Why does Explore use "Verifiable" while Identify uses "Casual"? Are they different?

Hello!

I am helping onboard a researcher who focuses on street tree species. This has forced me to start learning how to access and deal with Casual observations. I’ve stumbled into an issue/difference I don’t fully understand and searching around has not given me a clear answer.

In Identify, you can clearly select Casual as an option in the UI and utilize “&quality_grade=casual” or “&quality_grade=needs_id,research,casual” in URLs. This does not seem to work when using Explore, which instead offers &verifiable=any” or “&verifiable=false.

Is there a meaningful difference between a Casual observation and a “&verifiable=false” observation? What is the purpose of this distinction?

It’s quite confusing, especially given the location, time, photo, seasonality, etc can certainly be verified and revisited for a cultivated plant. In contrast, the logic that cultivated plants are Casual as a current iNaturalist category is clear. I totally understand the Verifiable hover text “Eligible for Research Grade” as a category describer, but that’s not inherently synonymous with being verifiable, as the word is commonly used. One of these trees that has been growing in the same spot for over 100 years is definitely verifiable, in the usual sense. Thus my confusion!

Thank you for your help clarifying. I want to get this new researcher hooked and want to be sure I’m passing on good info. Casual observations have value too!

no

because the Explore page defaults to verifiable=true under the hood, you need to explicitly set verifiable=any (or verifiable=false) if you want to be able to specify casual observations.

for example:

2 Likes

This might be a useful place to point them to for some definitions of terms: https://help.inaturalist.org/en/support/solutions/articles/151000169936-what-is-the-data-quality-assessment-and-how-do-observations-qualify-to-become-research-grade-

3 Likes

As you might know, multiple reasons for an observation being casual, including being captive
If you want to see casual observations that have all other DQA correct except captive then you might want to use the filters: &mappable=true&photos=true
unfortunately there is currently no way to exclude observations that are missing a date @pisum correct me if I am wrong

i noticed recently that they added a whole series of failed_dqa_xxx parameters. so you could use that to filter for missing date. ex: https://www.inaturalist.org/observations?verifiable=false&fails_dqa_date.

EDIT: sorry. i just tested fails_dqa_date, and it’s actually filtering for “date accurate”, not “has date”.

i think the definition of verifiable in here is inaccurate. it says:

A verifiable observation is an observation that:

  • has a date
  • is georeferenced (i.e. has lat/lon coordinates)
  • has photos or sounds
  • isn’t of a captive or cultivated organism

but look at what this returns: https://www.inaturalist.org/observations?captive=false&verifiable=false&id=19386749. that observation has a date, is georeferenced, has a photo, and is not captive or cultivated. according to that definition above, it should be verifiable, and yet gets picked up as verifiable=false. as far as i can tell, the converse of the definition of casual is the actual definition of verifiable.

4 Likes

Is there a way for these to be added to the https://www.inaturalist.org/pages/search+urls
Would be nice to have exclusions of these as well.

2 Likes

while there doesn’t seem to be a date presence/absence-related parameter specifically to filter out observations missing dates, you should still be able to exclude them through other means, namely setting the start date in the date observed range (d1) to something like, say, 0001-01-01. observations without a date won’t fall within any range of dates, so should be excluded from any date-based search results accordingly. (if you also want to exclude observations with messed-up dates extending back a couple thousand years, you can set it to a more realistic start date. i’m not sure there’s currently any way to do the reverse, searching specifically for observations lacking dates, though)

2 Likes

concerning https://www.inaturalist.org/observations?captive=false&verifiable=false&id=19386749

the user himself has marked it as casual because of an “inaccurate date”

1 Like

it doesn’t matter. the criterion is simply that there is a date, accurate or not. it has a date, and so by that definition it should be verifiable.

but just to prove the point, i took away downvote on date is accurate, and i downvoted recent evidence instead. you get the same result. anything that makes the observation casual also makes the observation verifiable=false.

1 Like

you can use positive filters instead of negative:

https://forum.inaturalist.org/t/large-number-of-app-users-with-date-observed-missing/11498/3

it is not. the casual is superset of non- verifiable as @bouteloua linked above , to be precise

Casual <=> (¬Verifiable ∨ DQA markings ∨ opted_out_and_CID_clash)

1 Like

if you’re saying that there are observations that are casual and also verifiable, please provide an example to prove that case.

here’s another observation that in my mind shows that anything casual is verifiable=false: https://www.inaturalist.org/observations?id=28874013&verifiable=false&captive=false

This would be SO COOL!! Maybe we should make a request thread.

About that Page

This page was a collaboration of many members of the iNaturalist community. Want to add or edit something? Any iNaturalist Curator can edit this page - the link to edit is on the bottom right. Otherwise, reach out to @bouteloua and she’d be happy to work with you on making changes or additions.

EDIT: Added here: https://www.inaturalist.org/pages/search+urls#dqa

1 Like

I know your question was directed to another user but I love doing searches. This assumes that captive should be casual - I think the following satisfies part of the query you asked : https://www.inaturalist.org/observations?captive=true&subview=map&verifiable=true
Oddly: some of them are research grade; some although they appear captive, I can’t see that being acknowledged in the DQA or elsewhere; and none of them are marked casual. Am I missing something?

1 Like

ah sorry i see your point now. the actual verifiable filter is getting affected just as casual filter in similar way from all DQA flags.

but according to help article explanation there it shouldn’t have - so not sure if its bug or how it should be and then article is wrong.

If its a bug, solving it can help inturn in this feature request i guess? - https://forum.inaturalist.org/t/make-captive-cultivated-not-automatically-no-id-needed/112
where they can use verifiable filter for IDing those, if fix can also push that non-wild condition too out of verifiable in addition to DQA conditions as is now. the casual flag can still apply for all these.

it’s likely just bad indexing. if you vote and then unvote any of the DQA flags on any of these, these should drop out of captive.

even if the intent was to do what the article was describing, because it’s such a core part of the system, i think it would be better to adjust the article to reflect what the system is actually doing right now (and has been doing for years).

1 Like