multiple individuals observing the same organism

[cray man]

Today, five people observed the same crayfish held by the same person and apparently all identified it with the identotron, 4 to one species a 1 to another species, all of which are wrong.  What a heaping mess this is.  Any ideas for dealing with the multiple observer issue?

[Upupa epops]

Creating the option to link shared observations so that a change in community ID on one will affect all of them. :)

This would work for multiple people viewing the same organism at the same time, as well as observations of the same organism throughout the year(s) (for example the same tree as it grows and changes through the seasons).

[Charlie Hohn]

Being able to link interactions was listed as a priority after the iNat retreat, hopefully that would include some of this too, but of course programming that sort of stuff takes time so in the mean time it's tricky. I guess you just have to ID them all correctly, what else can you do?

[Chris Cheatle]

Interesting, I interpreted "linked observations" in the list to mean cases where you are documenting the same organism over time, not sharing observations between users. Perhaps if Carrie sees this, she can confirm the meaning.

Functionally, any shared observation process will need to include:

- a way to "request" or question users if their observation is the same as observation X, and if so get permission to merge.

- some means to deal with when what is clearly the same thing is entered, but under different ID's. For example, right on on the Identity pool, the same dead rodent is submitted at least 10 times by different users, under 3 different ID's

There needs to be some means to centrally manage sharing records. Let's be honest, if it takes 20 seconds to do an observation on the mobile app, but 2 minutes to ask "what's your user name, and what's yours, and yours ?" etc, and 2 minutes to type all that it, then I'd wager 95% of the time it won't get done.

[Carrie Seltzer]

There are 3 ways in which observations could be linked:

1. 2+ users observing the same organism at basically the same time, resulting in 2+ observations of the same individual.

2. Separate observations of 2+ species interacting, which could be linked with a defined set of interaction fields.

3. Separate observations of the same individual organism over time (by 1 or more users). 

During the retreat we mostly talked about type 2, but also type 1, since each of us made our own observations of many things we saw together. I usually have one of the worst cameras in the group and would love to easily link my poorly photographed (bird) observations to someone else's much better documented ones. 

We don't have clear plans or timelines on these types of inter-observation linkages, but will let the community know when we do. In the meantime, I think the best course of action is to identify them individually like Charlie said. 

Best,

Carrie

--
You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist+unsubscribe@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

--

Carrie Seltzer

Stakeholder Engagement Strategist

iNaturalist

California Academy of Sciences

[Charlie Hohn]

in terms of multi user observations, I think it might also be able to nice to tag more than on person on an observation without creating another one. if i am with Ken Ichi and he has a nice camera and i have my old iphone, and we both see a bird, i'd rather just be tagged into his observation rather than have to take a blurry bird photo myself. But, either way is fine really, if we get there.

To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.

To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

[Lincoln Durey]

Carrie, here is a big "Case 1" example:  When I was in the Jack-in-the-Pulpits dividing out the
Arisaema quinatum 5-leaved Jacks, I found this one plant photographed 8 times by 8 observers in a bio-blitz:  (note all timestamps are between 11:50 and 12 noon)
https://www.inaturalist.org/observations/5089814   RG:3
https://www.inaturalist.org/observations/5089816   RG:4
https://www.inaturalist.org/observations/5089826   RG:3
https://www.inaturalist.org/observations/5089837   RG:4
https://www.inaturalist.org/observations/5089838   RG:3
https://www.inaturalist.org/observations/5089840   RG:3
https://www.inaturalist.org/observations/5090047   RG:3
https://www.inaturalist.org/observations/5091739   RG:4
(we have enough views there to generate a 3-D model and print one!)
So I had to pester a lot of people a lot of times to get those flipped into RG as the new species, esp. b/c many of the blitzers are no longer around...

Now, I'm reasonably sure the AI can't be asked to link these, b/c they don't all get a species level ID immediately, and that would be the only way to match them up...,
But we humans can certainly do this.

I'm imagining that a single user's single observation remains inviolate, but that if we see a blitz-like observation like this, we make a new "thing", a "pooled/group observation".
you hit a button, and enter a list of of observation #'s, and these are checked by the computer to ensure they are:
1) within a 1hr time span of the others in the group,
2) within a 50 meter (pick something) range (to account for GPS variances, like those is TX)
(There would also be a note "Please be SURE to link ONLY if you are certain this is the same physical organism")
Each observation (on iNat) would get an Is_member_of_Group=Group.id number (most would be empty)
When someone views an observation, they have a button to "See the Group Observation" this belongs to, that would go to a page showing all users who contributed to the observation, all the observed images on one page, with links to the member observations, and (if we got one added wrongly), a way to remove a member observation.

Another use of this Group Obs, could be for time-lapse growth observations.  Take a new observation of a growing plant every day, and then you can add each subsequent new obs to your Grouped Obs, and then people could see the development over time.

We could also use this to solve the 1 user makes 5 observations of one thing, one image each, we add them all to a Group Obs, and they all get the community ID of the consensus of all IDs given each.

[Charlie Hohn]

i would definitely use some sort of grouped observation like this to track the phenology of plants in our field. It would be neat.

[Colin Purrington]

There's one particular Cuscuta in LA that is extremely photogenic and it would be great to have a tool to link all the observations in some way. Currently ID'd by observers to two different species, sometimes just to genus. In addition to reducing wasted ID effort, the linkage would add a nice social factor for people who hike in the same area.

[Charlie Hohn]

ow wow we can have famous plants. There is one sycamore in Vermont that is huge and visible along a road and on the edge of its range and it has at least five or six observations. it can be 'the' Vermont iNaturalist tree

[Charlie Hohn]

here's another one

https://www.inaturalist.org/observations/9506035 etc

ID is probably wrong but more importantly it's a planted tree and none of them are tagged as such. Also no photos.  sigh...

[Cullen Hanks]

Hi Carrie et al.

#1, linking observations, or sharing an observation with another person would be huge.

1. Often documentation is a team effort.  I will often work with another person to find and photograph an animal.  It may be my photo, but it was our observation.

2. It is a great recruitment tool.  Sharing observations with new users helps open the door to participation.

Of course, I see the utility of all three.

Best,

Cullen

To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.