Hi,
I use dotplots to give me an idea of where populations lie, and I get the
computer to count events in the regions that I set. I don't feel that I
need to act as a densitometer, so I don't fiddle around getting the
"correct" contour set-up. Nonetheless I find that I'm not more than 5-10%
(of estimated percentage) out on most estimates that I'm asked to make at
run time for populations in the range of 5% to 50% of total. The proximity
of similar dots gives an impression of increasing density, that works not
unlike halftoning used in the print industry for representing grey scales.
As for low frequency populations: I had no problem seeing a cluster of 31
transfectants in 2 million cells yesterday using an accumulating dot plot
(whether I managed to sort any of them is a separate issue). It didn't
matter that the negatives appeared as a solid blob, I could still count 28
positive dots on the screen (presumably one or two of the dots represented
more than one event - the computer counted them for me as well so who
cares?). How would you go about choosing contour levels to best represent
this population? Could you really choose levels that would allow you to say
that there is a 0.005% population, or even to visualise it, in less time
than it takes to count the dots and divide by the total?
Ray
ps I realise that the sample size isn't really large enough for statistical
significance, so let's not open that can of worms, but here's an
interesting quote from the Feedback section of New Scientist magazine on
the Issue:
"The following is a recent conversation Feedback held with a representative
of the Office for National Statistics(ONS), after being referred there for
a concise definition of the term 'statistical significance'.
Feedback: "Could you tell me the definition of statistical significance?"
ONS: "What do you mean?"
Feedback: "You know, if you sample a population, when does that sample
become statistically significant? When is it taken into
account?"
ONS: "Well, we sample the population every other year, we
collect the
data, print it...and then it becomes significant""
Ray Hicks
________________________________________________________________________
|University of Cambridge |Tel 01223 330149 |
|Department of Medicine |Fax 01223 336846 |
|Level 5, Addenbrookes Hospital |e-mail <rh208@cus.cam.ac.uk> |
|Hills Road Cambridge |Web http://facsmac.med.cam.ac.uk |
|CB2 |ftp server ftp://131.111.80.78 |
|UK | |
|_________________________________|_____________________________________|
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:50:11 EST