I appreciate my own observations being agreed with, often because I’m uncertain - and, for instance, if it’s a vascular plant I feel reassured if experts such as JoC, JoP or lavateraguy confirm my tentative identification (and, contrarily, I feel equally grateful when they correct my errors). iSpot is about lifelong learning - we all learn from the observations of others, the experience of others and the mistakes of others. And iSpot only survives and thrives, if it does, upon user engagement and participation. So I try to agree to others’ IDs when I feel I have the level of knowledge to do so and hope that others within the core community continue to do likewise.
1 Like
I see that under another topic you raised the issue of making agreements to placeholder identifications at higher taxonomic levels. I agree there is a case for withholding agreements in that situation, pending the hoped for addition of a more precise identification.
are often misspells or of known synonyms, many of which have been reassigned to other Species and sometimes they are the result of an adverse reaction from the Server,
I still ‘feel’ there is no reasoning against adding a slightly higher-tax ID (to Validate) and marking the observation with the collecting tag taxonomy1. It matters little if the higher-tax ID receives an overload of agreements.
Your “… there is a case for withholding agreements…” suggests we should never agree to anything other than Species when there is a possibility of a better ID being added.
Just to clarify, I think Lavateraguy * and I are referring to the situation where the currently accepted ID is not in the dropdown. Those of us especially dejay, give the generic or higher level id to ‘ save’ the post.
.
This is working well. However, come the new dictionary, we will want to add the accepted id. If the Higher Level id has agreements it may be difficult to shift it.
.
*This is my November 2020 forum post in Shifting Banners that Lavateraguy refers to:
I am not disagreeing with the basic tenet of adding a Genus id. where it helps.
.
My point is about our adding further AGREEMENTS to that once it is designated ‘likely’.
.
I am not optimistic about future updates in any system… a pragmatic view is that unless the programmers know about it, then it will happen.
Well, Bluebird, I love being asked to help, so I’ve added something to your Polypody post.
Well I’ve always regarded the likely threshold on iSpot as far too low, things become ‘likely’ on the basis of very few agreements. So for my own stuff I like to see enough confirmations before I consider submitting a record to the relevant recording scheme if it is one I cannot ID myself.
1 Like
I also tend to look at who has agreed. If, for instance, I posted a seaweed I’d look out for an agreement from Derek; for a gall I’d be pleased to get an agreement from Mark; etc.
It is possible we might update the reputation system, it has been modified slightly a couple of times over the years.
I have been very quietly proposing for years that the Banner is NOT awarded to ANY ID until there is an agreement.
It will produce difficulties that could be overcome. But it would prevent the several thousands of WRONG ‘might be this’ IDs cluttering the Taxonomy Browser.
I would support David and dejay in the suggestion to make the Likely Banner contingent on at least one agreement.
1 Like
I disagree on need for an agreement in every case, eg in the case of an expert like martinjohnbishop who ID’d many foreign flowers, in many cases with no agreement
ChrisMcA
“…in the case of an expert”
But It is no great difficulty finding them and agreeing IF they are agreeable.
“Best is to have a sub-curation Taxa-team whose role might be to confirm IDs to the best of collective experience” is a quote from (way) above.
Interesting. Some ideas that were floating around in the past (you probably remember them) include
a) Have a higher threshold for likely ID (perhaps so that only an ‘expert’ can make it a LikelyID as the sole identifier).
b) Have a threshold for likely ID based on the species (higher for harder groups). This is probably beyond what can be done without a wholesale redesign.
c) Allow people to disapply their expertise when outside their area of knowledge when submitting an ID. This feels really useful for invertebrates where I feel bad when the knowledge I have in some groups (say lepidoptera) gets me a high reputation that swings the ID for something I know little about (like heteroptera).
d) Allow reputation to go down if you get things wrong. This was ruled out early on as iSpot wanted a system where ‘nothing bad’ could happen to users. But I think we are now seeing people in some groups who have got a high ID by just identifying common things so some way to stop that spreading into bad IDs elsewhere could help.
It’s an idea that is floated about a bit when this was discussed before. I think the challenge may be that the ‘LikelyID’ system currently doesn’t see ‘expert’ status, it’s just that experts have a high number of reputation points which count to swinging the ID (I think it is possible to actually earn more score than an ‘expert’ if you are on here doing enough).
But if the system can be tweaked to give expert a status beyond purely the high score that would be very useful.