Could I come to your talk? Or, if I can't make it, could I have a copy of
the talk?
cheers,
Louise.
On Mon, 15 Mar 2004, G.S.Clarke wrote:
> Hello
>
> I've been asked to give a half hour talk about statistics at a
> conference and was stuck with that rather annoying misquote 'lies, damn
> lies and statistics' as a title. I've told the organisers that I'm not
> going to talk about that ... but instead have retitled the thing 'how to
> look after your statistician'. I thought that I'd try to do some
> 'educating' on how best a statistican can help research projects, the
> sort of thing that a statistician *can't* do for them - and the sort of
> thing that really 'winds' statisticians up. (hopefully the talk wont
> just be me ranting for half an hour).
>
> I was hoping that some of you, might share with me - your personal
> opinions on how the 'user' of statisticians can best get along with and
> gain benefit from our humble profession.
>
> For instance. I tend to get a bit annoyed with people who have
> collected thousands of measurements from each of ten subjects (rather
> than say ten measurements on a thousand subjects), or who come and ask
> you to 'rescue their experiment' after the experiment has been
> completed. I also tend to get a bit irked with people who turn nice
> continious variables into categories (and then treat them as such) for
> no reason that I can see at all. Or possibly worst of all the
> experimenter who comes with a pile of data, drops it on your desk and
> says 'what does this tell us' (to which I usually reply, what question
> were you asking?)
>
> I would be really grateful if you would share your thoughts/experiences
> on how we (or they) can interact more effectively. I'll summarise to
> the list - with the names of respondants removed to 'protect the innocent'.
>
> Many thanks as always
>
> Graham Clarke
>
>
> --
>
> Dr G.S.Clarke
> Lecturer in Physiology & Biometery
> Faculty of Health Studies
> University of Wales, Bangor
> Fron Heulog
> Ffriddoedd Road
> Bangor
> Gwynedd LL57 2EF
>
> Tel: 01248 383157
> e-mail: [log in to unmask]
>
|