Research Grade vs. Verifiable


This could turn out to be a feature request but I wanted to post here first to see if I’m missing something.

There’s a checkbox on the observation page that declares an observation to be “as good as it can be.” That’s a powerful feature that assigns Research Grade to the observation via an algorithm. I don’t claim to fully understand the algorithm but I’d like to be able to use it in a search. For example: Select all red trilliums (Trillium erectum) in VT in 2019 that are either Research Grade (explicitly) or Research Grade according to the algorithm. Is this possible?

“Research Grade according to the algorithm” seems to be a happy medium between Research Grade and Verifiable. As I understand it, an observation with a single, unsubstantiated ID may be Verifiable. Unfortunately that’s too lax for my purposes.



“As good as it can be” will move an observation to Research Grade if it has a CID at Genus level or lower. It wouldn’t be a Trillium erectum observation though, but a Trillium observation. The below link shows such observations.

Verifiable means that the DQA does not have any flags against the observation that would prevent it from becoming Research Grade, such as “organism is wild”, “Date specified”, or “Location is accurate” being flagged as no. shows an example that the exact Trillium is in dispute, but does have a CID of Trillium, thus a Genus level Research Grade.


Technically verifiable is this, plus having a date, location and media attached to it. Records uploaded without one of these are not automatically flagged, only a human act can add a DQA flag.

Thanks for pointing out that my example was faulty. I’ll try to provide a real example from now on.

How do I obtain a single result that is the union of the previous result and the following result?

I assume the latter includes infraspecific taxa as well but I’m not sure.



“As good as it can be” will move an observation to Research Grade if it has a CID at Genus level or lower.

Actually is it subfamily or lower. Checking the box only matters for RG from the genus to subfamily levels. Anything that meets RG requirements at species or lower will achieve RG based on ids alone.

How do I obtain a single result that is the union of the previous result and the following result?

You can cut down on your results by selecting setting the Rank Low option to genus. That will give you all RG Trillium that are at genus level. You can also enter Vermont in the location box.

Oh wait I missed your link just like the one I described. Unless there are words in the tags or descriptions (which most observations don’t have, or at least aren’t use consistently or reliably), I don’t think you can exclude other taxa from your results.

No observation can reach research grade via an algoritm, only via human input. I guess you can say the calculation that scores and tallies the human input is an algorithm, but no status of rg can be achieved without a minimum of 2 people providing input.

Using the hrank=subfamily and lrank=complex filters should in theory show you the records at rg due to use of the no it cant be improved toggle.

For example

Specific groups can be excluded via the without_taxon_:id parameters that query above excludes plants.

I can’t think of a use case or even way to force a species level record to rg via the it cant be improved toggle, but maybe there is some combination of inputs that would allow it.

1 Like shows Research Grade Trilliums down to the infrahybrid level on 2019-06-10 (chosen so I don’t have to wade through pages to determine if it worked) in Vermont. It is all selectable via the filters interface, so the url doesn’t need to be saved, just examine the filters to determine what to do for various uses.

The computer does flag observations for Date specified, Location specified, Has Photos or Sounds, and Has ID supported by two or more when they are lacking with a resounding red X that is not open for voting.

Thanks for the info/correction.

1 Like

Wonderful! Thank you @clay_s!