"Disagreements" Filter added to Identify Page

News and Updates
, ,
1

Here’s a copy of the post I just added to the iNat Blog!

Hello, identifiers (and aspiring identifiers) of iNat! Thank you for all of the essential and difficult work you do on iNaturalist.

As a way to make your valuable identification time more efficient, we’ve added a small but helpful Disagreements filter to the Identify page, which you can access by clicking on the Filters box:

4540-original.png1418×876 116 KB

When you check that Disagreements box, the Identify page will show you only observations where at least one active identification is an explicit disagreement, such as:

4542-original.png1454×732 110 KB

Why add this filter?

This will help you find observations where more informed community input is most beneficial, at any taxonomic level. 

In the example above, the observation now has a Community Taxon of Class Arachnida due to the two disagreeing identifications, and would benefit by more people seeing the observation and weighing in with an informed opinion. If two more people agreed to the Harvestmen (Opiliones) identification there, for example, the observation’s Community Taxon and Observation Taxon would change to Harvestmen (Opiliones) and be more easily findable by someone who specializes in that taxon. It’s like a relay race, we’re doing everything we can to pass the baton to the next person.

In testing it out over the past week, I’ve found that for me it’s most useful when combined with other filters that restrict results by location and taxonomy. For example, restricting results to California shows me observations where I have enough knowledge to weigh in with an informed identification or comment. I’ve also found it useful when going through my own observations to see where I need to update or withdraw my identification.

Technical Details

If you’re familiar with editing search URLs, this filter adds the param disagreements=true to the URL. Other possible values are false and any (which is the default).

We hope this new filter will help you when you’re looking for observations to identify! Let us know how you’re using it or if you have any feedback. Thanks again for making iNat what it is!

44 Likes
2

Fantastic feature! I’ve used it to help go through my own observations and take another look at those which had disagreements and slipped under my radar.

I’ve also used it to look through observations at high taxonomic levels with disagreements- many of these are easily identifiable or can at least be refined, but have gotten lost among a sea of observations that are harder to identify.

11 Likes
3

Is there a way to view observations that are stuck at a rank above species due to disagreements where you can search for a specific species included in the disagreeing ID’s? Obviously, as it stands the only way to come across that record after there is disagreement is to find it under the higher taxon rank. I would love to be able to quickly find records that include someone ID’ing it as a specific species so that I may agree or disagree. From what I can tell so far, this feature just gives you everything that has disagreements under a higher taxon and I would have to do further sorting myself. In any case, this is still a great feature! Thanks.

2 Likes
4

i can see this as being useful the first time you look at a particular set of observations – say, by a particular location or by a particular taxon – but what happens if you look at the same set a week later to look for new disagreements?

for example, if there were 15 observations with disagreements found the first time and 16 found the second time, is there a good way to figure out which observation has the new disagreement?

i know that there is the reviewed flag that can be used, but what if the new disagreement happens on an observation that you’ve previously reviewed? or what if you didn’t mark all the observations as reviewed the first time you went through them? is sorting by date updated the best way to sort of get an idea of what might be new?

i’ve been using my own thing to periodically find disagreements in my area (based on GET /v1/identifications in the API). I think it handles that sort of maintenance workflow well, but it’s always been limited by its inability to filter by observer and project (since it’s not possible to filter identifications by observer or project, although it handles taxon and place just fine).

i like the fact that the new disagreements filter parameter on observations can be used with tons of other filter parameters (including observer and project), but i’m having a hard time visualizing a good workflow for using it to periodically find new disagreements in a set of observations.

i think the only reliable way to do exactly this is to search identifications, not observations. for example: https://jumear.github.io/stirfry/iNatAPIv1_identifications?per_page=100&disagreement=true&observation_taxon_id=47157&taxon_id=514918&observation_lrank=complex. however, this method comes with some of the limitations that i mentioned above.

2 Likes
5

Good questions! If a particular observation is of interest, add an ID, even if it’s merely a supporting ID. Otherwise mark the observation as reviewed. I don’t know if that’s optimal but at least it keeps you in the loop via notifications. That assumes you haven’t turned off notifications, however.

6

Does marking as reviewed (without adding an ID or comment) really mean you get notifications?

1 Like
7

Just marking as reviewed doesn’t subscribe you to notifications on an observation, you have to select to “follow” it too.

4 Likes
8

You could try sorting by ‘date updated.’ Presumably the new disagreement would be the observation that has been most recently updated, or at least one of the most recently.

4 Likes
9

Yes, you can replace taxon_id= with ident_taxon_id= in the search url, which will give you all observations in which at least one person has suggested that taxon, even if it’s not the record’s observation taxon. If you’re searching at species-level, then also set the lowest rank filter (lrank=) one level up from species (i.e. Complex) to ensure you’re only getting records with disagreements. Works on both Observations and Identify pages.

5 Likes
10

while i can’t think of a better option than sorting by date updated either, i don’t think it would be reliably true that the observation updated most recently would be the one with the new disagreement. (after the disagreement, you could get other identifications, or other updates like annotations, projects, observation fields, etc.)

i think this is just a limitation of filtering by observation using a single binary filter parameter though. it’s better than nothing, i guess.

2 Likes
11

Excellent tool, thank you.

Also please consider letting people who identify but doesn’t upload observations change example photos of species.

12

I’ve granted you the privilege of adding photos and common names.

2 Likes
13

Sorry, let me try to be more precise. Do one of two things: either add an ID or mark the observation as reviewed. If you add an ID, the observation will be automatically marked reviewed, but you will receive subsequent notifications (unless you’ve turned them off). If you simply mark the observation as reviewed (press “R”), you will never see it again (unless you override the default filter of the Identify tool).

2 Likes
14

Very true, I guess I was thinking about it in the context of checking for disagreements very frequently. But an update could be just about anything.

15

Great improvement. Is there a way to exclude settled disagreements (i.e. those where enough people gave the same correction, successfully overriding the observer’s wrong initial ID)? I’d love to look for my own errors (and maybe (rarely) where I’m the one who thinks somebody else is wrong), but https://www.inaturalist.org/observations/identify?reviewed=any&quality_grade=needs_id%2Cresearch%2Ccasual&verifiable=any&place=any&disagreements=true&ident_user_id=ralfmuschall also gives me dozens of observations where observers thought harvestmen were spiders and we fixed that (e.g. https://www.inaturalist.org/observations/261914508, but I want to keep e.g. https://www.inaturalist.org/observations/111922397). Removing research from quality_grade didn’t change the result.

2 Likes
16

If you were looking for disagreement at a consistent level, you could exclude anything with an ID below a certain level (order/family/genus etc)?

17

I rarely do fine IDs anyway. What I want is to see observations where my ID disagrees with a later one by somebody else (or the spider thing where I supported spiderness by saying Theridiidae and hope to see what species it is (maybe Steatoda something, but it is too blurry and has no location)), and I want to suppress the other way around (particularly if ID1=X and ID2=ID3=ID4=Y and I am one of 2,3,4 (like the harvestmen example)).

1 Like
18

Maverick or Pre-Maverick ?

19

I’m happy to be corrected, but I don’t know of any tool where you can filter by order of disagreement, sorry.

20

there’s not really a human interface within the system, but there’s this:

it doesn’t get exactly at ralfuschall’s thing exactly, but you can view identifications by identification date.

1 Like