And viceversa, what is set to be important in a propaganda / communication strategy often leads researchers and information officers - but also journalists, bloggers and communicators etc - to find statistical significance in the most improbable facts or phenomena simply because they want consciously or unconsciously to please somebody or something.
I remember myself ages ago - when I was an employee, before the independent consultant hat gave me a certain degree of freedom in respect of such issues - having very difficult negotiations with end-users / commissioners of a certain type of enquiries. They were asking to provide the data that could fit certain assumptions whereas I was expecting them to ask investigative or appreciatives enquiries in order to prevent / avoid biased processes.
In sum, how to recognise and avoid bias in design of analytics applications seems to me at present the great challenge for many commercial and educational projects. That is why in the article I advertised yesterday - from my web column imc2:re - I insist on the idea of a "data ownership" principle.
Brunella Longo
Information Management Adviser
Open Data Assurance
http://www.brunellalongo.co.uk
:::
Telephone +44(0)7549921488
email: [log in to unmask]
:::
PO BOX 53880
London SE27 7BU
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|