I agree with Kathy. I haven't heard that 'massaging the data' means to make
the dtat say something that it doesn't.

The way I think about it is like this.

When you run an experiment you often end up with an enormous number of data
points. Could be hundreds through to tens of thousands. It is very
difficult to look at even a hundred points and be able to 'see' what is
happening so you might do something simple like averaging the points. You
then might be able to see a bit clearer what is going on BUT the trick is
to massage in just the right way so that you do see the true story and not
something that is not there at all.

Much to everyones frustration incorrect massaging (or poor experimental
design eg the unfortunate never breastfed vs ever breastfed) can result in
an untrue or at best unclear story. An example of where this happened with
the NZ SIDS study published in 92? It came out saying that cosleeping
increased the risk of SIDS. True story? not quite. The data were true but a
little more massaging was required to find that in fact cosleeping only
increased SIDS when the parents smoked. You can't believe everything you
read and how you present data is really important.

There may also be a number of stories that each set of data may be telling
you. You can massage it in a number of different ways to get different yet
true stories. You have to work out which story you want to tell, which is
the most important in your eyes. There can be a large amount of
subjectivity involved.

I see data as being like a nebulous blob of clay. There is a lot there but
if it just sits in a blob its not much use. You can however, massage it
into something beautiful with a bit of insight, skill and hard work.

Karleen Gribble
Australia
Now back to massaging some data into a paper

             ***********************************************
The LACTNET mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software together with L-Soft's LSMTP(TM)
mailer for lightning fast mail delivery. For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html