[open-science] Centre for Global Development adopts open data policy for publications

[Paola Di Maio]

> As Ohm notes, this illustrates a central reality of data >collection: "data can either be useful or perfectly >anonymous but never both."


Thank you Peter for the nice quote

sure, 'privacy' cannot actually be taken lightly

However being a pragmatic scientist, having to make a choice, I would
necessarily aim for (esp re data paid for with public money)  it being
useful, and imperfectly anonymous (or as some would put it, anonymous
enough...)

Even the need for privacy must be weighted against
the context, I mean privacy of personal data (ie, my bank pin, genetic
code, phone calls record or clinical and health data etc) are a
stronger requirement than the need to protect my legitimate views on
'open access' aired in the context of a scientific enquiry the results
of which will be published in aggregated and anonymized form.

Especially not in the case of subjects who  otherwise take every
opportunity to make their views known....

PDM

On Fri, Aug 5, 2011 at 12:46 AM, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:
>
>
> On Thu, Aug 4, 2011 at 8:28 PM, Paola Di Maio <paola.dimaio at gmail.com>
> wrote:
>>
>> Thank you Chris
>>
>>
>> >  Key Perspectives are independent  consultants, so not subject to FoI
>> > >except where they hold data on behalf of others, including JISC who
>> > >commissioned this work. I have some sympathy for their practice;
>> > >qualitative data are often this kind of one to one interview data and
>> > >potentially very sensitive.
>>
>> I suggested that the data could be anonymized instead of being
>> destroyed  - so that I would not have to know the name of the person,
>> but still have an idea of their demographics, for example - job title,
>> institution,department, city, town etc. I would really like to analyse
>> such data according to different segmentation in relation to
>> geographical differences for example, would there be differences
>> between north and south, or between different institutions, and
>> disciplines, as this could serve the public good and provide
>> additional science for the same buck.
>
>
> It is generally recognized that it is impossible to anonymize data
> completely
>
> See
> http://arstechnica.com/tech-policy/news/2009/09/your-secrets-live-online-in-databases-of-ruin.ars
>
> with the quote:
>
> As Ohm notes, this illustrates a central reality of data collection: "data
> can either be useful or perfectly anonymous but never both."
>
> --
> Peter Murray-Rust
> Reader in Molecular Informatics
> Unilever Centre, Dep. Of Chemistry
> University of Cambridge
> CB2 1EW, UK
> +44-1223-763069
>