Requirements before identifying?

At one point I read a suggestion that people not be allowed to identify until they had posted X observations. Did that ever happen?

I want to know because I’m trying to recruit some plant people to make identifications, and it won’t go well if I tell them how to do it but have left out a step like posting 50 observations.

5 Likes

It doesn’t seem that way to me. I’ve seen users who have like 10,000 Identifications and 0 observations, although the users I’m thinking of are also curators so maybe the rules are different to them.

3 Likes

There is no requirement. However, it would be interesting to see if the community would want to make new users wait a week or so before being able to identify.

8 Likes

Thanks! I’ll step up my recruiting efforts now.

5 Likes

I’m not sure if making an observation requirement is the way to go, but maybe making it necessary for new users to actually read the guidelines and protocols of iNat would be a good idea.

Sometimes I have wildly wrong IDs assigned to my observations (as in, totally wrong continent, with a species that is only vaguely similar), and usually when I check those it’s a brand new user.

17 Likes

I agree with this and not just for new users. The site is full of both little-known and unspoken rules to the extent that many experienced users operate under their own unofficial, bizarre policies. Some kind of unifying baseline would be a good idea.

I think for the sake of rando users wanting to ID yard flowers with the app, perhaps a good compromise would be allowing observations with no restrictions, but if you want to ID anyone else’s observations, there should be short training ala the discobot in the forums. Some sort of dummy observation that has you run through some various scenarios, ID disagreement, etc.

7 Likes

There was a requirement for something re. new users, but I (typically) forget what is was!

1 Like

you have to have a certain number of observations before you can create a project, I think.

1 Like

Many of the projects under the Zooniverse umbrella have short tutorials associated with them before you’re allowed to offer IDs.

The Galaxy Zoo project is a good example of how they’re done.

Maybe something along those lines?

1 Like

Ideally something more interactive, but practically speaking yes, something like that would be a good option. Independent of gating interactions, maybe it would be nice to have more info/help buttons across the entire page. Overloading a tutorial will just make people skip it, but if there’s something short for each section, people can get help on them as needed. Like “Data Quality Assessment” already has.

2 Likes

One useful option would be for species getting IDs that are badly out of range, something like," Are you confident of your ID? X species is endemic to Y location," or, “Has not been documented in Y location.”

One of the issues with that, of course, is that there are a lot of species that are now out of their ranges, and that the documentation for what species are native, endemic, introduced, etc is not evenly applied across the iNat site.

4 Likes

I don’t have an opinion on that one. My interest is in microbial life, which is pretty cosmopolitan.

I recently participated in a county-wide bioblitz as a new user and new to the IDs of most (except birds and some plants & fungi). As a new user I found that when I submitted via my PC the location could be left off until the ID was attempted, leaving a much longer list of options (creating confusion and later corrections by others). One was a fungi which had many reports in my area but are obviously continents away. (Some helpful zealots are correcting those errors now.) So it would be helpful to either force location first or order the form so it becomes more useful.
Time and number of submissions are not requirements for expertise, which is what we need here. But having to walk back IDs on a regular basis might help to give us pause when something is outside our own expertise. Having less than a Species level ID shouldn’t be seen as less helpful - but a learning process.

2 Likes

If you know what you’re uploading adding ids first is just easier, adding precise location manually requires much more time, plus you can have observations without location at all, or date, or id. People shouldn’t blindly click on suggestions anyway, it’s a big downfall of the system.

1 Like

The pedant in me wants to say “fungus”.
If the Zealots are chiming in, the system is working as it is supposed to. ‘Experts’ come in all varieties - from people who have studied a Taxon to people who like, and are familiar with a taxon. It doesn’t matter how many observations they have, but how precise they are in identification.
In all observations, location and date are critical, and should be done first. The ID of the organism may change, but observation data do not.

3 Likes

I’ve also thought of this before. Specifically the AI (ID suggestions) could only show in-range taxa. One way could be for AI to only suggest taxa once they’ve been observed/identified in the country (or larger region). That way wouldn’t require many manual steps in defining range boundaries, and adapts to range expansions.

Adding to the idea and similar to what you said, iNat would ideally also caution identifiers if searching for out of range taxa (but they could still ID them).

1 Like

It is already set to suggest only seen nearby species.

2 Likes

That’s true by the default. There’s also a way to turn on and off the “nearby” filter.

Somehow for example, many identifiers will ID bees from US in Asia and vice versa (e.g. Apis mellifera and Apis cerana). Do you think that means the AI may be suggesting the wrong ones as “nearby” due to nearby wrong identifications?

If we talk about new observations, yes, it’s partly because of old incorrectly ided observations and partly because when cv is not sure what it sees it starts showing all the species it thinks it can be, from everywhere.

1 Like

If I recall, the ‘found nearby filter’ is pretty new.

Before, cv suggestions mostly had ‘found nearby notations’, rather than a filter. It was sort of easy to overlook the notations.

2 Likes