[Merged] Immortality & Bayesian Statistics

Status
Not open for further replies.
abaddon said:
I'm gonna take a wild guess here.

Even if you addressed him as Sir Richard of the clan Savage, occasionally known by the moniker "jabba", who signs his own posts "Rich".

You will still be ignored. I don't know why, and, well, tumbleweeds. Why Jabba thinks what he thinks is...unknowable, if he won't tell us.
Click to expand...

TA for the Lelz, friend!
 
Jabba said:
Humots,
- I'm not sure of some of what you're asking, but my (at least) quick interpretation of their claim is that their results show a 99% probability that the "null hypothesis" (that their population of results fall within the non Higgs Boson population of results) is incorrect.
Click to expand...

Jabba, this is a real world example of a scientific model and the data supporting it. What I am asking you is: How do you fit this result into your idea of how conditional probability and Bayes' Theorem applies to scientific models and data?

Scientific Model = Standard Model of particle physics

Data = "more than 99 percent certainty they had found a new elementary particle weighing about 126 times the mass of the proton that was likely the long-sought Higgs boson."

Which of the two statements

1. The Standard Model has a 99 percent certainty of being valid. P(Standard Model|Data) = .99.

2. Evidence that supports the Standard Model has a 99 percent chance of being valid

do you see as following from this result? I think you would say the first statement, while I think it would be the second statement.
 
Jabba said:
Humots,
- I think that the following statement shows that I don't need to say anything about the theory being true...
From http://plato.stanford.edu/entries/bayes-theorem/:
1. Conditional Probabilities and Bayes' Theorem

The probability of a hypothesis H conditional on a given body of data E is the ratio of the unconditional probability of the conjunction of the hypothesis with the data to the unconditional probability of the data alone.
Click to expand...

Humots said:
Note: this is simply a statement of the definition of conditional probability:

P(H|Data) = P(H and Data) / P(Data)
Click to expand...


But that equation is one step away from Bayes' Theorem, since P(H & Data)=P(Data|H)P(H), where P(H) is the prior probability of the hypotheses.

I think you may be confusing two different uses of the word "hypothesis".

Hypothesis: John Doe died in 2000

Hypothesis: a = G M / R^2

The first hypothesis is about an event.

The second hypothesis is a scientific model.

I don't believe that probability can be applied to both hypotheses in the same way, but I may be wrong. My knowledge of Bayes' Theorem is mostly about the math, not about its applications.
Click to expand...


In Bayesian statistics it is perfectly acceptable to talk about the probability of a hypothesis; indeed, that's kinda' the whole point of Bayesian statistics. You begin with a prior probability that H is true, based on all the available evidence before an experiment is conducted, and then you update that probability by combining the evidence for the H from the experimental data with the prior probability to form a posterior probability for H, P(H|Data).

Consider the following statement (Higgs boson based on the new data from the experiment, to performconfirmed):

"Physicists announced on July 4, 2012, that, with more than 99 percent certainty, they had found a new elementary particle weighing about 126 times the mass of the proton that was likely the long-sought Higgs boson."

Does this mean that
- the Standard Model has a 99 percent certainty of being valid, or
- evidence that supports the Standard Model has a 99 percent chance of being right?
Click to expand...


What the statement literally means is that the probability for the discovery of a new boson, given the data, is greater than .99; that is, as worded, the statement is a posterior probability of a hypothesis. However, the statement—widely quoted as it has been—is false, because physicists did not report a posterior probability; they reported a p-value, which is the probability of the observed data, or more extreme data, assuming that the null hypothesis is true.
 
jt512 said:
But that equation is one step away from Bayes' Theorem, since P(H & Data)=P(Data|H)P(H), where P(H) is the prior probability of the hypotheses.

In Bayesian statistics it is perfectly acceptable to talk about the probability of a hypothesis; indeed, that's kinda' the whole point of Bayesian statistics. You begin with a prior probability that H is true, based on all the available evidence before an experiment is conducted, and then you update that probability by combining the evidence for the H from the experimental data with the prior probability to form a posterior probability for H, P(H|Data).
Click to expand...

OK, but my basic question was, is Jabba correct in stating that the word hypothesis in the above statement can refer to a scientific model or theory in its entirety?

I am assuming that this is part of what Jabba has been going on about.

jt512 said:
What the statement literally means is that the probability for the discovery of a new boson, given the data, is greater than .99; that is, as worded, the statement is a posterior probability of a hypothesis. However, the statement—widely quoted as it has been—is false, because physicists did not report a posterior probability; they reported a p-value, which is the probability of the observed data, or more extreme data, assuming that the null hypothesis is true.
Click to expand...

I'm not clear on the meaning of the term "null hypothesis". In this case, would the null hypothesis be that a new boson has been discovered? And that "the probability of the observed data, assuming that the Higgs boson does exist, is .99?"
 
jt512 said:
B
What the statement literally means is that the probability for the discovery of a new boson, given the data, is greater than .99
Click to expand...

And the discovery of a new boson is support or evidence for the validity of the Standard Model?

And, as Jabba seems to think, does this have anything to do with the probability that the Standard Model is true?

Note: I strongly suspect that the statement "the probability that the Standard Model is true" is not the correct way of speaking. I just think that Jabba, in his application of conditional probability and Bayes' Theorem to scientific models, is saying something similar.
 
Humots said:
And the discovery of a new boson is support or evidence for the validity of the Standard Model?

And, as Jabba seems to think, does this have anything to do with the probability that the Standard Model is true?

Note: I strongly suspect that the statement "the probability that the Standard Model is true" is not the correct way of speaking. I just think that Jabba, in his application of conditional probability and Bayes' Theorem to scientific models, is saying something similar.
Click to expand...

Yeah but it's quantum. Quantum allows us to make up any durn thing and claim it's true, as any fule kno.

This is no different than the quantum shroud claim. You know, the one about the levitating radioactive cloth baloney. Same here. It's quantum, therefore any invention is equiprobabably true.
 
Humots said:
OK, but my basic question was, is Jabba correct in stating that the word hypothesis in the above statement can refer to a scientific model or theory in its entirety?
Click to expand...


Yes, a hypothesis can be that a particular scientific model or theory is true, and you can make a probability statement about it.

I'm not clear on the meaning of the term "null hypothesis". In this case, would the null hypothesis be that a new boson has been discovered?
Click to expand...


No. That would be the alternative hypothesis. The null hypothesis would be that there is no new elementary particle in the energy range studied.

And that "the probability of the observed data, assuming that the Higgs boson does exist, is .99?"
Click to expand...


No. The quoted p-value means that the probability of the observed data, or more extreme data, would be (greater than) .99 if there were no new elementary particle.
 
Humots said:
And the discovery of a new boson is support or evidence for the validity of the Standard Model?

And, as Jabba seems to think, does this have anything to do with the probability that the Standard Model is true?
Click to expand...


I'm not a physicist, but as I understand it, the Standard Model predicted the Higgs boson, so discovery of the boson would be evidence in favor of the Standard Model.

Note: I strongly suspect that the statement "the probability that the Standard Model is true" is not the correct way of speaking. I just think that Jabba, in his application of conditional probability and Bayes' Theorem to scientific models, is saying something similar.
Click to expand...


Jabba is saying something similar, but it is perfectly acceptable (and common statistical practice) to speak about the probability of a scientific theory. Jabba has gotten that much correct (though not much else).
 
Slowvehicle said:
Mr. Savage: When I called you "Rich", or "Richard", you complained about it. Are you honestly going to take time from your glacial progress to complain about being formally addressed? Shall I address you as "Ootliender Pedoonkee"?

When do you intend to address the problem of the utter lack of any evidence of the existence of a "soul"; or the problem of a fixed number of immortal "souls" compared to increasing populations, or the fact that interrupted existence without continuity is not immortality?
Click to expand...
Slowvehicle,
- "Rich," by itself, is not condescending; "Oh Rich" is. I wasn't complaining about "Rich"; I was complaining about "Oh Rich."
 
jt512 said:
...
Jabba is saying something similar, but it is perfectly acceptable (and common statistical practice) to speak about the probability of a scientific theory. Jabba has gotten that much correct (though not much else).
Click to expand...
Jay,
- Good to hear from you.
- You've probably told me already what you think is incorrect about my argument. Please refresh my memory -- one or two issues would be enough to get me started. Thanks.
 
Last edited:
Jabba,

You've been posting to this thread for months. Could you perhaps summarize what you think the main position/positions of the people you are arguing with are? The arguments don't seem do have changed too much for quite some time. You surely don't need Akhenaten to provide you with discussion points that the rest of the forum believe you haven't addressed.

How about the "special snowflake" argument, for example. Why do you feel that fails? Why do you think other forum members think it demolishes everything you're claiming?
 
Jabba said:
Jay,
- Good to hear from you.
- You've probably told me already what you think is incorrect about my argument. Please refresh my memory -- one or two issues would be enough to get me started. Thanks.
Click to expand...

Didn't you make a list of issues and questions so that we wouldn't have to play the "refresh my memory" game anymore?
 
Jabba said:
Slowvehicle,
- "Rich," by itself, is not condescending; "Oh Rich" is. I wasn't complaining about "Rich"; I was complaining about "Oh Rich."
Click to expand...

That's rubbish. When I leave the office, I do not accuse my colleagues of condescension if they say "Oh, Dave grab me a sammich on your way back"
 
Status
Not open for further replies.

Back
Top Bottom