I am Curious as to what your thoughts are on this, I feel like there are times when one ID should not be good enough to get it to research-grade, especially when the observer adds a tentative ID first then an ID is added that the observer agrees with the ID without knowledge of what they have decided to, I know that I am guilty of this on occasion with some taxons. I am not saying that there is something wrong with Inat, I am simply wondering what the thought is behind two IDs being needed for it to achieve RG. Thanks to anyone who takes the time to respond.
4 Likes
Iâve been involved with iNat for less than a year, but one thing Iâve already learned is to NOT take research grade very seriously.
There are lots of observations that ARE research grade, that shouldnât be, and there are lots of observations in limbo (casual, or at some in-between taxonomic level that nobody looks at) that SHOULD be research grade, but arenât.
So ânumber of species observedâ and metrics like that are a kind of indicator or trend, but they should be taken with a grain of salt.
You will often see taxon specialists talk about âcleaning upâ their taxon, because so many observations are mis-assigned or mis-categorized. So they will go through every observation and try to sort it out.
But, this is not a criticism of iNat, because it is at least as accurate, if not MORE accurate, than any museum collection or âofficialâ source of data.
30 Likes
I fully agree with @AdamWargonâs reply above. Also, 2 agreeing IDâs is probably the best option. Having it at 1 would of course not work, and having it at 3 would require many more identifiers, which is a lot to ask for taxa that donât have many.
Also, if you feel like the evidence given is not enough to confirm the species even if the observation is RG, you can check âYesâ here at the bottom of the DQA:
4 Likes
There are many taxa and many parts of the world that simply do not receive a lot of attention. Observations of these taxa / in these regions would never reach RG. Itâs also worth noting, RG does not mean the observation cannot be subject to further review. Observations often get corrected, even if years later.
11 Likes
Because consensus is better than disagreement. However, just because multiple voices state the same doesnât make it correct, i.e. âargumentum ad populumâ.
https://forum.inaturalist.org/t/rename-research-grade-discussion-and-polls/590 where only 7% voted for RG.
Even I, as a non-scientist would not describe - two people agree - as research! I would prefer that ( = two agree) to be labelled CID. It is up to a taxon scientist to decide if the particular two who agree add up to âresearchâ.
2 Likes
Unless youâre a taxon expert and can categorically say you know for a fact nobody can identify the sp in the observation, you shouldnât be using that portion of the DQA to remove a RG assigment.
If you think the observation has reached research grade without enough basis, put in a disagreeing genus level ID with a comment.
You can withdraw it later if one of the existing identifiers on the obs returns with a comment giving information for why they put it at RG based on their expertise using details that werenât apparent to you, or when your disagreement is overturned by followup IDs (if weâre talking about a taxon that is likely to get followup IDs).
By using the DQA in the way youâve mentioned, the observation will be pushed back without anybody knowing, the observer and the identifiers will never see that this has happened, you wonât get any notifications about further IDs to rescind the DQA later down the line, and the onus will fall on somebody else in the future noticing something is wrong to counterbalance your vote.
6 Likes
I think you misread what was said. If you tick Yes rather than No, the only negative consequence is that it stays in Needs ID until someone else realises thatâs why the observation isnât RG - which isnât ideal but isnât the same as disappearing into Casual or changing the ID.
1 Like
Iâm a bit unclear about your question. A lot of people seem to be responding to your title but ignoring what youâre written in your post.
I feel like there are times when one ID should not be good enough to get it to research-grade
That is true, and thatâs how the system currently works. That is the answer to why two agreeing IDs are needed for research grade.
1 Like
Yes is of course better than no in this case, but itâs still not the right way to contest an ID for the reasons I state in the last paragraph. If you think the ID is not correct, then adding a disagreeing identification at a higher level will serve the same purpose, without the downsides.
I also think that while there are cases when it can and should be used, those should be reserved for when the user understands all the downsides that come with it. I donât think itâs a solution that should ever be blanket recommended to others like it was above, because they will just unknowningly be creating more work for other identifiers in the future at best (if they press yes), and inconveniencing both observers and idenitifers (if they press no).
2 Likes
I definitely agree that it shouldnât be blanket recommended, and I canât imagine doing it myself. But if youâre uncertain about the ID, but not enough to actively disagree with it, I can see that clicking âyesâ does make sure that someone else will need to look at it for it to get to RG. So Iâm not convinced that it should be blanket forbidden either.
2 Likes
Actually, checking the box âID cannot be improvedâ automatically subscribes you to the observation (at least when using the identify module), so you should get notifications for any IDs added at a later date.
I have seen cases where someone adds a disagreeing ID to an observation because they are skeptical about the current ID but donât know enough to actually evaluate whether it is correct or not. This is NOT useful to anyone, particularly if the disagreement puts the observation back at a fairly high level.
If you think an observation has become research grade too quickly on the basis of uncritical agrees, there are other, more productive ways to respond â you can ask the other users about their ID, you can tag another specialist in the taxon, etc.
Disagreeing on principle because you donât want the observation to be RG when you canât evaluate the ID yourself should not be the first way to deal with this.
5 Likes
CID = Community ID?
1 Like
Yep! it simply gives that obs the deeper scrutiny it supposedly deserves, allowing any extra identifier to intervene (confirm/infirm), eventually - if and when they want.
Thus, no need for pushing in-person a disagreement at some tricky level (with unfathomable consequences, I still havenât figured the whole disagreement algorithm); no need for much energy and time arguing with anyone (diplomatically or not⌠assuming there is some language in common). Just push it back to âneeds IDâ for further eyeballing by dedicated experts, and forget it. Or forget it not, if interested in the outcome - e.g. by âfavorite-ingâ that obs.
(My emphasis)
The problem is that a lot of people who use this DQA to make an observation âneeds IDâ do precisely this â they check the box and forget it and then you end up with observations that have half a dozen IDs by experts who are trying to make the observation RG and do not understand why it is not. So it wastes a lot of peopleâs time and it does so in perpetuum unless someone happens to notice that the observation is not behaving as expected and knows enough to check the DQA, which is very much not intuitive and fairly hidden on both observation pages and the Identify module.
Whereas a comment or tag communicates the issue effectively and in many cases sorts out the situation fairly quickly. We have words â why not use them? The various buttons available on iNat are not meant to replace actual discussion, merely to streamline certain processes. But they are not words and they are therefore limited in what they can communicate.
(A common language is helpful but not necessary â I have been involved in discussions in multiple languages where each participant writes in whatever language they are comfortable with and uses an automatic translation service or their own rusty skills from their schooldays to understand the other peopleâs comments.)
10 Likes
Donât call it âresearch gradeâ. I canât think of a better moniker for now, but there should be - maybe something probabilistic. How about âmore likely than notâ? :-)
2 Likes
There are topics on this [Does âResearch Gradeâ actually mean anything?](/t/does-research-grade-actually-mean-anything/37293
2 Likes
Didnât iNat require 3 agreeing IDs in the early days to reach RG but it later was reduced to 2? I dimly recall reading that somewhere.
1 Like
Not everyone has the time and linguistic proficiency and social/diplomatic skills to deal with identifiers. Just like thereâs an âagreeâ button (not requiring details and explaining and reasons), thereâs a âdisagree with the current state of affairs but donât feel like guessing which language to use and which diplomacy to enactâ checkbox. Sort of. âSneakyâ maybe, but what other solution?
(edit: as for âwastes a lot of peopleâs timeâ, Iâm not so sure whether additional confirming IDs are wasteful - I remember it was discussed at length on other forum threads)
1 Like
But thatâs not true - only one agreeing ID is required, which seems to be the main point of the question. Someone (often the observer with only the CV to guide them) must firstly suggest a tentative ID, and it then only takes one agreement to achieve RG status. So I suppose it comes down to asssessing the relative value of the original suggestion versus the subsequent agreements. Currently, every vote counts exactly the same on iNaturalist, which is very hard to argue against, given the general ethos of the site. Equality tends to encourage participation and continued engagement.
2 Likes