See my reply to one of your above posts. This is all about the benefits of removing the Agree button.
2 Likes
As I’ve mentioned previously, there is a huge problem with people piling incorrect IDs onto observations (ex: https://www.inaturalist.org/observations/2619553). So many of these will not make it to RG and will be stuck at needs ID due to the lack of IDers.
Most top identifiers are good. But, there are a few, not going to call anyone out here, whose IDs make it insanely difficult to flip incorrect observations (and then don’t respond to pings).
3 Likes
There is certainly some suggestive evidence out there, but whether that justifies the suspicions about “most people” looks quite dubious. There is a forum post in a thread about “speed identifying” that I found quite disturbing. I find it hard to see how aiming to do fifty thousand IDs in a month can be good for either the site or the individuals involved. Identification isn’t (or shouldn’t be) purely a numbers game. Although there may only be a small number of people doing this, I think iNaturalist has an obligation to explicitly discourage it (even if it’s well-intentioned). There is no problem with gamification per se - but, as with so many things in life, it needs to be done in moderation.
So I personally think the changes are aiming in the right direction, but the specific steps taken so far are probably not the best ones. However, maybe the most effective way to prove this is to treat it as an experiment and thereby show that it either doesn’t solve the problems it is supposed to, or has too many negative side-effects to justify continuing with it.
4 Likes
Can you please clarify why it takes more time?
If it is wrongly identified, you would not be using the agree button anyway, I guess unless someone has already added a correcting ID to which you cant agree.
If it is correctly identified, and you don’t want to see that record again , it is the same amount of work to click reviewed as it is to click agree. Yes, it may take a day or two to get the muscle memory going to click reviewed, but it still does what you want with the same effort.
4 Likes
Reviewed is definitely not the same as Agreed. If I agree to an identification, then should someone come along later and add an alternative identification or comment then I will not be alerted if I tick reviewed. Also I want to see if a taxonomic expert has Agreed to an identification of mine, as the identity of the people matter. And as an expert in certain taxa, why people want me to add an Agreed identification to theirs.
4 Likes
But in my case, for example, I want to mark it reviewed, follow it, and indicate to others that I have reviewed the observation. So “1 click” becomes “a click (reviewed), a slow to appear drop-down and another click (follow), a click on the comments tab, a bunch of words to the effect that I have been involved here, followed by a click to make the comment commit, and finally a check to make sure I didn’t typo anything”.
In NZ Araneae back when I started iNat there were 2 IDers, and they mostly made comments to what they thought something was, and only place an ID if they were 100% certain. We have grown to maybe a dozen active spider enthusiasts (in the NZ fauna), of which I would say there are 4 that actively review RG obs as well. We tag each other a lot, often just a “check this one out”… But also “you know these ones better than any of us” and “another one related to that review you are considering”. This sub-community that we have formed is largely due to being able to see each others activity in identifying. For me, this is exactly the “build community” mission of iNat in action, and I can only see this new change as impacting negatively on that ability to see each others activity.
6 Likes
I can respect that some folks have anecdotal experience with this problem (though many others don’t), but this still isn’t hard data on the problem as @fogartyf mentioned. iNat has literally millions of observations, posting a few links here and there isn’t strong evidence. I’m not saying that evidence doesn’t exist, but it hasn’t been presented.
Additionally, I’d like to point out that there’s often a big difference in the “cost” of problematic observations. You argue that due to poor use of the Agree buttton, so observations remain at Needs ID that could be (correctly) RG. Observations that will remain in “Needs ID” due to incorrect conflict are annoying, but, in one sense at least, aren’t particularly costly: they aren’t being used as bad data.
I would argue that the more costly scenario is incorrectly identified observations reaching RG and being sent off to GBIF. Due to the weighting of votes in community ID, even a moderate level of conflict in IDs is usually enough to remove an observation from RG. And I think folks on both sides of this “agree button conflict” will agree that reducing the number of incorrectly IDed RG observations with just two IDs is one of the biggest and most difficult to solve problems on iNat (which this agree button change doesn’t address).
Taking many posters at their word for now, removing the Agree button is going to reduce the number of identifiers looking at RG observations. This will likely result in more incorrectly IDed observations remaining at RG that could be corrected. I would predict that this change could actually exacerbate the problem of incorrectly IDed RG observations. I think that is one of the major potential negative outcomes of this change.
11 Likes
I agree that ticking “reviewed” takes the same amount of time as using the agree button, but many of folks don’t think that is an ideal solution because it represents a total loss of time/effort that go into assessing an observation’s ID (see point #1 on a previous post I wrote here).
3 Likes
I concur that people who support the change to remove the agree button but don’t feel as strongly may be less likely to chime in (which would affect the overall count of commenters pro/against). Though totally anecdotally, looking at the past couple posts on iNat announcing updates and site changes, there are lots of supportive/pro comments on those, and a much lower number of negative comments. So the response to this change does seem quite different to me.
However, even if the total number of users concurring with the agree button change is higher (due to lower likelihood to post), that shouldn’t impact the apparent differences between users for/against in bothvolume and usage of iNat in frequency of observation and IDs (unless there is some kind of interaction/bias between likelihood to post and iNat usage patterns).
I do think that higher volume IDers are more likely to be the ones reviewing RG observations though. So that could explain some of the apparent difference in experiences between users for/against the change.
3 Likes
I specifically wrote if your objective is to not see the record again.
For my purposes, one specific benefit would possibly be not being inundated with notifications on my dashboard that I don’t want, as I’ve mentioned in other topics (and I do have my settings set to not receive notifications of IDs that match my own exactly). But I’m looking forward to the rollout of the new notifications system that @tiwane wrote about in the past couple of days because that may solve that issue for me.
The other benefit for me, as someone who tries to help with observations in the State of Matter Life category, would be that if identifiers would slow down enough to be able to be responsive to comments, there would not be so many identifications that are incorrect because the identifier identified the wrong organism in the photos and pays no attention when that is pointed out.
5 Likes
Well, I have just checked on one especially prolific agreer: he has almost 258 000 IDs on his account (during 3 years). It seems he is not very daunted by the change. Still agreeing in full speed. Hopefully at least the rate of his IDs will slow at least slightly.No hope though that he will start to consider what he is agreeing with.
3 Likes
Unfortunately, posting links is about as strong evidence you can get to a systemic problem. So far today, I’ve helped deal with around 12 different problematic observations. The worst was https://www.inaturalist.org/observations/34353760. Almost all of the ones I looked at were in my specialty, birds, and only today. How many more are there in insects, plants, and other living organisms? How many have and will fall through the cracks? Not just that, but most of these birds were of only one species. There are plenty of other hard-to-tell-apart species too.
That’s a valid point, but this should (hopefully) slow down the rate at which these obs are being created.
3 Likes
Yeah… they still haven’t removed the “agree” button from the identify grid page, and that’s probably the one that’s being used the most by the mass “agreers”.
8 Likes
How often does the circumstance that this covers actually happen ? It requires:
- a record that is already RG before you look at it
- you as an expert agree that it is properly identified
- someone comes along, who is not trolling or otherwise acting in bad faith comes along and adds a dissenting ID, saying all of you are wrong.
- assuming that the dissent is not based in a dispute about the taxonomy (not the identity) in which case no number of agreeing ID’s will likely prevent them from doing so.
1 Like
Impressive. And much much worse that Gerald case.
2 Likes
It really is baffling that they didn’t consider changing the identify grid page before making the other changes. It would probably still have been controversial, but on the face of it, it seems a much more effective way of slowing down the speed identifiers (assuming that is one of the main aims). The fact that this has gotten so little attention suggests that those who are most strongly against the other changes don’t use the agree buttons on that page very much.
7 Likes
My point exactly! Gerald has done more to build community in iNat than any observation that has only two IDs make it RG!
6 Likes
Wait… so you’re saying Gerald was a good thing? I’m not completely certain I understand.
3 Likes
Yes! Is there anyone that actually thinks the Gerald obs was a bad thing?
6 Likes