Why I have stopped using eBird
-
ledzep
- Posts: 852
- Joined: Tue Jul 07, 2009 7:30 pm
- Location: Wellington
- Contact:
Why I have stopped using eBird
Apologies if I offend anyone. I was an early adopter of eBird, and I still have entries in there going back to 2008, but I have been progressively removing my logs. There is now only about 10% of my personal sighting logs which I keep separately. Why? I'm sure many others have received an email out of the blue from someone in eBird questioning an observation, or wanting the location changed. Sometimes this is useful, as occasionally I do get things wrong. I recall logging the wrong species of Thornbill in Australia once, and I've also made the occasional data entry error (White Heron instead of White-faced Heron). Being contacted about an obvious problem within a couple of weeks of submitting an observation is fine. But then earlier this year (ie 2024) I get an email about a sighting made on tour in Peru in 2012. That is 12 years ago. Earlier this week, I got an email about the Mallard's I saw in Norfolk Island in 2011 - that is 13 years ago! Not questioning the sighting, but wanting it changed from Mallard to "Mallard (domestic type)". I see there are now around 40 choices for Mallards and Mallard hybrids. I don't care - I just want to record it as a Mallard. It was a wild bird. I also log Muscovy Ducks as Muscovy Ducks, with a note that they are domesticated (I have also seen real wild Muscovy Duck in Costa Rica - for me, same species). eBird has all sorts of categories for recording species when we aren't quite sure - for example, I am not 100% sure whether all my Giant Petrels are correctly logged as Northern or Southern. I usually assume Northern unless I get a good look at the bill, but I don't want another category in my sightings log for Giant Petrel species. Similarly, I've seen Australian Magpies at Staglands that were clearly "domesticated", and the Black-billed Gulls in Wanaka scavenging chips from the locals are heading down the "domesticated" route as well, not as feral as the ones in Pauatahanui inlet, but I don't want to log all these separately. So we get to the knub of the problem - how does eBird maintain data cleanliness for data that they don't own? For the Mallard, I don't care if they want to now classify my Mallard sighting as "Mallard (domesticated type)". But instead of them doing it, they send me an email and ask me to do it. They send emails asking me to change locations to hotspots, and want further information on loggings of rare birds. We did log something like the 20th record of a Sandwich Tern in Iceland, in that case I had an excellent photo and the Iceland eBird guy was very happy - all communication happened within a couple of days of the sighting. But I couldn't find a photo of the Slaty Becard we saw with the guide in Peru in 2013, so eBird doesn't recognise it. That's fine, but I'm still logging it in my list because I assume the guide was right. We had a similar situation being questioned 8+ years after logging a sighting at Cape May observatory with an official Cape May observatory guide. I remember the situation, and the wader was a hazy dot in the distance through the scope, no hope of getting a photo even for id purposes. Whether it was or wasn't, if the guide is being questioned it needs to be done at the time of logging, not 8 years later. Who owns my sighting logs? I do. I share them with eBird so they can use them for analysis, and also for sharing with other birders. This is where eBird have a problem that they haven't yet got their head around yet - how you manage data cleanliness when you don't own the data? From my day job in the Ministry here in NZ, this issue of data cleanliness for data you don't own is a problem that we have been grappling with for over a decade, and there isn't an easy answer. So whilst I have sympathy for eBird, their current procedures for handling sightings flagged for review is far from optimal. The overhead for me constantly responding to requests to modify past sightings is too high. I maintain my own sighting logs, and I'm pretty particular about the cleanliness of my data against IOC. eBird doesn't provide any value to me by holding my sighting logs in their rather clunky "My eBird" site. If they are going to continually pester me to make changes to my sighting logs from over a decade ago, my response is to remove my logs completely from eBird. There are still some there, but in most cases I'm not recording new sightings there, and may well remove everything in the future to avoid these pestering emails. There are ways of tackling the problem and eBird has had a decade to make progress and they haven't. So I'm sorry if this offends keen eBird users. I was enthusiastic about eBird back in 2010, but time has moved on and there are other places and means to record bird sightings. If eBird wants my data, they need to offer me value and stop pestering me with annoying emails.
-
Trina
- Posts: 16
- Joined: Wed Aug 08, 2012 3:20 pm
- Location: Auckland
- Contact:
Re: Why I have stopped using eBird
You make good points. My bugbear with eBird is the obviously wrong sightings here in NZ - particularly from overseas birders triggering a new record for an area or just making a mistake with data entry. NZ reviewers do their best to get punters to change their record but many requests just get ignored. Why can't they delete obvious mistakes? or do they get dismissed as outliers somehow?
-
RussCannings
- Posts: 1241
- Joined: Wed Sep 24, 2014 5:23 am
Re: Why I have stopped using eBird
Hello to you both,
As one of the eBird review volunteers for NZ I'll try and respond to a few of the key issues raised--and don't worry--no offense taken at all--I think we're all solution-driven so feedback and suggestions should always be welcomed for eBird or anything.
1) Dealing with checklists/records ebird "doesn't own." People tend to use eBird for three main reasons: 1) To contribute to citizen science and 2) For their own record-keeping 3) To use the tools provided (e.g. Being alerted to rare sightings). If we reviewers started manually changing peoples' checklists (e.g. To correct a misIDed bird, or move a location) you could imagine how ticked off users would be. The core people at eBird do have powers to override checklists (e.g. Change your mallard to a domestic type) if they really wanted to, but the reason we ask users to do it themselves it because it needs to involve consent, and sometimes the original checklist is correct after all. By checking, this can help to 'purify' further, and it is always the right of the user to NOT change their checklist details. So in other words, you can call your mallard or sandpiper anything you want--it will remain their for your records in perpetuity (It may just be removed from public consumption if it does not adequately clear the bar for confirmation depending on context). So I'm not sure what you're suggesting here as a better way. If eBird reviewers didn't check for errors and follow up on them, there would be a lot more than there currently are. If you want reviewing to be far more timely a rigorous then you'd likely need professional reviews and those costs might ultimately be passed on to users. Part of what makes eBird one of (if not) the most popular citizen science projects in the world is that it's free.
2) Getting emails questioning records/locations long after the fact: This is certainly not ideal but also inevitable when you think about it. The filters that catch these errors (or possible errors) are constantly being revised and updated. Sometimes this is because the original ones were quite generic (e.g. For all of NZ instead of a specific region), and other times it's simply manually updated based on updated data. In many cases (e.g. For places like Peru), you might now have far more competent and enthusiastic reviewers that you did 10 years ago. More reviews means more effort and coverage---more likelihood to pick up anomoles and issues. You also have issues with taxonomy splits etc. that can cause havoc on what to do with less specific records. So the question a review faces is either 1) to follow up in case clarity can be achieved 2) Just hide any ambiguous records or 3) Be satisfied with subpar/ambiguous/possibly erroneous records. Like you I get random emails out of the blue, especially for overseas trips. Sometimes I can provide details but other times I just throw up my hands and say "I don't know". I don't begrudge these volunteers for trying but sometimes a satisfactory result just isn't possible and that's fine. If you get annoyed by these ocassional emails well I suppose removing your records completely is an option (But that seems drastic and unnecessary to me?)
"There are ways of tackling the problem and eBird has had a decade to make progress and they haven't." While I'm not involved in any eBird decision-making I would love to pass on any suggestions you might have. I know they are constantly trying to smoothen out all aspects of ebird--and I know you've been using it since 2008 so you'll know how far many aspects have come since the early days of Paul Scofield being the only one in the country trying to do everything!
3) To Trina asking about why obvious errors aren't removed -- Believe me, we are trying! If it's obvious and we see it, we change it when we get a change. There are sometimes thousands of these things to deal with so some slip through. Please feel free to point them out by flagging the checklist.
Hope this helps a tiny bit.
Russ
As one of the eBird review volunteers for NZ I'll try and respond to a few of the key issues raised--and don't worry--no offense taken at all--I think we're all solution-driven so feedback and suggestions should always be welcomed for eBird or anything.
1) Dealing with checklists/records ebird "doesn't own." People tend to use eBird for three main reasons: 1) To contribute to citizen science and 2) For their own record-keeping 3) To use the tools provided (e.g. Being alerted to rare sightings). If we reviewers started manually changing peoples' checklists (e.g. To correct a misIDed bird, or move a location) you could imagine how ticked off users would be. The core people at eBird do have powers to override checklists (e.g. Change your mallard to a domestic type) if they really wanted to, but the reason we ask users to do it themselves it because it needs to involve consent, and sometimes the original checklist is correct after all. By checking, this can help to 'purify' further, and it is always the right of the user to NOT change their checklist details. So in other words, you can call your mallard or sandpiper anything you want--it will remain their for your records in perpetuity (It may just be removed from public consumption if it does not adequately clear the bar for confirmation depending on context). So I'm not sure what you're suggesting here as a better way. If eBird reviewers didn't check for errors and follow up on them, there would be a lot more than there currently are. If you want reviewing to be far more timely a rigorous then you'd likely need professional reviews and those costs might ultimately be passed on to users. Part of what makes eBird one of (if not) the most popular citizen science projects in the world is that it's free.
2) Getting emails questioning records/locations long after the fact: This is certainly not ideal but also inevitable when you think about it. The filters that catch these errors (or possible errors) are constantly being revised and updated. Sometimes this is because the original ones were quite generic (e.g. For all of NZ instead of a specific region), and other times it's simply manually updated based on updated data. In many cases (e.g. For places like Peru), you might now have far more competent and enthusiastic reviewers that you did 10 years ago. More reviews means more effort and coverage---more likelihood to pick up anomoles and issues. You also have issues with taxonomy splits etc. that can cause havoc on what to do with less specific records. So the question a review faces is either 1) to follow up in case clarity can be achieved 2) Just hide any ambiguous records or 3) Be satisfied with subpar/ambiguous/possibly erroneous records. Like you I get random emails out of the blue, especially for overseas trips. Sometimes I can provide details but other times I just throw up my hands and say "I don't know". I don't begrudge these volunteers for trying but sometimes a satisfactory result just isn't possible and that's fine. If you get annoyed by these ocassional emails well I suppose removing your records completely is an option (But that seems drastic and unnecessary to me?)
"There are ways of tackling the problem and eBird has had a decade to make progress and they haven't." While I'm not involved in any eBird decision-making I would love to pass on any suggestions you might have. I know they are constantly trying to smoothen out all aspects of ebird--and I know you've been using it since 2008 so you'll know how far many aspects have come since the early days of Paul Scofield being the only one in the country trying to do everything!
3) To Trina asking about why obvious errors aren't removed -- Believe me, we are trying! If it's obvious and we see it, we change it when we get a change. There are sometimes thousands of these things to deal with so some slip through. Please feel free to point them out by flagging the checklist.
Hope this helps a tiny bit.
Russ