Re: Ratio or Mean

From: Geoffrey Osborne (geoff.osborne@anu.edu.au)
Date: Tue Mar 27 2001 - 16:21:59 EST


Hello Tomas,
	I thought I'd stick my neck out on this one and give you my two cents worth.
	It's been discussed numerous times in the flow list over the years, and my
take on this issue is this: given the fact that flow cytometry is a
"relative" science, in that you are always comparing a control with a
responding sample and looking to compare the relative shift, then it is
appropriate to use a ratio to reflect the relative amount of change, e.g.
expressing a result as a 2 fold increase relative to a control is meaningful.
	However, the statistic you choose to base this ratio upon is a more
difficult proposition. As Dave Coder (and others) said recently, the
geometric mean is appropriate given a log normal distribution, and like
Dave I'd suggest using the median fluorescence intensity if the population
is skewed at all, even though the median will also be slightly affected by
a skewed distribution.
The other point I think is useful here, from a purely non statistical point
of view, is that when you start talking about the median, almost everyone
knows what it is and that the median is an indicator of the central
tendency of a population.
	Other people may have a different interpretation of this topic, so I'll
leave you with my final thought on this. Consistency in the choice of
statistic is important. Choose a statistic which you can justifiably say is
a good indicator of the trends in your data sets and then use that method
consistently.
	Hope this helps in some way, can't have you all going daft! :-)
Geoff


At 10:14 PM 3/23/01 -0000, you wrote:
>   I wonder whether anyone is willing to put their  neck on the line with
>this apparently innocuous question.   It relates to the ongoing discussion
>regarding  whether the mean , geometric mean or median should be used.
>When measuring neutrophil respiratory burst or  phagocytic activity , using
>Phagoburst from pharmigen , should you use the  absolute difference between
>geometric means ( means , medians ) of the control  versus the activated
>samples , or should you use the ratios. We were mostly  using the
>differences or ratios of the geometric means but there is a fierce  debate
>ongoing in our laboratory regarding this issue and preliminary results
>from comparative studies show the use of ratios of geometric means coming
>out on  top.	 Please help we're all going daft and would love an
>inpartial arbitration.   Tomas Corcoran Dept.of Anaesthetic research  Cork
>University Hospital Cork City Republic of  Ireland
Geoffrey Osborne

Specialist, Flow Cytometry,
John Curtin School of Medical Research,
The Australian National University,
Canberra, 0200, ACT. AUSTRALIA
email: geoff.osborne@anu.edu.au
http://jcsmr.anu.edu.au/facslab/facshome.html

(61 2) 6125 3694.



This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:01:13 EST