[texbirds] Re: e-bird question- Quis custodiet ipsos custodes?
- From: MBB22222@xxxxxxx
- To: texbirds@xxxxxxxxxxxxx
- Date: Fri, 1 Feb 2013 12:34:54 -0500 (EST)
Eric,
I'm not sure you have the right idea about what I wanted to say as I am
afraid that you seem to did not understand what I wrote. Perhaps this is my
English so I will try again to stress out a few things one more time and
hopefully I will do a better job this time. I see that your position in eBirds
structure makes you a perfect person to speak for this organization so me
as well as others can learn more about this great project.
I absolutely do not have problem with the eBirds process and the time
length it takes to review records. I will take it further - process should take
even longer (to review records) if that what is needed to get better
results. I personally do not care at all how long it will take to approve any
of
my records if I decide to enter some. Sure, a faster approval benefits
these who look for the newest records - I usually check long term and seasonal
distributions (and only a few species) and already have my own records
available to me anyway. So you are the one who wants to get timely processed
data (which is a good approach but hard to achieve with limited staff) and I
would like to see it happened but do not see it necessary and I am
absolutely against speeding the process if in result the quality of data is
compromised.
Now I will provide a couple examples illustrating both situations when
reviewers either can or cannot correct the data.
Let me present you as an example something I witnessed. A small group of
people where standing in the front of egret rookery counting birds and taking
detailed notes. Great so far - seemed like they are working on detailed
record to enter somewhere (eBirds?). I overheard them talking about a large
number of nesting Snowy Egrets they are just observing. After a moment I
decided to interrupt their conversation and say something. Excusive, I said, I
heard that you are talking about Snowy but from here you only see a large
number of Cattle Egrets; there are a few Snowy nests not far away but from
here you only can see Cattle. I received a very strange look from all of
them. Couple of them stepped aside took a bird guide from the pocket and
study it for a few minutes looking back and forth at egrets in front of them
and illustrations in the book. They all left soon after that without talking
to me again. What they reported I do not know. What I know that they could
report bunch of Snowy Egrets for sure if I was not there. To some perhaps
this is not a big deal - both common species, both were present (even that
in quite different proportion). To me this is a problem if those numbers
were reported incorrectly. In this case I do not blame neither observer
(working and reporting in good faith) or reviewer (no chance to catch the
mistake
if not familiar with the rookery) but I do have problem with the system
even that I do not see a solution how to improve it. The observers evidently
had not clue what they are looking at but databases like eBirds allow to
enter this kind of data, any kind of data. Again I see no solution but this
let me to be very skeptical about quality of data in the database. I am sure
this happen too often to be ignored when one tried to analyze data.
As you see I did not blame observers in the case presented above and I
think no observers acting in good faith can really be blamed for entering
incorrect records. But I could blame reviewers if in some cases they could and
should check/correct data but they didn't. I already described what is going
on with this year January Common Tern records in Texas so instead me
writing it again here is the link to that post:
_//www.freelists.org/post/texbirds/Forsters-Tern-with-misleading-impres
sion-of-having-a-dark-carpal-bar_
(//www.freelists.org/post/texbirds/Forsters-Tern-with-misleading-impression-of-having-a-dark-carpal-bar)
I hope you will find time to took at eBirds Texas January COTE records and
if you find that there is nothing wrong with flocks of 20-25 COTEs I will
accept that this is a way of eBirds accepting some records. To me any
accepted January COTE record in Texas (even single birds) should be
documented,
and for sure to be not accepted by the default. Perhaps you can explain to
me if you think it is not necessary to do so. BTW I still think that guards
should be guarded or mess will happen; sooner or later.
Just to post a few photos not only bare text here is an example of the FOTE
with real dark carpal bar quite well visible, not just an illusion this
time.
_http://www.pbase.com/mbb/image/148573810_
(http://www.pbase.com/mbb/image/148573810)
_http://www.pbase.com/mbb/image/148573811_
(http://www.pbase.com/mbb/image/148573811)
_http://www.pbase.com/mbb/image/148573812_
(http://www.pbase.com/mbb/image/148573812)
_http://www.pbase.com/mbb/image/148573813_
(http://www.pbase.com/mbb/image/148573813)
Mark
Mark B Bartosik
Houston, Texas
_http://www.pbase.com/mbb/from_the_field_
(http://www.pbase.com/mbb/from_the_field)
In a message dated 1/31/2013 11:05:41 P.M. Central Standard Time,
_ecarpe@gmail.com_ (mailto:ecarpe@xxxxxxxxx) writes:
Mark,
I'm not sure you have the right idea about the eBird review process.
The areas not being reviewed 'adequately' as you put it are areas that
are not being reviewed in a TIMELY (in the eye of the beholder) way
perhaps but the reviewers are all quite competent and often go above &
beyond to get documentation on sightings. Many requests for
documentation can go unanswered. The reason why the time lag in what
some folks seem to be expecting as real-time or near real-time
posting/vetting of every one of their sightings is because all the
reviewers are volunteers who often have a list of things they need to
be or rather be doing (many of them are often out birding instead!)
then this task.
Who is guarding the guards? Well, the "guards" as you put it are very
few in number who are both willing and able to do this and don't
really need guarding at this point but would love to have additional
guards helping them instead. They are part of a behind-the-scenes team
that is making eBird data quality better and better and are constantly
revisiting old records in a never-ending process of separating the
wheat from the chaff. In the end, the "misleading" records are indeed
removed from the database and/or never make it past the reviewers in
the first place.
There is a lot more data and interesting sightings in eBird these days
then there is in Texbirds, either the listserve version or the
Facebook version, and I would say that both those forums suffer issues
of quality/lack of vetting with no appointed "guard". Mining eBird
can produce some real diamonds in the rough that would otherwise never
have been documented. Thanks to adoption of smartphone apps like
BirdLog and Bird Watcher's Diary by birders in the field, much of that
data now comes in almost real-time (or at least on the same day).
Unfortunately, that amount of data can not be vetted instantly simply
because there aren't that many hours in the day. Just one month into
the year, and there are over 24,000 bird records in eBird for 2013 in
Harris County (the core of Houston birding); in the Austin area, there
are over 16,000 database entries from January for Travis County. The
validated records for these areas paint a much more accurate picture
of bird distribution (for regularly occurring species) than any other
medium/forum we have.
As Mary suggested, providing documentation w/o being asked (BirdLog
will even let you know what sightings require documentation) will
speed the process along. Reviewers certainly aren't perfect and tend
to err on the conservative side as might be expected, but they do a
pretty bang up job in trying to guard the eBird store.
-- Eric
Eric Carpenter
Austin
(full-time eBird reviewer for Travis County and part-time/backup for
other areas of the state)
On Thu, Jan 31, 2013 at 10:01 PM, <_MBB22222@aol.com_
(mailto:MBB22222@xxxxxxx) > wrote:
Well, just to make sure that nobody misunderstand my intention - I truly
like the idea (eBirds) and this database could be helpful, but ⦠There is
usually a âbutâ, isnât it?
It is not important how many and how long hours reviewers spend when
checking records - what counts is quality not quantity. I read with quite
interest these only few responses and it was not a surprise that only some
regions are thought to do reviews adequately (hope more than listed).
So another question needs to be asked: Who is guarding the guards?
Evidently some reviewers are having problems with verifying records (hopefully
they do not use vote method which is so often used on popular fora). IMHO
another layer of guards checking guards could improve the situation so would
removal of those who for whatever reason cannot correctly qualify records. I
understand that this is a new project that still needs a lot improvements
but there is no excuse to keep and accumulating worthless and misleading
records. This database (eBirds) should not be a replacement for places where
birders look for something to chase - Texbirds and other similar websites
are good enough for that. BTW I only recently learned that there is
difference between birder and bird watcher ⦠So birders have plenty of
places to
check for something to chase and records on eBirds should be rather delayed
than incorrect. Often less is more â¦
Mark B Bartosik
Houston, Texas
_http://www.pbase.com/mbb/from_the_field_
(http://www.pbase.com/mbb/from_the_field)
Edit your Freelists account settings for TEXBIRDS at
//www.freelists.org/list/texbirds
Reposting of traffic from TEXBIRDS is prohibited without seeking permission
from the List Owner
Other related posts:
