Issues with the "Can the community ID be confirmed/improved" question

[Tim. Reichard]

I want to point out some shortcomings of the following question on the observation details page and suggest and idea to potentially fix it.

Q1: Can the Community ID still be confirmed or improved?

Issue #1: If I'm entering the first ID for an observation and I know its evidence cannot support a species-or-lower ID, then I want to indicate this when I enter my ID.  Then I can help it reach Research Grade with a higher-than-species-rank ID, its ultimate status.  Since my ID is the only one, there is no Community ID, so the question as stated can't be answered.  But since I want to indicate that the ID can't be as fine as species, I answer Q1 = No anyway.  The result is that the status changes from Needs ID to Casual, which should not happen.  The observation should still be kept in the Needs ID pool so other identifiers are likely to see it and help it reach Research Grade status.

Issue #2: When there is a Community ID after I enter my ID, Q1 is answerable.  But the Community ID could change later, and my answer to Q1 does not.  So later my answer to Q1 could change the status between Needs ID or Research Grade when it should not, even though my answer did the right thing when I entered it. With the current system, I would have to check every observation for which I've answered this question to see if my answer is still correct with respect to the current Community ID or if others have properly overruled it if necessary. And that's not feasible.

Issue #3: I'll give an example.  I enter a genus-rank ID and I know that a species-rank ID is not possible.  The Community ID was at the order rank and remains so because my ID is the 2nd ID (order + genus = order).  Q1 is asking if the order-rank community ID can be improved.  It can be improved.  But I can't answer the question of whether a genus-rank ID can be improved.  There is no way to indicate this without somehow remembering to revisit the observation later (days, weeks, years later?) when the community ID has become a genus-rank ID.

An idea to fix this:  Introduce a new question that would be attached directly to the identification and answered (or ignored) by the identifier:

Q2: Can the community improve upon *my* ID?  Yes, No, or No Answer.

And each identifier can answer this question with respect to their particular ID. 

At any time the quality grade is computed, for each Yes or No answer to this question (there could be several answers from several identifiers), the identifier ID would be compared to the Community ID. It would look at agreement vs. disagreement (does one ID match or contain the other) and identifier ID rank vs. community ID rank. That comparison would decide whether that identifier casts a Yes or No vote for Q1.  The votes on Q1 still decide whether to override the quality grade (no change to current behavior).

Thus Q1 remains but is no longer directly answerable by any person.  Q2 is the question that identifiers answer.  And the votes cast for Q1 now take into account the current community ID rather than the community ID at the time the question was answered. That should solve Issues #2 and #3.

For Issue #1, suppose I have entered a genus ID and Q2 = No, in hopes of allowing Research Grade status at a future genus-rank community ID.  There is no community ID, so a vote should not be cast for Q1, even though I answered Q2. The observation would have a Needs ID grade instead of Casual, so it retains a higher chance of reaching Research Grade.

Tim Reichard

[Tim. Reichard]

Here's a new example where answering the "Based on the evidence, can the Community ID still be confirmed or improved?" question changed the quality grade from Needs ID to Casual when it should have kept it as Needs ID.
http://www.inaturalist.org/observations/7165820
At the time of writing, there are two IDs: Order Lepidoptera and Genus Crambus. The second identifier (for the genus) answered No to the question. But the No applies to the order, not to the genus. This changed the quality grade to Casual, decreasing the chances of becoming Research Grade with a genus ID.

The No answer would have made sense if it could have been applied to the genus ID. We need a way to do that.

Furthermore, the change to the Casual quality grade removed the observation from the project it was in.
http://www.inaturalist.org/projects/howard-county-parks-moth-inventory
But instead, it should remain in the project with a Needs ID grade.

Tim

[Scott Loarie]

Hi Tim,

Thats because the community ID is still at Lepidoptera. So the community ID can still be improved (ie bring it forward to Crambus). I'd hold off on voting 'no' until the CID is as fine as it can be

--
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.

--

--------------------------------------------------
Scott R. Loarie, Ph.D.
Co-director, iNaturalist.org
California Academy of Sciences
55 Music Concourse Dr
San Francisco, CA 94118
--------------------------------------------------

[Tim. Reichard]

Scott,

Yes, it is working as designed.  But the design leads to unwanted side effects. So I'm arguing that the design needs improvement.  I'm curious what your thoughts would be for modifying the design.

I think the better behavior would be, assuming the CID becomes Genus Crambus at some point in the future:
1. Give the identifier a way to say "The community should be able to identify this observation as Genus Crambus but not to a lower rank," even when the CID is at a higher rank.
2. Maintain the Needs ID quality grade until some unknown future time when the CID reaches Genus Crambus.
3. Change the quality grade to Research Grade when the CID reaches Genus Crambus, without the identifier in #1 having to revisit the observation.

Thoughts?

Tim