annotation.

  • If you would like to bulk annotate your own observations, there is another tutorial for that.

That’s about it!

47 Likes

A post was split to a new topic: Community Annotation Challenges

Some helpful URLs for this. This post is a wiki, so please feel free to add your own links.

4 Likes

Can I just add that not every observation needs to have annotations applied. If you find yourself struggling to decide which value to make the annotation (“Is this observation of an exuviae an adult or a juvenile?”) then perhaps just leave the annotation out.

On that note, I wonder if having a “reviewed=no” version of the links above would allow people to “skip” those that they are unsure of… perhaps they can add it to the url if they want it…

I also wonder if it is acceptable to only mark “dead”, with “alive” being assumed? That is how I have been annotating alive/dead for new observations since that annotation came into effect…

@tiwane If you would rather this not be the place to discuss these thoughts, please feel free to delete this post, I have raised the thoughts on other topics before.

3 Likes

@tiwane, I don’t know if we have a page for how newbies can learn more while adding value, but this would be a great activity for that! Almost anyone could go through a set of observations and tell whether they’re budding, flowering, or fruiting.

As for that page, I’m thinking of a list of things like adding high-level IDs to Unknowns, annotating observations, using the Computer Vision Model to refine identifications, adding Observation Fields, tagging organisms that are obviously captive or cultivated, and the like. Each item on the list would link to the corresponding tutorial.

As for the value added, I’m amazed at how powerful those graphs of phenology are. For example, if I am looking at common privet (Ligustrum vulgare) and click the square for the month of June on the curve for Flowering, iNat will produce a page of all observations identified as that species and annotated as flowering in June. (I think there’s also a way to display all observations, regardless of month, that have that annotation. The site’s down, so I can’t check right now.) I can then see if everything identified as L. vulgare produces the same kind of flowers—in other words, whether or not everything identified as L. vulgare really is L. vulgare. I’ve used that as a first step in correcting misidentifications.

Of course, if the observations aren’t annotated, I can’t do that so easily. So if somebody who knows nothing more than what flowers, flower buds, and fruit look like will take the time to add those annotations, they’ll probably learn more about the species while they make the data associated with the observations more available to everyone.

People are already using this information, by the way. One person told me that a plant had stumped them because it was in full bloom when the Phenology data showed its species wasn’t blooming at all. As it turned out, nobody had entered annotations for that species, and they had thought that the 0 for Flowering meant that the species did not flower that month.

1 Like

This wiki may be what you’re looking for RE: new users:
https://forum.inaturalist.org/t/useful-inaturalist-tasks-for-non-experts-wiki/35034

4 Likes

Is there a topic dealing with missing annotations and annotations for other groups (e.g. for bacteria the only option under evidence is: “galls” - what about “symptoms”, “slimes and mats”, microscopy?). Perhaps a link from here would help?

1 Like

https://forum.inaturalist.org/t/lets-talk-annotations/627

4 Likes

@Tiwane Who can create annotations for a particular order or phylum ? Can a mere user somehow add more choices of annotations for a particular group of organisms.
Bacteria only has the annotation, gall ! Can’t we have organism, petri dish culture, disease symptoms on host etc etc for bacteria ? Do we users flag for improved annotations for an order ? Why do some orders or phyla have alive/dead while others dont ?

You need to leave that comment on the link Marina has above

1 Like

Staff members. Annotations are meant to be iNat-controlled, as opposed to the “wild west” of observation fields. We try to keep them pretty limited and broadly applicable, and their main purpose is for taxon photo browsers and graphs.

4 Likes

@tiwane What is the reasoning for sorting the observations randomly?

1 Like

It’s certainly not necessary but I like that I’m not just annotating new observations. I like to show some love to older observations as well.

3 Likes

Thanks for the link! I might put some effort in adding alive and dead annotations to portuguese observations.
A question: When I use the Random order, only one page of results is displayed. Is this a bug?

1 Like

@joaolemoslima I noticed the one page thing too. My guess is that the sequential pages are populated by time so when you randomize the time you no longer have sequential pages. I found View More will load a whole new page. I still have over 7,000 pages of damselflies to look at. :face_with_spiral_eyes:

Fortunately I’m only looking for nymphs so it’s a quick review. But still. 7000.

3 Likes

Yes, with random there’s no pagination.

2 Likes