RE>>data display 9/30/97
The only thing that I would say against this line of reasoning is that the
point seems to be that the intended audience isn't bright enought to interpret
contour plots so we'll give them simple dot plots. Dot plots do stress that
each dot is a cell. The problem is that they aren't always as good at
conveying information on populations which is what I believe is the purpose of
most flow cytometry data. As I have acknowledge before, contour plots (with
outliers) do have a steeper learning curve. However I think that the extra
information provided is worth the effort. If data is presented as a dot plot
certain information could be lost, and if it's lost in the publication it's
effectively lost forever as far as the reader is concerned. I also have
greater expectations of the abilities of our colleagues to be able to learn
how to read these plots. I think it's incumbant on those of us who use this
technology in our daily research to help educate the populace. I challenge
anyone to tell me that reading a contour plot is more difficult that
understanding MHC restriction or the various signal transduction pathways.
People cope. Yes the most important thing is to get the point across, but
does that mean that we need to sink to the lowest commmon denominator of data
presentation?
Alan Stall
--------------------------------------
Date: 9/30/97 1:28 PM
To: cyto-inbox
From: Calman Prussin
In further support of dot plots. The purpose of data display is to
convey information. Many non-cytometrists have told me that their IQ
drops by about 20 points when viewing flow data. A figure composed of
dot plots is simpler for the average reader to understand. Furthermore
it stresses the single cell nature of the data- each dot is a cell.
Bottom line: what is most important to you? Conveying information, or
being right, but not getting your point across.
Calman Prussin
----------
From: Alice.L.Givan@dartmouth.edu
Sent: Monday, September 29, 1997 12:48 PM
To: Cytometry Mailing List
Subject: re: data display
I know that dot plots can be misleading for all the reasons that
Mario Roederer
describes --- BUT I also know that, by choice of the contouring
algorithm, you
can make a contour plot look any way you want: shoulders on
peaks can be
emphasized or can be made to disappear, double peaks can be made
to look like
single peaks, etc etc etc.. These problems are not solved by
showing the dots
that are below the contouring threshold, as they relate to the
levels of
coutours above the threshold. Dot plots can be misleading,
but contour plots
are a can of worms. OK Mario (and anyone else) -- looking
forward to your
response!
Alice
Alice L. Givan
Englert Cell Analysis Laboratory
Dartmouth Medical School
Lebanon, New Hampshire
NH 03756 USA
tel 603-650-7661
fax 603-650-6130
e-mail givan@dartmouth.edu
------------------ RFC822 Header Follows ------------------
Received: by pharmingen.com with ADMIN;30 Sep 1997 13:23:04 -0700
Received: from flowcyt.cyto.purdue.edu ([128.210.60.31]) by fw.pmgsd.com
via smtpd (for [204.182.230.3]) with SMTP; 30 Sep 1997 20:19:20 UT
Received: by flowcyt.cyto.purdue.edu (940816.SGI.8.6.9/930416.SGI.AUTO)
for cyto-sendout id MAA21222; Tue, 30 Sep 1997 12:11:51 -0500
Received: from atlas.niaid.nih.gov by flowcyt.cyto.purdue.edu via ESMTP
(940816.SGI.8.6.9/930416.SGI.AUTO)
for <cytometry@flowcyt.cyto.purdue.edu> id IAA19671; Tue, 30 Sep 1997
08:12:59 -0500
Received: by atlas.niaid.nih.gov with Internet Mail Service (5.0.1458.49)
id <S94YNWX0>; Tue, 30 Sep 1997 08:54:18 -0400
Message-ID: <8C2006559F32CF119E0000AA00B7845F01E5B958@atlas.niaid.nih.gov>
From: Calman Prussin <CPRUSSIN@atlas.niaid.nih.gov>
To: cyto-inbox
Subject: RE: data display
Date: Tue, 30 Sep 1997 08:54:17 -0400
X-Priority: 3
X-Mailer: Internet Mail Service (5.0.1458.49)
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:50:11 EST