Statistical Tests and the Law

Originally posted by bbarr
He's claiming you're being disingenuous.

It is interesting to note that in the U.S., there do in fact
exist at least two separate and distinct null hypotheses, depending
on whether you are operating under criminal procedure
or civil procedure.

In a criminal trial, the jury is usually instructed to return
a not-guilty verdict unless they have been convinced
"beyond a reasonable doubt" of the guilt of the accused.

In a civil procedure, the "preponderance of the evidence"
must show the plaintiff's claims to be true.

This may be like sliding your confidence threshold, with
a much higher one needed to convict a criminal.

Further, the judge in a criminal case writes his own instructions
to the jury on a case by case basis before they deliberate.
The variations in these instructions would serve as subtly different
null hypotheses for each case that is heard.

Dr. Cribs

Originally posted by royalchicken
Ivanhoe, in addition to being sarcastic, they are discussing a serious subject, namely the relationship between the concept of a null hypothesis and the idea of ''innocent until proven guilty''. Please respect their right to discu ...[text shortened]... ent, and if you have ideas on the subject, please join the debate.

Originally posted by bekieke
Give it a rest rc, unless you live in a cave it's obvious what's going on.

Go ahead, bump me, ban me, send me straight to hell. I really don't care anymore after reading all this dribble that started a few days ago as a result of a thread I initiated.

There have been people talking about what ruined other sites for them and I see the same thing happ ...[text shortened]... e alley behind the damn bar. Don't bother me while I enjoy a beer watching baseball on the tv.

Originally posted by royalchicken
This thread has two purposes, at least one of which is to have a serious discussion about legal presumption.

Originally posted by Cribs
I think 'legal presumptions' are a kind of beast that are
partially analogous to halting and modifying a hypothesis
test mid-experiment, modifying the hypothesis mid-experiment,
and then continuing to the conclusion.

In other words, during the course of an "innocent until
guilty" proceeding, a 'legal presumption' could be triggered
that switches the burden of proof mid-trial.

Dr. Cribs

Originally posted by royalchicken
In inferential statistics, one makes a guess about the value of a parameter in the population--the null hypothesis--and compares the value to measured statistics, finally reaching a conclusion about whether or not the null hypothesis is likely to be true.

In a legal proceeding or informal discussion about someone's culpability, the notion of ''inn ...[text shortened]... ocent until proven guilty'' might better suit certain situations? Is the analogy a bad one?

Originally posted by royalchicken
While I respect the fact that you go out of your way to (perfectly) speak a language other than your mother tongue to post on RHP, and that I am not capable of doing this, can you please translate?