Subject: | |
From: | |
Reply To: | |
Date: | Mon, 18 Jun 2007 07:20:28 -0400 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Dear all:
I was thinking about the creamatocrit discussions and research that is based on conflicts
of interest and it dawned on me how easy it is to stagger results if people are not aware
of proper research techniques.
I wish I had the details of the talk I went to at hand to illustrate why the creamatocrit
would be a lousy diagnostic tool. As I remember from the talk, the variability from feed
to feed, breast to breast, minute to minute within the same woman was astoundingly
large! If one looks at test weighing, I was surprised that it backs up what all of us know
from observing infants feed. Infants usually take a bit more in the morning, but the
variability feed to feed in total intake is far less than changes in the fat content.
So, it dawned on me one could simply cheat and stagger the results by having the mother
pump herself almost dry with masssage, turning the machine on and off, using better
fitting flanges and then do the creamocrit. Of course this would not be ethical and it
would not help educate other health care practitioners about how the fat content changes
throughout a feeding, nor the body of research cited by Cathy Genna that supports
volume being more important than measurements of fat content.
Similarly, the easiest study in the world to design is one that shows no results. All you
have to do is simply use a smaller sample size or do sloppy measurements. The first will
reduce your statistical power. The second will increase your variance.
No test of any kind, no matter how sophisticated is 100% accurate, nor 100% precise.
All tests have wiggle room or variance. Some have more variance than others. The
more variance, the greater the number of measurements you have to take to
compensate. The more measurements you have to take, the more intrusive the test.
As Rachel pointed out, the most important issue is how you are going to use the
information from the test to make decisions or to educate parents about how to observe
their infants cues and trust their own observations. In rare cases, you must really look at
everything that is going on and one lone diagnostic tool is not really going to give you the
answer without astute listening and observing and putting together a comprehensive
picture.
Best, Susan.
***********************************************
Archives: http://community.lsoft.com/archives/LACTNET.html
Mail all commands to [log in to unmask]
To temporarily stop your subscription: set lactnet nomail
To start it again: set lactnet mail (or [log in to unmask])
To unsubscribe: unsubscribe lactnet or ([log in to unmask])
To reach list owners: [log in to unmask]
|
|
|