The fable of the superiority of randomized studies no longer holds up against the reality of observational studies

[Christophe]

because it neutralizes all the confounding factors: by drawing lots in the SAME control group, all other parameters are statistically equal except the one you want to test.

While if you compare the mortality of young people who came on foot to be tested at the IHU with that of old subclaquants hospitalized in Paris, well you don't just have the treatment that changes, you have plenty of confusing factors.

If you haven't understood this basic principle, or don't admit it, it can be argued forever ....

Sorry ? So it's more of randomization ...

Random is no coincidence, isn't it? I didn't know that now meant hacking

A good study (not only in the medical field) is a study that corresponds as closely as possible to the reality on the ground ... not a study where risks have been neutralized ...

We have clearly seen the limits of randomized vaccine studies ... which have skipped a lot of side effects, some of which are fatal ... side effects that are fine in observational studies ... right?

[GuyGadeboisTheBack]

"A good study is a study whose results drive up the actions of the group that manufactures and markets the product under study." (WJB Rockabilly III)

[ABC2019]

because it neutralizes all the confounding factors: by drawing lots in the SAME control group, all other parameters are statistically equal except the one you want to test.

While if you compare the mortality of young people who came on foot to be tested at the IHU with that of old subclaquants hospitalized in Paris, well you don't just have the treatment that changes, you have plenty of confusing factors.

If you haven't understood this basic principle, or don't admit it, it can be argued forever ....

Sorry ? So it's more of randomization ...

Randomization means that you randomly select the people who take the treatment you want to test and the people who don't take it, without it being correlated to any other criteria. If you pick up people in a large train station and randomly pick who takes a pill and who doesn't (or who takes a placebo, a pill that looks the same but has no product in it), it's "random". , you will have as many men, women, old people, young people in the two samples, the only difference statistically speaking is the pill.
If you don't shoot at random but you take a criterion to choose who you give the pill to, it's no longer random, it's biased. And you introduce a factor of confusion because you will not know if the possible measured effect is really due to the pill or to the choice criterion that you applied to know to whom you are giving it.

Same for "observational" studies where you do not act to give something but you try to correlate two criteria (for example knowing if men or women have more than a certain disease), you can find a spurious correlation due to a phenomenon correlated with the two criteria which you ignore, and which seems to produce an effect opposite to the real one: it is the paradox of Simpson.

[ABC2019]

"A good study is a study whose results drive up the actions of the group that manufactures and markets the product under study." (WJB Rockabilly III)

except that the article you quoted doesn't write that at all, it doesn't say that most of the studies are doctored. He just says that randomization is often not necessary. That said, there are studies that do show that many studies are biased, but this is true of both randomized studies and observational studies. One bias that can easily occur even in randomized studies is the choice of your starting population, which is not random. For example, in a station, you will mainly have people in good condition and having the means to travel, it's not random.

[Janic]

For example, in a station, you will mainly have people in good condition and having the means to travel, it is not random.

But the population of an entire country, or 67.000.000 individuals, is random! Limiting bias much more strongly than on a few hundred or thousand individuals. But it is already too late because we must not confuse fear and reason!

[Obamot]

As I did not understand Simpson ... I read diagonally and it gives Bigard:

- "If you pick up people in a big station and you shoot at random ..."
- "To the selection criteria that you applied to know to whom you are giving it ..."
- “You introduce a confounding factor”
- “identical in appearance but without any product in it”
- “you will have as many men, women,”
- “old, young”
- “it's biased”
- “because you won't know if the possible measured effect is really due ...”
- “If you don't shoot at random”
- “You don't act to give something”
- “Which seems to produce the opposite effect to the real one”

As we are “off real”, Very concept“physical”, I dare not comment, it's too poetic ...

[ABC2019]

For example, in a station, you will mainly have people in good condition and having the means to travel, it is not random.

But the population of an entire country, or 67.000.000 individuals, is random!

it's crazy to be clogged at this point ... if you treat the population of an entire country, you have no point of comparison, you need a treated sample and an untreated sample to know the effect, and the choice of the treated sample must be RANDOM.

And it is totally unnecessary to treat tens of millions of people to find out, a few thousand or tens of thousands is more than enough. But they must have been chosen randomly.

[Obamot]

Reassure us, you “tires”At random in a large station, but at the office in the wagon anyway?

[Janic]

it's crazy to be clogged at this point ... if you treat the population of an entire country, you have no point of comparison, you need a treated sample and an untreated sample to know the effect, and the choice of the treated sample must be RANDOM.

Of course if, you have other countries which have not applied this vaccination (for example) which allows these statistical comparisons.

And it is totally unnecessary to treat tens of millions of people to find out, a few thousand or tens of thousands is more than enough. But they must have been chosen at random.

still clogged up. When in phase 4, side effects appear that have not been identified in phase 2 and 3, it is because these small samples are too weak. as for the distilbene whose effects impacted the following generations, or the vioxx which also made tens of thousands of deaths, etc ...

[ABC2019]

it's crazy to be clogged at this point ... if you treat the population of an entire country, you have no point of comparison, you need a treated sample and an untreated sample to know the effect, and the choice of the treated sample must be RANDOM.

of course yes, you have other countries that have not applied this vaccination (for example) which allows these statistical comparisons.

So you confirm: you did not understand what a randomized sample is.

We suspected a little notice ...