It comes down to sample definition.
The smaller the sample size the less sensible it is to use a statistical
research mechanism because you will be chasing results from sub populations
of one.
For the purposes of the type of study you outline it seems more sensible to
"follow the individual" (ignoring any DP implications) rather than to
attempt a nonsensical anonymisation.
-----Original Message-----
From: Smith, Tony [mailto:[log in to unmask]]
Sent: 17 February 2005 14:04
To: Tim Trent; [log in to unmask]
Subject: RE: 'pure' research
Surely there could be "pure" research, for example, that looked at the time
that it took local people in a small area to get rehoused. The results may
then say that in one specific ward there were x people who were white who
were waiting to be rehoused and they were so within 6 months and that there
was 1 person who was not white who was not rehoused for a year. Would this
be pure research? It should not take too much research to find out who the
person was! Additionally, pure research could look at why the person needed
rehousing in the first place!
Tony
-----Original Message-----
From: This list is for those interested in Data Protection issues
[mailto:[log in to unmask]]On Behalf Of Tim Trent
Sent: 17 February 2005 12:23
To: [log in to unmask]
Subject: Re: 'pure' research
One has to look at what "research" is.
Pure research is an anonymised statistical study of a group of unidentified
research subjects. It creates an analysis of results by set. If it
identified an individual that would limit its value as pure research since
perceptions of the individual would be used to interpret the results.
Passing the results of the individual to the individual are the passing of a
sample of one to a sample of one. The individual has a right to see the
answers, naturally.
Passing attributable results to a third party are outside the terms of
reference of pure research. It is possible to declare at the start of a
study "By taking part you agree that your attributable results may be passed
to (list of third parties or statement about third party access) for the
following purposes (list of purposes)", and that removes the DPA issue.
It also taints the research since the sample is necessarily skewed to
include only those who do not mind their data being passed to third parties.
Conducting pure research on a skewed sample invalidates the results of the
research. There are circumstances where one could argue against this,
including the comparison of purposely skewed samples against other purposely
skewed samples, but that is way beyond the initial question.
-----Original Message-----
From: This list is for those interested in Data Protection issues
[mailto:[log in to unmask]] On Behalf Of Ian Welton
Sent: 17 February 2005 12:15
To: [log in to unmask]
Subject: Re: [data-protection] 'pure' research
Tim Trent on Thursday, February 17, 2005 at 10:25 AM said:-
> At the point of delivering attributable data to the school (or to any
> third
> party) the research is no longer pure.
Are all the exemptions contained within the DPA permissible, or is research
purity tainted in the same way?
Ian W
> -----Original Message-----
> From: This list is for those interested in Data Protection issues
> [mailto:[log in to unmask]] On Behalf Of Tim Trent
> Sent: Thursday, February 17, 2005 10:25 AM
> To: [log in to unmask]
> Subject: Re: 'pure' research
>
>
> At the point of delivering attributable data to the school (or to any
> third
> party) the research is no longer pure.
>
> -----Original Message-----
> From: This list is for those interested in Data Protection issues
> [mailto:[log in to unmask]] On Behalf Of Okey, Andrew
> Sent: 17 February 2005 10:21
> To: [log in to unmask]
> Subject: [data-protection] 'pure' research
>
> Hi folks,
>
> Today's HE teaser is -
>
> Research staff undertake work in schools. A group of pupils, some with
> learning difficulties, see a play and are then asked what they
> remember about it (the point of the research is to establish whether
> children with learning difficulties are potentially less reliable
> witnesses in court). As part of the set-up for the work, all children
> likely to be involved are given an adapted IQ test to inform the
> analysis of their later recollections.
>
> So far, so 'pure' and, of course, pure research enjoys certain
> exemptions from DP law, notably from access under section 7, and from
> principles 2 and 5.
>
> Now, the researchers suspect that they may get enquiries from parents
> whose children are involved in the work and who are interested in
> knowing their IQ results. That looks OK, because section 33.5 of the
> act says that research is still pure even where the data is disclosed
> to "the data subject of a person acting on their behalf". However, the
> researchers are concerned that the schools involved may also
> ask/demand to know such results. Now, research is only pure when it
> does not "support measures or decisions" made on/against individuals,
> and one could imagine that what the schools would be trying to obtain
> are measures of children's IQ on which they might conceivably make
> decisions - e.g. about streaming.
>
> At which point, if the researchers gave this data to schools would the
> research no longer be pure, thereby removing all the exemptions they'd
> usually enjoy?
>
> Andrew Okey
> Lancaster University
>
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> All archives of messages are stored permanently and are
> available to the world wide web community at large at
> http://www.jiscmail.ac.uk/lists/data-protection.html
> If you wish to leave this list please send the command
> leave data-protection to [log in to unmask]
> All user commands can be found at : -
> http://www.jiscmail.ac.uk/help/commandref.htm
> Any queries about sending or receiving message please send to the list
> owner
> [log in to unmask]
> (all commands go to [log in to unmask] not the list please)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> All archives of messages are stored permanently and are
> available to the world wide web community at large at
> http://www.jiscmail.ac.uk/lists/data-protection.html
> If you wish to leave this list please send the command
> leave data-protection to [log in to unmask]
> All user commands can be found at : -
> http://www.jiscmail.ac.uk/help/commandref.htm
> Any queries about sending or receiving message please send to the list
> owner
> [log in to unmask]
> (all commands go to [log in to unmask] not the list please)
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
All archives of messages are stored permanently and are
available to the world wide web community at large at
http://www.jiscmail.ac.uk/lists/data-protection.html
If you wish to leave this list please send the command
leave data-protection to [log in to unmask]
All user commands can be found at : -
http://www.jiscmail.ac.uk/help/commandref.htm
Any queries about sending or receiving message please send to the list owner
[log in to unmask]
(all commands go to [log in to unmask] not the list please)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
All archives of messages are stored permanently and are
available to the world wide web community at large at
http://www.jiscmail.ac.uk/lists/data-protection.html
If you wish to leave this list please send the command
leave data-protection to [log in to unmask]
All user commands can be found at : -
http://www.jiscmail.ac.uk/help/commandref.htm
Any queries about sending or receiving message please send to the list owner
[log in to unmask]
(all commands go to [log in to unmask] not the list please)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
All archives of messages are stored permanently and are
available to the world wide web community at large at
http://www.jiscmail.ac.uk/lists/data-protection.html
If you wish to leave this list please send the command
leave data-protection to [log in to unmask]
All user commands can be found at : -
http://www.jiscmail.ac.uk/help/commandref.htm
Any queries about sending or receiving message please send to the list owner
[log in to unmask]
(all commands go to [log in to unmask] not the list please)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|