CFLarsen said:
Now, what if the numbers were temperatures? The statistician does not know what the numbers mean.
The statistician would know they were temperatures (just like he knows they are blood pressures), he just wouldn't have access to the group information.
What does he compare them to?
The other groups, other numbers, an "alpha" level, like he normally would in any statistical analysis.
Stop. There are problems.
No, there aren't any. We are just doing 'normal statistics', but are recoding the group information.
Where, in your example, do you present the average blood pressure data?
The point, as already noted, was to show how the statistician would be blinded to group membership information. I really can keep repeating this all day if need be.
Nope.
You give an example, and when it is explained that it is thoroughly inadequate, you deem it "a quick example".
Too bad you weren't convincing on why it is supposedly thoroughly inadequate as you believe.
No? When I ask for examples, I am not asking for sloppiness.
Yawn.
As noted already, my example was to show how one could analyze data without the group information. It's not really difficult.
I was not referring to the specific numbers, but to the fact that:
1) You did not know how blood pressure was measured.
2) You had no idea in what range a blood pressure would be.
All in all, you had no idea what you were doing.
Then you are incredibly nieve to think 1) and 2). It is very amusing as both of my theses directly involved blood pressure.

One was predicting heart disease, and the other was determining the reliability of blood pressure measuring devices. Do let me know if you want autographed copies of either, and I'd be happy to oblige.
But continuing on.. I mentioned that the numbers were entirely made up. I already mentioned that the example was to show how one could analyze data without the group information.
..., you have come up with an entirely inadequate and flawed example that does not explain what triple-blind is.
I showed how one could recode group information to make the real group labels unaware to the evaluator. That is the "guts" of a triple blind. The rest proceeds as any standard statistical analysis would, and therefore doesn't need to be shown, as it is entirely standard after that point.
Of course, when cornered, just claim that people are stupid.
That's what you do to apparently. Observe:
"All in all, you had no idea what you were doing."
"Could it be that you simply don't know?"
How does triple-blinding work?
Already answered. The group information is recoded so the evaluator doesn't know real group membership of individuals.
Oh, and in anticipating your next post:
Already answered. The group information is recoded so the evaluator doesn't know real group membership of individuals.