If there is an observation where the sequence of IDs has been:
ID 1 (User 1): Species A
ID 2 (User 2): Higher taxon with at least one intervening taxonomic level (family, order, etc.)
ID 3 (User 3): Intermediate ID for a group that Species A is in
then if User 2 realizes that ID 3 is more precise and clicks Agree, the system gets rid of the hard disagree at the previous level and does not give User 2 the hard/soft disagree choice when they agree with ID 3. As a result, Species A goes back to being the leading ID. It seems like the only way to avoid this is for User 2 to withdraw their ID and type in their own version of the intermediate ID, which will then give them the hard/soft disagree choice.
Is this working entirely as intended? If so, is there some way of publicizing the most efficient way of retaining the hard disagreement with the initial species ID?
This happens when you leave a new ID without manually withdrawing the old ID first. I’m not sure if it’s intentional behavior, but it’s easily avoided by first withdrawing your initial disagreement and then leaving the new ID, which will prompt you if you want to disagree again. My guess is that when you’re making the new ID, the system is checking based on the current community taxon, and not what it would be after your initial ID is withdrawn, so it does not detect the potential disagreement.
Thanks for linking to that old thread…I didn’t find it when I was searching earlier.
In that case, I know the Feature Request needs to be something more specific than “make it not do this anymore”. Does anyone have thoughts on what would be a relatively uncomplicated change that would address this problem?
The two possible approaches I can think of (but don’t know how complicated they’d be) are:
Ask anyone making an ID upstream of the leading ID whether their ID is a hard or soft disagree with the leading ID rather than comparing it to the Community Taxon. This would also address the issue described in 2019 of User 2 withdrawing or deleting their ID and turning User 3’s ID into a soft disagree by default. But is there always a leading ID on every observation? And would it be a big philosophical change in how ID disagreements work?
Calculate what the Community Taxon will be without the user’s previous ID and use that to determine whether the hard/soft disagree question should be asked. I can imagine there might be issues with asking the system to split the instant of clicking “Agree” up into calculating the Community Taxon minus the user’s previous ID and determining whether the hard/soft disagree question should be asked before the new ID takes effect. And whether User 3’s intermediate ID was a hard or soft disagree with the species/leading ID would still be ambiguous.
It is not a solution, but since iNat is ‘unpredictable’, I always check that my ID has both the effect I intend, and that I expect. Or I persevere till iNat agrees with me.
