ISF Logo   IS Forum
Forum Index Register Members List Events Mark Forums Read Help

Go Back   International Skeptics Forum » General Topics » General Skepticism and The Paranormal
 


Welcome to the International Skeptics Forum, where we discuss skepticism, critical thinking, the paranormal and science in a friendly but lively way. You are currently viewing the forum as a guest, which means you are missing out on discussing matters that are of interest to you. Please consider registering so you can gain full use of the forum features and interact with other Members. Registration is simple, fast and free! Click here to register today.
Reply
Old 21st May 2020, 08:26 PM   #1
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
When is an event evidence for it's explanation?

In the politics thread, there was a link to a story of one state swapping dates for daily infections, making the numbers look better....three different times.

Replies indicated this was evidence it was on purpose.

This is a concept in general that has seemed flawed to me the many times it comes up...


X is some event (getting it wrong three times)
Y is the theory for the reason of X (on purpose)
X is implied to be the evidence for Y (it was on purpose because it was done three times).

Can an event be evidence for the explanation? Or does the explanation of an event require evidence outside of X?
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 21st May 2020, 09:29 PM   #2
The Great Zaganza
Maledictorian
 
The Great Zaganza's Avatar
 
Join Date: Aug 2016
Posts: 12,090
Originally Posted by BobTheCoward View Post

Can an event be evidence for the explanation? Or does the explanation of an event require evidence outside of X?
yes and yes (because of false dichtonomy)
__________________
ETTD
Everything Trump Touches Dies
The Great Zaganza is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 21st May 2020, 10:50 PM   #3
Swordfishtrombone
Illuminator
 
Swordfishtrombone's Avatar
 
Join Date: May 2015
Location: Helsinki, Finland
Posts: 3,609
^Agreed with The Great Zaganza.

X is evidence for Y, but not proof of Y.

To establish Y, more evidence may be needed, such as other forms of deceit showing a pattern, or a clear agenda, by the state involved.

Also, how strong evidence X is for Y, depends on how many alternative explanations for X there are, and how likely they are to happen.

Is it easy to mix up the data for the dates on accident, or is it unlikely?
Swordfishtrombone is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 02:38 AM   #4
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
Originally Posted by BobTheCoward View Post
Can an event be evidence for the explanation? Or does the explanation of an event require evidence outside of X?
As others have said, evidence is not a black-and-white concept. Getting something wrong three times, also, is not a single event; it's three separate events. So we can look at the probabilities of getting it wrong once, twice, and three times, and note that for every repetition the odds are altered. Of course, we don't have actual statistics, so there's a large degree of interpretation involved, bordering on guesswork. But if we note that, hypothetically, the error has occurred three times in one state and never in any other, then we can assume first that the probability of error is very low, secondly that an occasional isolated error is unremarkable, and thirdly that the probability of three errors closely spaced is so remote that a more likely explanation should be sought. If all errors are then found to be in the same direction, that's suggestive of manipulation; however, it's weaker evidence, because the odds are one in eight that all errors should be in the direction that makes the state look good, not an outlandish improbability.

The explanation of the event doesn't absolutely require evidence outside of the event itself. If a whistleblower came forward and stated that the numbers had been manipulated, that would be stronger evidence; but still not irrefutable, because they might be lying for political or personal reasons.

Ultimately, there's no objectively rigorous standard by which evidence can be declared sufficient or insufficient; there's only human judgement.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 05:53 AM   #5
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by Swordfishtrombone View Post
^Agreed with The Great Zaganza.

X is evidence for Y, but not proof of Y.

To establish Y, more evidence may be needed, such as other forms of deceit showing a pattern, or a clear agenda, by the state involved.

Also, how strong evidence X is for Y, depends on how many alternative explanations for X there are, and how likely they are to happen.

Is it easy to mix up the data for the dates on accident, or is it unlikely?
Is it evidence of a prerequisite?

For an explanation of the reason behind an event, the event must have occurred. But is it any better evidence for a pedestrian explanation than an outlandish one?
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 06:20 AM   #6
Loss Leader
I would save the receptionist.
Moderator
 
Loss Leader's Avatar
 
Join Date: Jul 2006
Location: Florida
Posts: 27,573
So, here it goes in real life: The driver of a hit-and-run in Vermont is noted to have been in a red Ford Taurus with Nevada plates. Police (using magic) search for all red Ford Tauruses with Nevada plates that were in Vermont on that date.

Is the identification of Joe Shmoe as one of those drivers evidence that he committed the hit-and-run?
__________________
I have the honor to be
Your Obdt. St

L. Leader
Loss Leader is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 06:22 AM   #7
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by Loss Leader View Post
So, here it goes in real life: The driver of a hit-and-run in Vermont is noted to have been in a red Ford Taurus with Nevada plates. Police (using magic) search for all red Ford Tauruses with Nevada plates that were in Vermont on that date.

Is the identification of Joe Shmoe as one of those drivers evidence that he committed the hit-and-run?
This conversation is about explanations for events. Identification is separate.
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 06:32 AM   #8
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
OK, how about this.

I roll a pair of dice. Both come up 6.
I roll them again. Both come up 6.
I roll them again. Both come up 6.

Is this evidence that both dice are weighted or otherwise rigged so that they always, or almost always, roll a 6?

We have two possibilities: they are rigged, or they are not. If they are not, then a series of events with a probability of about 1 in 50,000. has just occurred. If they are, then a series of events with an unknown but very high probability has occurred.

I would consider this evidence that the latter, not the former, has occurred.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 06:34 AM   #9
Bikewer
Penultimate Amazing
 
Bikewer's Avatar
 
Join Date: Sep 2003
Location: St. Louis, Mo.
Posts: 12,723
I couldn’t recall the principal, had to look it up...

“Hanlon's razor” is a saying that reads: Never attribute to malice that which is adequately explained by stupidity.

Possible in this case?
Bikewer is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 06:40 AM   #10
Thermal
Penultimate Amazing
 
Thermal's Avatar
 
Join Date: Aug 2016
Location: It was Lothian's idea
Posts: 11,040
Originally Posted by Dave Rogers View Post
OK, how about this.

I roll a pair of dice. Both come up 6.
I roll them again. Both come up 6.
I roll them again. Both come up 6.

Is this evidence that both dice are weighted or otherwise rigged so that they always, or almost always, roll a 6?

We have two possibilities: they are rigged, or they are not. If they are not, then a series of events with a probability of about 1 in 50,000. has just occurred. If they are, then a series of events with an unknown but very high probability has occurred.

I would consider this evidence that the latter, not the former, has occurred.

Dave
The roll of any three particular outcomes would be exactly the same odds, though. Isn't it just in out minds that this particular one is significant?
__________________

Whenever you find yourself on the side of the majority, it is time to pause and reflect -Mark Twain
Thermal is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:10 AM   #11
JayUtah
Penultimate Amazing
 
JayUtah's Avatar
 
Join Date: Sep 2011
Posts: 18,675
Originally Posted by Thermal View Post
The roll of any three particular outcomes would be exactly the same odds, though. Isn't it just in out minds that this particular one is significant?
Yes and no. You're completely correct in saying that if you have a uniform random variable, a particular set of outcomes of three trials is no more likely than any other set of three outcomes from the same variable.

But the hypothesis here is whether the dice are fair. That is, whether the variable is uniformly random. If you simplify the exercise to one die, you can say that over a number of trials, the distribution of single outcomes from such a variable is expected to fit a uniform distribution: a flat curve governed by a uniform p = 1/6 for each possible trial outcome. We can say with what probability an actual run of trials fits that -- or any -- distribution. That probability becomes more confident the more trials you perform. What David Rogers is saying with his particular formulation of the exercise is that even with N=3, we can see that the outcomes don't fit a uniform distribution according to a significant p-value.

It gets even more fun when you take the Bayesian approach.
JayUtah is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:25 AM   #12
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
An event E is evidence for a hypothesis H iff P(H|E) > P(H|E u ~E)

ETA: Note that this doesn't mean H should necessarily be considered true given E. For example consider the following:
P(H|E u ~E) = 5%
P(H|E) = 10%
then E is evidence for H but the probability of H being true given E is still only 10%, in other words there's still a 90% chance that H is false.
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin

Last edited by caveman1917; 22nd May 2020 at 07:27 AM.
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:37 AM   #13
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by caveman1917 View Post
An event E is evidence for a hypothesis H iff P(H|E) > P(H|E u ~E)

ETA: Note that this doesn't mean H should necessarily be considered true given E. For example consider the following:
P(H|E u ~E) = 5%
P(H|E) = 10%
then E is evidence for H but the probability of H being true given E is still only 10%, in other words there's still a 90% chance that H is false.
It seems like the event itself and the probability of the event is separate evidence. Stating X happened without stating the probability of X is what is occuring here.
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:39 AM   #14
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by BobTheCoward View Post
In the politics thread, there was a link to a story of one state swapping dates for daily infections, making the numbers look better....three different times.

Replies indicated this was evidence it was on purpose.

This is a concept in general that has seemed flawed to me the many times it comes up...


X is some event (getting it wrong three times)
Y is the theory for the reason of X (on purpose)
X is implied to be the evidence for Y (it was on purpose because it was done three times).

Can an event be evidence for the explanation? Or does the explanation of an event require evidence outside of X?

You might be concerned about [Affirming The Consequent]?

Model:
P1: IF A THEN B
P2: B
C: THEREFORE A

Example from the OP:
P1: IF [The State is deliberately lying] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is deliberately lying]

Is it a fallacy? Sure.

Here's the problem - the challenge is to determine whether this is the correct argument analysis model. Just to throw out a counterexample...

Example 2:
P1: IF [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed] THEN [the drug probably works]
P2: [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed]
C: [the drug probably works]

Uh-oh, we've just "proven" that the entire concept of modern medicine is a fallacious hoax? Well, I hope not. It's more likely that critical thinking about the real world is more complicated than just shopping around with an argument we don't like to find a fallacy hole to hammer it into.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:40 AM   #15
theprestige
Penultimate Amazing
 
Join Date: Aug 2007
Location: Hong Kong
Posts: 45,225
I'm not a logimatician, so I don't have fancy symbols or formal terms. But the way I see it is this:

First, it has to be predicted by the explanation. You don't actually have to say it ahead of time, but "if X is true, we should expect to see Y" has to be a true statement for explanation X and observed event Y.

Second, it can't be predicted by a competing explanation. An even that is evidence for both is effectively evidence for neither.

Third, evidence is not the same as proof.
theprestige is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:54 AM   #16
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by theprestige View Post
I'm not a logimatician, so I don't have fancy symbols or formal terms. But the way I see it is this:

First, it has to be predicted by the explanation. You don't actually have to say it ahead of time, but "if X is true, we should expect to see Y" has to be a true statement for explanation X and observed event Y.
In this case, I'll say the criteria is satisfied. "If the state wants to mislead the public they will publish false stats."


Originally Posted by theprestige View Post
Second, it can't be predicted by a competing explanation. An even that is evidence for both is effectively evidence for neither.
In this case, for example, "Incompetence" is a competing explanation. And I have to say, a compelling one.


Originally Posted by theprestige View Post
Third, evidence is not the same as proof.
Yep.

When I was doing ghost investigations, the ghost hunters would say things like, "I feel a presence," or they'd show me orb photos, sometimes they'd report that they literally saw a ghost. This is all 'evidence' but then we put it into the critical thinking meatgrinder and as you point out in point#2, there are satisfactory naturalistic explanations for these. The Natural Sciences don't really have proof, just increasing confidence in one explanatory model.

The challenge is that once you have competing models, you get into calculating probabilities which are often dependent on baselines that have to be assigned.

And this depends on worldviews, which involve trust of information sources and accumulated information drawn from them. We assign probabilities based on these worldviews.


Just as an example from this thread, above I mentioned that a perfectly good explanation for publishing misleading stats three times is incompetence. And unfortunately there's no research we can point to that tells us what the prior probability of incompetence vs deception is for this exact case. We're all just going to go with our gut, juggling a feeling about whether this is a case of 'government in general is incompetent, so high probability' or other partisan interpretations.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett

Last edited by blutoski; 22nd May 2020 at 07:55 AM.
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 07:58 AM   #17
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by Dave Rogers View Post
OK, how about this.

I roll a pair of dice. Both come up 6.
I roll them again. Both come up 6.
I roll them again. Both come up 6.

Is this evidence that both dice are weighted or otherwise rigged so that they always, or almost always, roll a 6?

We have two possibilities: they are rigged, or they are not. If they are not, then a series of events with a probability of about 1 in 50,000. has just occurred. If they are, then a series of events with an unknown but very high probability has occurred.

I would consider this evidence that the latter, not the former, has occurred.

Dave
Let's make this more fun

Suppose I have a die. I roll it once, it comes up 6.

I am considering two hypotheses, H1 = "the die is fair" and H2 = "the die is rigged to produce the pattern 6,5,4,3,2,1,6,5,4,3,2,1,...".

Is me rolling a 6 evidence for H2?
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin

Last edited by caveman1917; 22nd May 2020 at 08:00 AM.
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:14 AM   #18
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
Originally Posted by blutoski View Post
You might be concerned about [Affirming The Consequent]?

Model:
P1: IF A THEN B
P2: B
C: THEREFORE A

Example from the OP:
P1: IF [The State is deliberately lying] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is deliberately lying]

Is it a fallacy? Sure.

Here's the problem - the challenge is to determine whether this is the correct argument analysis model. Just to throw out a counterexample...

Example 2:
P1: IF [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed] THEN [the drug probably works]
P2: [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed]
C: [the drug probably works]

Uh-oh, we've just "proven" that the entire concept of modern medicine is a fallacious hoax? Well, I hope not. It's more likely that critical thinking about the real world is more complicated than just shopping around with an argument we don't like to find a fallacy hole to hammer it into.
Uh, no. You've confused your antecedent and your consequent. In your medical example, A is [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed], B is [the drug probably works], and your statement boils down to:

P1: IF A THEN B
P2: A
P3: THEREFORE B

Which is, of course, logically sound.

And, of course, the medical example doesn't work as a good one anyway, because we expect a drug that works to show significantly significant improvement over a control group; we can therefore say that IF B THEN A is as true as IF A THEN B, because A is in fact the definition of B.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:16 AM   #19
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
Originally Posted by caveman1917 View Post
Let's make this more fun

Suppose I have a die. I roll it once, it comes up 6.

I am considering two hypotheses, H1 = "the die is fair" and H2 = "the die is rigged to produce the pattern 6,5,4,3,2,1,6,5,4,3,2,1,...".

Is me rolling a 6 evidence for H2?
Very, very weak evidence, yes. With subsequent rolls the evidence gets stronger, but can never reach the status of absolute proof. And, of course, one roll disagreeing with the hypothesis falsifies it however many preceding rolls have not.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:17 AM   #20
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by caveman1917 View Post
Let's make this more fun

Suppose I have a die. I roll it once, it comes up 6.

I am considering two hypotheses, H1 = "the die is fair" and H2 = "the die is rigged to produce the pattern 6,5,4,3,2,1,6,5,4,3,2,1,...".

Is me rolling a 6 evidence for H2?
Note that what you are showing with that experiment is that the outcomes are no longer random. But not why. Was the die deliberately rigged, or was it a manufacturing error or later damage?

The problem is that it also reinforces the alternative explanation: incompetence.

What you have is support for any model that doesn't make the results independent of each other. It supports any model where the publications have a shared root cause. It won't help you distinguish between malevolence or incompetence, it just increases confidence that there's a consistent methodology to their publication process.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett

Last edited by blutoski; 22nd May 2020 at 08:18 AM.
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:22 AM   #21
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by Dave Rogers View Post
Very, very weak evidence, yes. With subsequent rolls the evidence gets stronger, but can never reach the status of absolute proof. And, of course, one roll disagreeing with the hypothesis falsifies it however many preceding rolls have not.

Dave
Well... evidence for what, though? This is why Affirming The Consequent is a fallacy - it doesn't address alternative models.

eg:

Model:
P1: IF A THEN B
P2: B
C: THEREFORE A

Example from the OP:
P1: IF [The State is deliberately lying] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is deliberately lying]


Or, another example:
P1: IF [The State is incompetent] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is incompetent]

The evidence is kinda neutral using this argument structure.

If both models produce opposite results from the same evidence, then this is not a good roadmap to answering the question of whether the state was malevolent or incompetent.

What I'm getting at is: we need to ask a different question.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:24 AM   #22
JayUtah
Penultimate Amazing
 
JayUtah's Avatar
 
Join Date: Sep 2011
Posts: 18,675
Originally Posted by Dave Rogers View Post
Very, very weak evidence, yes. With subsequent rolls the evidence gets stronger, but can never reach the status of absolute proof. And, of course, one roll disagreeing with the hypothesis falsifies it however many preceding rolls have not.

Dave
For real-world evidence there will always be that inductive leap. For nearly all scientific purposes, leaps of p < 0.05 are considered acceptable. For other purposes, the width of the leap is difficult to reckon because we don't have a good idea what numbers to assign as likelihoods of the various events that occur. We tend to reason intuitively, and as Vegas teaches us, that often leads to wrack and ruin.

Broadly speaking, evidence is any information that tends to change the likelihood that a particular hypothesis is true. Beyond that, we use words like "compelling," "solid," and so forth to describe the degree to which the likelihood is altered. Only in some cases can we quantify that effect to any helpful degree of confidence, and only because we've constrained the observations to produce that kind of data. Usually the reasoning alludes to statistical approaches, but can't always apply them.
JayUtah is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:27 AM   #23
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by Dave Rogers View Post
Very, very weak evidence, yes. With subsequent rolls the evidence gets stronger, but can never reach the status of absolute proof. And, of course, one roll disagreeing with the hypothesis falsifies it however many preceding rolls have not.

Dave
I would rather say that it's pretty strong evidence but that H2 is a very, very weak hypothesis. Depends on how you define things, but let's define the strength of evidence E for hypothesis H as:

P(H|E) / P(H|E u ~E)

and the strength of hypothesis H before accounting for evidence E as:

P(H|E u ~E)

and the strength of hypothesis H after accounting for evidence E as:

P(H|E)

Then we should say that H2 is a very, very weak hypothesis that quickly gets stronger because each die roll consistent with it is strong evidence for it. There's a difference between the strength of evidence (how much the relative probability of a hypothesis changes after accounting for said evidence) and the strength of a hypothesis (the prior and posterior probability of a hypothesis). Suppose we live in a world with two types of dice equally distributed, those which follow H1 and those which follow H2. Then after rolling a 6 you would be reasonably sure you've got one of the H2 dice, but the strength of the evidence is the same.
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin

Last edited by caveman1917; 22nd May 2020 at 08:55 AM.
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:32 AM   #24
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by Dave Rogers View Post
Uh, no. You've confused your antecedent and your consequent. In your medical example, A is [stage 1, 2, and 3 double blinded placebo controlled clinical trials show the experimental group achieves statistically significant improvement over the control group and the results are replicated independently and peer reviewed], B is [the drug probably works], and your statement boils down to:

P1: IF A THEN B
P2: A
P3: THEREFORE B

Which is, of course, logically sound.

And, of course, the medical example doesn't work as a good one anyway, because we expect a drug that works to show significantly significant improvement over a control group; we can therefore say that IF B THEN A is as true as IF A THEN B, because A is in fact the definition of B.

Dave
Ah, you're right. I need to do this after I get my coffee. I was jumbling this with my Fallacy Of Composition example. (clinical trials are fallacious because they are fallacy of composition).

But the point is that generally science modelling is hypothesis based, models are antecedents, results consequents. eg: if the big ban is true, we should see cosmic background radiation. Observing CMB supports BB theory, but doesn't prove it.

So, to use a clinical trial example:

Model:
P1: IF A THEN B
P2: B
C: THEREFORE A

Example from the OP:
P1: IF [drug A works] THEN [clinical trials show positive results]
P2: [clinical trials show positive results]
C: THEREFORE [drug A works]

This is fallacious, and that's OK, because we have to shift to more advanced critical thinking techniques to assign probabilistic outcomes and use those for real world decisionmaking.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:39 AM   #25
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
Originally Posted by blutoski View Post
Example from the OP:
P1: IF [The State is deliberately lying] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is deliberately lying]


Or, another example:
P1: IF [The State is incompetent] THEN [they publish false stats over and over]
P2: [they publish false stats over and over]
C: THEREFORE [The State is incompetent]

The evidence is kinda neutral using this argument structure.
Yes, true. However, the event in this case was not that the State published false stats repeatedly; it was that the State published false stats repeatedly in such a way that they supported the narrative the State wanted them to support better than the real statistics (let's define that as "positive spin"). One might reasonably expect that incompetence would be randomly distributed between positive and negative spin, so when all the errors observed result in positive spin there's a suggestion that this is deliberate rather than incompetent. However, it's a very small data set, so I wouldn't expect the probability of the hypothesis being true to be compellingly high.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:39 AM   #26
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by caveman1917 View Post
I would rather say that it's pretty strong evidence but that H2 is a very, very weak hypothesis. Depends on how you define things, but let's define the strength of evidence E for hypothesis H as:

P(H|E) / P(H|E u ~E)

and the strength of hypothesis H as:

P(H)

and the strength of hypothesis H after accounting for evidence E as:

P(H|E)

Then we should say that H is a very, very weak hypothesis that quickly gets stronger because each die roll consistent with it is strong evidence for it. There's a difference between the strength of evidence (how much the relative probability of a hypothesis changes after accounting for said evidence) and the strength of a hypothesis (the prior and posterior probability of a hypothesis). Suppose we live in a world with two types of dice equally distributed, those which follow H1 and those which follow H2. Then after throwing a 6 you would be reasonable sure you've got one of the H2 dice, but the strength of the evidence is the same.
I guess where I was going is that this is evidence for a botched die, but not necessarily evidence for a specific cause for the wonkiness.

But once wonkiness is established, that can be a premise for an argument about whether it's intentional on the part of the player. Did they deliberately buy a weighted die? Did they find it in the street? Was it given to them by a friend who wanted to frame them for cheating?

I think the question in the thread is about the state's intention, which can't really be sussed from knowing how probable it would be to hooch three reports the same way. Whether they're consistently idiots or consistently deceptive, we have to ask different questions.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:41 AM   #27
bruto
Penultimate Amazing
 
bruto's Avatar
 
Join Date: Jun 2005
Location: Way way north of Diddy Wah Diddy
Posts: 26,581
I would think one problem in sorting this out might be the fact that lying and incompetence are not mutually exclusive - as we have abundantly seen on the national scene.
__________________
I love this world, but not for its answers. (Mary Oliver)

Quand il dit "cuic" le moineau croit tout dire. (When he's tweeted the sparrow thinks he's said it all. (Jules Renard)
bruto is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:46 AM   #28
blutoski
Penultimate Amazing
 
blutoski's Avatar
 
Join Date: Jan 2006
Posts: 12,244
Originally Posted by Dave Rogers View Post
Yes, true. However, the event in this case was not that the State published false stats repeatedly; it was that the State published false stats repeatedly in such a way that they supported the narrative the State wanted them to support better than the real statistics (let's define that as "positive spin").
Originally Posted by Dave Rogers View Post
One might reasonably expect that incompetence would be randomly distributed between positive and negative spin
And this is where I disagree. I don't think that's reasonable. For example, I don't think my math errors while doing my taxes half support and half complicate state policy. I think they're a product of the complexity of the calculation and rules of each specific line item. I miscalculate my pension adjustment pretty consistently because the formula keeps changing and it has a lot of iterative steps. I have never miscalculated my gross income because it's pretty straightforward addition.

I think we need to use the specifics of the activity and the known capabilities of the persons involved as to how often an error falls on either side of something as vague as state policy. Have they published this type of information reliably before, for example. If they've done this just fine a million times before, then a new error consistent with state policy is much more suspicious.


Originally Posted by Dave Rogers View Post
..., so when all the errors observed result in positive spin there's a suggestion that this is deliberate rather than incompetent. However, it's a very small data set, so I wouldn't expect the probability of the hypothesis being true to be compellingly high.
I still think that's the wrong assumption. But this is the challenge I mentioned earlier: we assign baseline probabilities with personal biases based on worldview. It's a problem but at least it is transparent and we can point to what exactly is in dispute.
__________________
"Sometimes it's better to light a flamethrower than curse the darkness." - Terry Pratchett

Last edited by blutoski; 22nd May 2020 at 08:51 AM.
blutoski is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:53 AM   #29
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by blutoski View Post
I guess where I was going is that this is evidence for a botched die, but not necessarily evidence for a specific cause for the wonkiness.

But once wonkiness is established, that can be a premise for an argument about whether it's intentional on the part of the player. Did they deliberately buy a weighted die? Did they find it in the street? Was it given to them by a friend who wanted to frame them for cheating?

I think the question in the thread is about the state's intention, which can't really be sussed from knowing how probable it would be to hooch three reports the same way. Whether they're consistently idiots or consistently deceptive, we have to ask different questions.
I thought the question in the thread was whether the presented event is evidence for the presented hypothesis, to which the answer is yes, and what generally makes an event evidence for a hypothesis. Sure, in this specific case it's also evidence for the state being incompetent, but that doesn't stop it also being evidence for the state lying. It depends on which hypotheses you consider.
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin

Last edited by caveman1917; 22nd May 2020 at 08:56 AM.
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:56 AM   #30
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by blutoski View Post
This is fallacious, and that's OK, because we have to shift to more advanced critical thinking techniques to assign probabilistic outcomes and use those for real world decisionmaking.
This is not really this thread.....but it seems the best answer is to reject the claim it was on purpose and reject the claim that it was the accident.

It seems like something you don't really need to have a position on.
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 08:57 AM   #31
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by caveman1917 View Post
I thought the question in the thread was whether the presented event is evidence for the presented hypothesis, to which the answer is yes, and what generally makes an event evidence for a hypothesis. Sure, in this specific case it's also evidence for the state being incompetent, but that doesn't stop it also being evidence for the state lying. It depends on which hypotheses you consider.
But it is also evidence for dragons altering it with magic before publishing.
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:02 AM   #32
JayUtah
Penultimate Amazing
 
JayUtah's Avatar
 
Join Date: Sep 2011
Posts: 18,675
Originally Posted by BobTheCoward View Post
But it is also evidence for dragons altering it with magic before publishing.
True, but for that hypothesis you could say you start with a very low initial probability, derived from a prima facie evaluation. The information may affect that probability, but it unless it affects it dramatically, that hypothesis remains very unlikely compared with others that start with a greater prima facie probability.
JayUtah is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:03 AM   #33
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by BobTheCoward View Post
But it is also evidence for dragons altering it with magic before publishing.
Yes, that is very true. It depends on the hypotheses you're considering. In this case the choice seems to be "is the state lying or not" and not "are dragons altering publications with magic or not".
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:05 AM   #34
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by JayUtah View Post
True, but for that hypothesis you could say you start with a very low initial probability, derived from a prima facie evaluation. The information may affect that probability, but it unless it affects it dramatically, that hypothesis remains very unlikely compared with others that start with a greater prima facie probability.
Does the event asserted with the proposed explanation come with the prima facie evaluation and I am being uncharitable by saying the event is not evidence for the explanation? Or is it incumbent on the person making the claim to assert that?
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:06 AM   #35
BobTheCoward
Penultimate Amazing
 
Join Date: Nov 2010
Posts: 19,160
Originally Posted by caveman1917 View Post
Yes, that is very true. It depends on the hypotheses you're considering. In this case the choice seems to be "is the state lying or not" and not "are dragons altering publications with magic or not".
I don't think evidence is the right word. This is sounding more like a prerequisite for a claim.
BobTheCoward is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:07 AM   #36
Dave Rogers
Bandaged ice that stampedes inexpensively through a scribbled morning waving necessary ankles
 
Dave Rogers's Avatar
 
Join Date: Jan 2007
Location: Cair Paravel, according to XKCD
Posts: 30,941
Originally Posted by caveman1917 View Post
Yes, that is very true. It depends on the hypotheses you're considering. In this case the choice seems to be "is the state lying or not" and not "are dragons altering publications with magic or not".
And if you're considering three hypotheses, "Is the state incompetent," "Is the state lying," and "The state is both honest and competent," it's primarily evidence against the third.

Dave
__________________
Inspiring discussion of Sharknado is not a good sign for the audience expectations of your new high-concept SF movie sequel.

- Myriad
Dave Rogers is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:25 AM   #37
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by BobTheCoward View Post
I don't think evidence is the right word. This is sounding more like a prerequisite for a claim.
Well a word is a word, it's an arbitrary bunch of squiggles, what matters is the concepts.
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:28 AM   #38
caveman1917
Philosopher
 
Join Date: Feb 2015
Posts: 7,216
Originally Posted by Dave Rogers View Post
And if you're considering three hypotheses, "Is the state incompetent," "Is the state lying," and "The state is both honest and competent," it's primarily evidence against the third.

Dave
Yes exactly.
__________________
"Ideas are also weapons." - Subcomandante Marcos
"We must devastate the avenues where the wealthy live." - Lucy Parsons
"Let us therefore trust the eternal Spirit which destroys and annihilates only because it is the unfathomable and eternal source of all life. The passion for destruction is a creative passion, too!" - Mikhail Bakunin
caveman1917 is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:33 AM   #39
JayUtah
Penultimate Amazing
 
JayUtah's Avatar
 
Join Date: Sep 2011
Posts: 18,675
Originally Posted by BobTheCoward View Post
Does the event asserted with the proposed explanation come with the prima facie evaluation...
Yes.

Quote:
...and I am being uncharitable by saying the event is not evidence for the explanation?
A hypothesis that presumes the existence of something not known to exist can be said to have a low prima facie probability of explaining anything. That's just made worse when there are specific facts to explain. The questionable existence of the cause and its capacity to bring about the specific observation now have to be combined in our prima facie analysis. For that reason we generally dismiss hypotheses summarily that require agency whose existence is not proven. That's the rule of parsimony.

If we say that evidence is information that tends to make a hypothesis more or less likely, what we really want to say is that it makes some hypotheses more or less likely than others. Information becomes evidence only when applied to a hypothesis. But really it becomes evidence in the useful sense only when applied to two or more hypotheses, and only if it serves to distinguish them. If, as you suggest, the rising tide of evidence raises all boats, then it would be hard to consider it evidence in the operative sense.
JayUtah is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Old 22nd May 2020, 09:43 AM   #40
theprestige
Penultimate Amazing
 
Join Date: Aug 2007
Location: Hong Kong
Posts: 45,225
Originally Posted by BobTheCoward View Post
But it is also evidence for dragons altering it with magic before publishing.
The problem here is you're begging the question that magic dragons are a competing claim for anything.
theprestige is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top
Reply

International Skeptics Forum » General Topics » General Skepticism and The Paranormal

Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -7. The time now is 03:16 AM.
Powered by vBulletin. Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum began as part of the James Randi Education Foundation (JREF). However, the forum now exists as
an independent entity with no affiliation with or endorsement by the JREF, including the section in reference to "JREF" topics.

Disclaimer: Messages posted in the Forum are solely the opinion of their authors.