Discussion:
[Adastra] Consultation: online verification of biological records
Charles Roper
2013-12-03 20:31:37 UTC
Permalink
The Biological Records Centre (BRC; http://www.brc.ac.uk/) is carrying out
a consultation for Natural England, with funding from the Defra Fund for
Biodiversity Recording in the Voluntary Sector. They would like to hear
from people who are actively engaged with wildlife recording schemes and/or
local environmental records centres, and who carry out a role as a verifier
and/or a data manager in such schemes and centres. They would be grateful
for your response by 31 December 2013. Please pass this on to anyone within
your respective groups who does any form of record checking work.

https://docs.google.com/forms/d/1pFQiFi50jWwbuAaaaJLEABtZggrBAkGQHhhHbVe4sfo/viewform

If anyone would like to discuss the consultation, or any issues surrounding
it, please feel free either here or via the private email of Penny and I. A
good old phone call would also be welcome.

Patrick Roper has already pointed out to me that use of the term
"verification" is problematic. To verify
(definition<http://dictionary.reference.com/browse/verify>)
something means to prove by evidence, to ascertain (to find out for
*certain*), or to conclusively determine the truth. Given this definition,
very few records *at all* could truly be verified. Only those records with
voucher specimens and arguably photos (although only for some species, and
they can be faked) could be verified. In the strictest sense, only where
the verifier is there in the field with the recorder and can confirm the
specimen taken can a record be verified. What, I think, is meant by
"verification" is actually closer to "checking", "corroborating" or
"lending support" based on, primarily, expert knowledge and knowledge of
the recorder; i.e., qualitative judgement rather than strictly objective.
This may sound like semantic hair-splitting, but to use the verb "verify"
lends undue objective weight to what work has actually been done *and is
able to be done* on a record. For the end-user of data, it sends the wrong
signals that can lead to faulty expectations and assumptions.

The following article by Jeni Tennison<http://theodi.org/team/jeni-tennison>,
Technical Director of the Open Data Institute, makes for interesting and
insightful reading and is related to the above. I was certainly nodding
furiously as I read it. Here is the closing thought:


Making data open for other people to look at provides lots more
opportunities for people to spot errors. This can be terrifying ? who wants
people to know that they are running their organisation based on
bad-quality data? ? but those who have progressed through the five stages
of data grief find hope in another developer maxim:

* Given enough eyeballs, all bugs are shallow.*

*? Linus?s Law, The Cathedral and the Bazaar by Eric Raymond*


The more people look at your data, the more likely they are to find the
problems within it. The secret is to build in feedback mechanisms which
allow those errors to be corrected, so that you can benefit from those eyes
and increase your data quality to what you thought it was in the first
place.


Here is the full article: http://theodi.org/blog/five-stages-of-data-grief

Charles

*Charles Roper*
Sussex Biodiversity Record Centre | http://sxbrc.org.uk | 01273 497554

Sussex Wildlife Trust is a company limited by guarantee under the Companies
Act. Registered in England, Company No. 698851. Registered Charity No.
207005. VAT Registration No. 191 305969. Registered Office: Woods Mill,
Henfield, West Sussex BN5 9SD. Telephone 01273 492630
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sxbrc.org.uk/pipermail/adastra_lists.sxbrc.org.uk/attachments/20131203/91ff3a7c/attachment.html>
Patrick Roper
2013-12-03 21:33:44 UTC
Permalink
Good email ? well done. I did, btw, read the article by Jeni Tennison. I think, again, there is a definition problem, namely what do we mean by data? Much data is, of course, fuzzy but can be helpful if analysed according to probability theory etc. Facial composites are put together by descriptions from as many witnesses as possible, but the data they provide, though essential, is approximate rather than absolutely correct. In the past it was verified that the sun went round the earth, and one could check it every day from the data - it rose and set.



Perhaps we should ask Defra to verify that their preferred new chair for Natural England really is Andrew Sells.



From: Adastra [mailto:adastra-bounces at lists.sxbrc.org.uk] On Behalf Of Charles Roper
Sent: 03 December 2013 20:32
To: Adastra discussion group
Subject: [Adastra] Consultation: online verification of biological records



The Biological Records Centre (BRC; http://www.brc.ac.uk/) is carrying out a consultation for Natural England, with funding from the Defra Fund for Biodiversity Recording in the Voluntary Sector. They would like to hear from people who are actively engaged with wildlife recording schemes and/or local environmental records centres, and who carry out a role as a verifier and/or a data manager in such schemes and centres. They would be grateful for your response by 31 December 2013. Please pass this on to anyone within your respective groups who does any form of record checking work.



https://docs.google.com/forms/d/1pFQiFi50jWwbuAaaaJLEABtZggrBAkGQHhhHbVe4sfo/viewform



If anyone would like to discuss the consultation, or any issues surrounding it, please feel free either here or via the private email of Penny and I. A good old phone call would also be welcome.



Patrick Roper has already pointed out to me that use of the term "verification" is problematic. To verify ( <http://dictionary.reference.com/browse/verify> definition) something means to prove by evidence, to ascertain (to find out for certain), or to conclusively determine the truth. Given this definition, very few records at all could truly be verified. Only those records with voucher specimens and arguably photos (although only for some species, and they can be faked) could be verified. In the strictest sense, only where the verifier is there in the field with the recorder and can confirm the specimen taken can a record be verified. What, I think, is meant by "verification" is actually closer to "checking", "corroborating" or "lending support" based on, primarily, expert knowledge and knowledge of the recorder; i.e., qualitative judgement rather than strictly objective. This may sound like semantic hair-splitting, but to use the verb "verify" lends undue objective weight to what work has actually been done and is able to be done on a record. For the end-user of data, it sends the wrong signals that can lead to faulty expectations and assumptions.



The following article by Jeni Tennison <http://theodi.org/team/jeni-tennison> , Technical Director of the Open Data Institute, makes for interesting and insightful reading and is related to the above. I was certainly nodding furiously as I read it. Here is the closing thought:



Making data open for other people to look at provides lots more opportunities for people to spot errors. This can be terrifying ? who wants people to know that they are running their organisation based on bad-quality data? ? but those who have progressed through the five stages of data grief find hope in another developer maxim:



Given enough eyeballs, all bugs are shallow.

? Linus?s Law, The Cathedral and the Bazaar by Eric Raymond



The more people look at your data, the more likely they are to find the problems within it. The secret is to build in feedback mechanisms which allow those errors to be corrected, so that you can benefit from those eyes and increase your data quality to what you thought it was in the first place.



Here is the full article: http://theodi.org/blog/five-stages-of-data-grief



Charles




Charles Roper
Sussex Biodiversity Record Centre | http://sxbrc.org.uk | 01273 497554


Sussex Wildlife Trust is a company limited by guarantee under the Companies Act. Registered in England, Company No. 698851. Registered Charity No. 207005. VAT Registration No. 191 305969. Registered Office: Woods Mill, Henfield, West Sussex BN5 9SD. Telephone 01273 492630

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sxbrc.org.uk/pipermail/adastra_lists.sxbrc.org.uk/attachments/20131203/23d92fa8/attachment.html>
Charles Roper
2013-12-03 23:20:35 UTC
Permalink
The question "what is data" was looked at on the same site recently:

http://theodi.org/blog/what-is-data-exploring-the-question-that-no-one-asks

It's a bit abstract, but draws 3 simple conclusions based on perspective.
These describe data as:

1. objective facts
2. subjective observations
3. communications

So data means different things to different people. Data should as such be
protected as far as is possible in its original form, with layers of
interpretation, expert judgement, etc. layered on top as annotations, as
with a curating a collection of physical artifacts. The article has a link
to a paper, but that link no longer works, so I've attached the paper to
this message. Well worth a read.

A concrete example of why it is important to preserve original data is with
the Old Weather project. Old Weather seeks to digitise the log books of old
ships so as to obtain a picture of historical weather:

http://www.oldweather.org/why_scientists_need_you

"[Philip Brohan] Charles Darwin's voyage on the Beagle is most famous for
his great work on evolution, and his great zoological work. But actually
Robert Fitzory, the Captain, and his crew also recorded the weather in
their logs of the records at every point the ship visited."


I'm sure the sailors of the 17- and 1800's - and Darwin himself - would be
marvelling at the ways the data gathered on their voyages would be put to
such novel and unexpected use some 300 years later.

*Charles Roper*
Sussex Biodiversity Record Centre | http://sxbrc.org.uk | 01273 497554

Sussex Wildlife Trust is a company limited by guarantee under the Companies
Act. Registered in England, Company No. 698851. Registered Charity No.
207005. VAT Registration No. 191 305969. Registered Office: Woods Mill,
Henfield, West Sussex BN5 9SD. Telephone 01273 492630
Post by Patrick Roper
Good email ? well done. I did, btw, read the article by Jeni Tennison. I
think, again, there is a definition problem, namely what do we mean by
data? Much data is, of course, fuzzy but can be helpful if analysed
according to probability theory etc. Facial composites are put together by
descriptions from as many witnesses as possible, but the data they provide,
though essential, is approximate rather than absolutely correct. In the
past it was verified that the sun went round the earth, and one could check
it every day from the data - it rose and set.
Perhaps we should ask Defra to verify that their preferred new chair for
Natural England really is Andrew Sells.
*From:* Adastra [mailto:adastra-bounces at lists.sxbrc.org.uk] *On Behalf Of
*Charles Roper
*Sent:* 03 December 2013 20:32
*To:* Adastra discussion group
*Subject:* [Adastra] Consultation: online verification of biological
records
The Biological Records Centre (BRC; http://www.brc.ac.uk/) is carrying
out a consultation for Natural England, with funding from the Defra Fund
for Biodiversity Recording in the Voluntary Sector. They would like to hear
from people who are actively engaged with wildlife recording schemes and/or
local environmental records centres, and who carry out a role as a verifier
and/or a data manager in such schemes and centres. They would be grateful
for your response by 31 December 2013. Please pass this on to anyone within
your respective groups who does any form of record checking work.
https://docs.google.com/forms/d/1pFQiFi50jWwbuAaaaJLEABtZggrBAkGQHhhHbVe4sfo/viewform
If anyone would like to discuss the consultation, or any issues
surrounding it, please feel free either here or via the private email of
Penny and I. A good old phone call would also be welcome.
Patrick Roper has already pointed out to me that use of the term
"verification" is problematic. To verify (definition<http://dictionary.reference.com/browse/verify>)
something means to prove by evidence, to ascertain (to find out for
*certain*), or to conclusively determine the truth. Given this
definition, very few records *at all* could truly be verified. Only those
records with voucher specimens and arguably photos (although only for some
species, and they can be faked) could be verified. In the strictest sense,
only where the verifier is there in the field with the recorder and can
confirm the specimen taken can a record be verified. What, I think, is
meant by "verification" is actually closer to "checking", "corroborating"
or "lending support" based on, primarily, expert knowledge and knowledge of
the recorder; i.e., qualitative judgement rather than strictly objective.
This may sound like semantic hair-splitting, but to use the verb "verify"
lends undue objective weight to what work has actually been done *and is
able to be done* on a record. For the end-user of data, it sends the
wrong signals that can lead to faulty expectations and assumptions.
The following article by Jeni Tennison<http://theodi.org/team/jeni-tennison>,
Technical Director of the Open Data Institute, makes for interesting and
insightful reading and is related to the above. I was certainly nodding
Making data open for other people to look at provides lots more
opportunities for people to spot errors. This can be terrifying ? who wants
people to know that they are running their organisation based on
bad-quality data? ? but those who have progressed through the five stages
* Given enough eyeballs, all bugs are shallow.*
*? Linus?s Law, The Cathedral and the Bazaar by Eric Raymond*
The more people look at your data, the more likely they are to find the
problems within it. The secret is to build in feedback mechanisms which
allow those errors to be corrected, so that you can benefit from those eyes
and increase your data quality to what you thought it was in the first
place.
Here is the full article: http://theodi.org/blog/five-stages-of-data-grief
Charles
*Charles Roper*
Sussex Biodiversity Record Centre | http://sxbrc.org.uk | 01273 497554
Sussex Wildlife Trust is a company limited by guarantee under the
Companies Act. Registered in England, Company No. 698851. Registered
Woods Mill, Henfield, West Sussex BN5 9SD. Telephone 01273 492630
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sxbrc.org.uk/pipermail/adastra_lists.sxbrc.org.uk/attachments/20131203/b649c69d/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ballsun-stanton2010-Asking_about_data_Experimental_philosophy_of_Information_Technology.pdf
Type: application/pdf
Size: 325059 bytes
Desc: not available
URL: <http://lists.sxbrc.org.uk/pipermail/adastra_lists.sxbrc.org.uk/attachments/20131203/b649c69d/attachment-0001.pdf>
Loading...