trishrhymes
Senior Member
- Messages
- 2,158
Unfortunately stat-check is unlikely to pick up the fact that PACE is a 'pile of crap' to quote David Tuller.
As I understand it, all stat-check does is pick up anomalies in calculated p-values which may be little more than rounding errors, or the correct application of more sophisticated corrections to p-values. As some have said, it's little more than the statistical equivalent of a spell checker.
I doubt it would show up anything in PACE, or if it does show anything, it won't be the fundamental flaws in the methodology.
Stat-check cannot reach the heart of the problem:
- inherent bias in un-blinded psychological, questionnaire based 'research',
- using the wrong cohort of patients (fatigue, not ME)
- outcome switching,
etc.
all of which make PACE unreliable, indeed fraudulent, in a much more fundamental way.
As I understand it, all stat-check does is pick up anomalies in calculated p-values which may be little more than rounding errors, or the correct application of more sophisticated corrections to p-values. As some have said, it's little more than the statistical equivalent of a spell checker.
I doubt it would show up anything in PACE, or if it does show anything, it won't be the fundamental flaws in the methodology.
Stat-check cannot reach the heart of the problem:
- inherent bias in un-blinded psychological, questionnaire based 'research',
- using the wrong cohort of patients (fatigue, not ME)
- outcome switching,
etc.
all of which make PACE unreliable, indeed fraudulent, in a much more fundamental way.