An idea for a new identification/agreement system

I would like this to be a feature request, but would first like to get feedback.

I feel the current system of identifications, agreements and comments is very hard to track especially on observations with lots of some or all of the above, and very especially on obs with taxon swaps, e.g. https://www.inaturalist.org/observations/10827632 . Trying to find the ID one wants to agree to is a time-consuming exercise.

I also feel that the Identifier and the Agree-ers are two different things and should be separated. This will also limit the number of people who agree compulsively just to go up the leaderboard.

I am very aware of how hard the devs work, and the absolutely amazing functionality of this fantastic platform. I am not in any way intending to be dismissive of the current setup. This is just a suggestion from a userā€™s POV for perhaps streamlining the visualisation of the huge amounts of data.

So my proposals are thus:

  1. Separate IDs from agreements-to-IDs and compact the ID section.
  2. Separate the comments section from the IDs section.
  3. Add a confidence level to IDs.
  4. Add an Other Observations section under the IDs of thumbnails of other obs with the same community ID.
  5. Remove the blocks of thumbnails from the bottom of the page (Seen Nearby, Other Obs by so-and-so etc.) - these take up bandwidth and most people donā€™t even know theyā€™re there anyway, and I donā€™t believe I have ever gone to look at them on purpose.

So, I envisage blocks of info (on the left of the page) like so:
IDpagelayout2

And it would look something like this:

And here are some notes on the layout:

Note: I have left out the right-hand side of the page because that works fine as is, and reduces the size of the images uploaded on this comment.

When someone adds an original ID, I think a confidence level should be mandatory - something like Maybe, Probably, Certainly. Just like when someone adds a higher level taxon and the reason choice is obligatory.

Then, on the Identification leaderboards, only original identifications are counted. Agreements would be a separate thing and can have their own leaderboard, if necessary.

If someone tries to add a duplicate ID (rather than an agreement) a popup tells them that the ID already exists and if they agree with it they should hit the Agree button. One can only agree with one ID on an ob at a time.

I would also be very keen to change Research Grade to the simple Likely ID.

Note: I have not used the apps and donā€™t know how stuff differs, or if any of this could be incorporated in them. These suggestions are based purely on my experiences on the website.

7 Likes

I use these, so I hope theyā€™re not removed. Mostly I use the one for other observations seen by the user on that dateā€“it sometimes helps in figuring out what the intended subject of an observation is when there is more than one possibility, and sometimes I just want to see what else was out there.

3 Likes

I also use this a fair amount, as well as ā€˜nearby obsā€™ as it often helps with an ID. but it might be an idea to hide/view these according to user preference (i.e if you hide them once they stay hidden until you unhide them).
I like IDs and comments staying inline, often comments flow from the ID sequence. The IDs can get in the way of comments though, so if one could have a toggle to hide IDs that might help. It would also help follow the conversation if one could reply to an ID or a comment.
(As an aside I find it frustrating when a user goes and adds the nth confirmation of an obvious ID like ā€˜elephantā€™. Why do that, it isnā€™t needed. Sometimes I wish iNat had a cut off for the number of agreements to an ID.)
I like Karoopixieā€™s other suggestions, esp recording a level of certainty in an ID.

5 Likes

Just my 2 centsā€¦

Ideally this is not the intention. The agree buttons are provided as a convenient shortcut to add oneā€™s own independent identification when it happens to match someone elseā€™s, without having to re-type. Unfortunately they donā€™t always get used that way (for reasons discussed in many other threads). But I imagine there would be a lot of pushback if that convenience got removed or re-purposed. In my opinion it would be preferable to re-name the Agree button to something that better indicates its intended purpose.

4 Likes

I add a duplicate ID quite regularly because clicking on the agree button does not provide the ID text box in which I can make comments on why I am agreeing, what features I am keying in on. I could, under the proposed system, add a comment separately, but that explanation would spatially disconnected from my agreeing ID.

I also concur with @vynbos that comments flow from IDs, comments would lose their context in this arrangement. I view IDs and comments as a community conversation. If I am jumping in to contribute, I should know what has already been discussed. The result can be messy, and suboptimal for high volume IDers, but I find value in the conversation that IDs and comments make.

7 Likes

I can imagine a ā€œsummary boxā€ at the top of the comments/ID section would help a lot, but as you have it in the example, I can imagine people thinking it is similar to the forum where you can ā€œLikeā€ a post, and click it to see who else has likedā€¦

I also agree with the other comments about losing the context or flow of ID and discussion. I see it as being a group of people standing around something found on a bush walk, and part of that is someone saying ā€œI think it is aā€¦ā€ and then the discussion over that suggestion.

Iā€™m personally not fussed over the thumbnails at the bottom of page, wouldnā€™t be upset if they were gone, but neither do they bother me (even re bandwidth)

1 Like

I often ā€œagreeā€ to say ā€œyes this looks like a fitā€, after doing due diligence to see the species is in the area and the pictures look similar with some confidence. But Iā€™m not nearly skilled enough to independently identify, i.e. to say ā€œthere is nothing else in the area that looks like this and might cause confusion.ā€ So Iā€™d argue the distinction between ā€œagreeā€ and ā€œidentifyā€ is real and useful.

2 Likes

Functionally there is no difference though, so you agreeing ultimately has exactly the same effect as an expert identifying. To fix that weā€™d either need to eliminate the ā€œagreeā€ category you describe, or make a reputation system so that you can ā€œagreeā€ without affecting the community ID as much. I think the majority of users prefer the former option, which is what @jdmore described.

2 Likes

Thank you all very much for your responses!! It is so wonderful to have all your comments and a clearer idea of how people use the site. It has given me some ideas of my ownā€¦

I will come back with amendments to my proposal soon - it takes a while, especially to set up the graphics, so I thank you for your patience :-)

2 Likes

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.