Mime-Version: |
1.0 |
Content-Type: |
text/plain; charset="us-ascii" ; format="flowed" |
Date: |
Thu, 28 May 2015 12:15:26 +0100 |
Reply-To: |
|
Subject: |
|
From: |
|
In-Reply-To: |
|
Message-ID: |
<a06240801d18ca47cfbb0@[192.168.0.4]> |
Sender: |
|
Parts/Attachments: |
|
|
I can't find the comments, and there are no ads on my screen.
I expect the authors will come back with a robust riposte to the
comment about 'too large a sample'.
They will argue that they *have* taken measures to address potential
confounding factors.
My understanding of stats is (ahem) a work in progress, but I
understand that a very large sample is sometimes *not* a good thing,
though. The problem with large samples can be that the huge numbers
allow you to look for things and find them - simply 'cos with
sufficient subjects in your study, you are more likely to come up
with something that confirms your bias! You can go on asking a
zillion questions, until you hit a target, eventually. This is
particularly the case with a retrospective, observational study like
this one.
But with care, in particular controlling for your confounders as this
study seems to have done, this is not a problem. Also, having a
shortish list of precise questions - again what this study has done -
helps.
In fact, this study appears to confirm previous research about
hospitalisation and infant feeding in a western setting.
Heather Welford Neil
NCT bfc, tutor, UK
>While it is reasonably hard to leave me speechless, the commentary
>published in this on-line journal did manage to do exactly that.
>
>''Nevertheless, despite the authors having taken measures to address
>potential confounding factors, it is possible that sample sizes as
>large as this study's can lead to associations that are without
>academic merit. "
>
>"Nevertheless"? Too big sample? that is "without academic merit"?!
>
>Grrrr....
>
>But then look at the ads on both sides of the article...
>
>Alla (off my soap box almost speechless)
>
--
***********************************************
Archives: http://community.lsoft.com/archives/LACTNET.html
To reach list owners: [log in to unmask]
Mail all list management commands to: [log in to unmask]
COMMANDS:
1. To temporarily stop your subscription write in the body of an email: set lactnet nomail
2. To start it again: set lactnet mail
3. To unsubscribe: unsubscribe lactnet
4. To get a comprehensive list of rules and directions: get lactnet welcome
|
|
|