|
Did the jews learn to fight HIV | by hej | 2006-11-19 12:55:59 |
|
It's bogus | by Imp | 2005-08-04 11:44:07 |
|
The article said you can still get it if you are | by carthax | 2005-08-04 11:49:37 |
|
The study was unscientific. | by BloodyViking | 2005-08-04 11:55:43 |
|
The exposure is "randomised", just | by hej | 2005-08-04 11:59:47 |
| Clarification on the term "randomised study". |
by Pic |
2005-08-04 12:19:40 |
A double blind, randomised study is executed as follows:
You take your study poulation and randomly divide it into two groups (this is why it's called randomised).
Now, one part of your population is given the real thing you want to test, the other is given a placebo that is indistinguishable from the real drug without submitting it to chemical analysis (and they're really anal retentive on this. Every imaginable characteristic has to be identical). The only ones who know who are being treated with what are the ones directing the study, the patients and the doctors treating them have no idea what they are receiving/administering.
Only random thing is who gets what. After that everything you can imagine, and probably then some, is being controlled and measured. |
|
[ Reply ] |
|
How do you give a placebo circumsision? :) (n/t) | by MikeCDN | 2005-08-04 12:25:52 |
|
That is of course a problem, | by Pic | 2005-08-04 12:35:10 |
|
Are you kidding? | by BloodyViking | 2005-08-04 12:42:55 |
|
The random factor there is sort of inevitable. | by Pic | 2005-08-04 12:59:25 |
|
but we have just over a 2% difference | by MikeCDN | 2005-08-04 13:06:56 |
|
The relevance of the conclusion | by Pic | 2005-08-04 13:10:05 |
|
Of Course the methodology is inherently wrong. | by BloodyViking | 2005-08-04 13:15:15 |
|
So statistics is crap? (n/t) | by Pic | 2005-08-04 13:19:13 |
|
Frequently. In this particular case, definitely. | by BloodyViking | 2005-08-04 13:25:50 |
|
The best option I can think of, | by BloodyViking | 2005-08-04 13:10:18 |
|
That would indeed be useful, | by Pic | 2005-08-04 13:14:27 |
|
but statistically | by MikeCDN | 2005-08-04 13:17:11 |
|
That's not really my point. | by Pic | 2005-08-04 13:32:58 |
|
The sample is far from large enough | by BloodyViking | 2005-08-04 13:22:17 |
|
Side thought | by MikeCDN | 2005-08-04 13:29:51 |
|
That would invalidate the data. | by Peace_man | 2005-08-04 13:40:42 |
|
Yeah, | by Pic | 2005-08-04 13:46:18 |
|
I suspect this is just a 'quick and dirty' initial | by Peace_man | 2005-08-04 15:19:54 |
