The purpose of statistics is to project the possibility or the probability
of a given event occurring in the future. With Periostat, this means what is
the chance of a good outcome, say meaningful pocket reduction.
The data is obtained from a specific population, say 100 or 1,000 people.
The general population is 6 billion. We need a way of predicting possible
outcomes.
That is the purpose statistics or inferential analysis.
********
Newsgroups: sci.med.dentistry
Subject: Re: Perio treatment
Date: Mon, 23 Apr 2001 05:38:03 GMT
Organization: EarthLink Inc. -- http://www.***.com/
Quote:
> Joel, how do they deteremine the P value? ie. (P>0.05) I was under the
> impression that P values were of more value because the took into
> consideration standard deviation. This is what all the clinical trial
> studies seem to use. Is there something you know that the reviewers
don't?
**********
P value. I wll send another post about this tomorrow. It is a slightly
complicated issue why this is so.
**********
Quote:
> ******
> I'm confused. I asked you how P values are determined, not standard
deviation. You clearly have a vast knowledge of statistical calculation,
could you please explain the P value?
****
Tomorrow .....
*****
I have a vague impression that cement A works well in my hands. As
Quote:
> for patient response to cement A, how would I eliminate confounding
> variables?
I don't understand your question.
*****
Meaning, is cement A better, or is it that it only is better due to some
other unmeasured variable, that is occurring only in my office? This means
that cement B may be in fact, the better cement.
***********************
Quote:
> If you can't trust the research, and you can't trust you own observations,
> what can you do?
> Read Reality, CRA, etc.
Please be more specific. What is the "etc?" Are these publications that
can be trusted to provide more accurate and reliable clinical information
than those that typical dentists and periodontists rely on?
*****
These publications come the closest to a "Consumer Reports: type of
independent assessment of various materials.
Quote:
> Same with golf scores, same with Periostat data.
> Std. Deviation - an unknown to your golfing buddy!
Did a little research on my own....standard deviation takes on importance
only when you lack a control....double blind study, and the control is
valid, Std. Deviation takes on far less significance.
********
Stardard deviation tells people how far the measurement deviates (deviation)
from the mean (from the average). This is very valuable. I suspect that the
Periostat data has a very low standard deviation (meaning the data closely
follows the mean or average).
Your golfing buddy predicts a large standard deviation, meaning that the
average is one-half the thickness of a dime, while several patients of his
obtain three dime''s thickness worth of improvement.
********
Your analogy fails because the golf scores have no control to compare. When
the control and the subject are both placed under the same conditions, then
measured by the same standard, the control is considered zero and the
variance between the control and the test subject is the result.
P value is the probability of a given variance happening as a result of the
tested hypothesis.
He says he sees less bleeding, less pocket depths,
and less non-responding pockets when a patient is prescribed Periostat
and is compliant and faithful in taking the medication.
He is probably fooling himself .... or more likely, his patients!
Quote:
> Ah, so based on his observations, he is either incompetent or unethical?
Well Joel, which is it?
**************
No, I think he is not really that interested in whether or not it works. He
just does what he does. It does not harm anyone so why not do it anyway?
That is the attitude of most.
I am just being a bit more rigorous to encourage better and more effective
therapies.
Joel M. Eichen, D.D.S.
Mailgate References:
Perio treatment, Joel M. Eichen
--
Posted from [38.26.235.37] by way of oe61.pav1.hotmail.com [64.4.30.196]
via Mailgate.ORG Server - http://www.***.com/