Is “good evidence” a subjective or objective determination?

[T.G.G. Moogly]

Objective facts are true even if the overwhelming majority rejects them.

Before one can accept this one has to have decided what is real. I'm reminded of an experience from very early childhood. A police officer was telling us about traffic safety, crossing streets safely, looking in both directions, etc. He said that if you see a car coming toward you and you cover your eyes because you're afraid, the car doesn't go away. That really stuck with me.

 

[steve_bank]

By definition facts are indisputable. A car is going 50 mph.

In general the problem is in proving something is indeed a fact and not supposition or conjrecture. Again a Perry Mason episode.

 

[fromderinside]

Basically I watched the whole original Perry Mason series recently. There is no way a Perry Mason episode relates to establishing facts.

Very twisty. Something is always missing when Perry gets the loser to confess. There is no scene to which one can refer that the person actually did the deed. there is only the word picture Perry draws when accusing.

 

[steve_bank]

Basically I watched the whole original Perry Mason series recently. There is no way a Perry Mason episode relates to establishing facts.

Very twisty. Something is always missing when Perry gets the loser to confess. There is no scene to which one can refer that the person actually did the deed. there is only the word picture Perry draws when accusing.

It surely does. How the application of logic goes goes from tne subjective and circumstantial to objective truth.

Glad to see my old nemesis is still rolling around.

 

[Lion IRC]

By definition facts are indisputable. A car is going 50 mph.

In general the problem is in proving something is indeed a fact and not supposition or conjecture...

I thought the problem was that 50 miles per hour is a disputable 'fact' because space/time is relative and not objectively observable from inside the uni/multiverse.

 

[steve_bank]

By definition facts are indisputable. A car is going 50 mph.

In general the problem is in proving something is indeed a fact and not supposition or conjecture...

I thought the problem was that 50 miles per hour is a disputable 'fact' because space/time is relative and not objectively observable from inside the uni/multiverse.

There are 3 general levels of science. Relativistic where things go very fast and time dilation becomes significant. Newtonian here on the surface which governs things like baseballs and jets, and quantum where thongs are very small like atoms.

All measurers are relative in that they are refernced to a point or standard. If something weighs 1o kg it is relative to the kg standard.. The philosophical bombshell Einstein dropped was that there are no absolutes.

Velocity is relative to an in inertial frame an xyz set of axis in space. A space ship in motion is an inertial frame.


Two cars one at 50 mph and the other at 70 mph have velocities relative to the ground. To the 50 mph car the 70 mph car is moving at 20mph faster to its inertial frame.

People on the ISS are aging differently than those on the ground. Relativistic corrections for time dilation have to made for GP clocks on Seattleites ab\nd the ground.

Thanks for the rep comment. We all sorely need civility.

The global reference points are Stem International or SI. Google SI units NIST and you will find an explanation and list of units.

 

[skepticalbip]

By definition facts are indisputable. A car is going 50 mph.

In general the problem is in proving something is indeed a fact and not supposition or conjecture...

I thought the problem was that 50 miles per hour is a disputable 'fact' because space/time is relative and not objectively observable from inside the uni/multiverse.

If you thought that then you would be wrong. Someone in a different reference frame would still agree that the car was traveling at 50mph in the reference frame in which that measurement was made.

 

[ronburgundy]

"Good evidence" is contextual, but there are logical principles that should be applied.
It is basically a matter of relative probability and how much a piece of data increases the probability of a given theory. That increase in probability is logically determined by whether the theory was able to predict and explain the data with greater accuracy than the alternatives, and whether those alternatives make more or less unverified assumptions than the theory in question.

So, relevant factors are things like how many alternative theories can account for the data? If the alternatives mean the data can be easily accounted for even if theory X is false, then it isn't "good evidence" and in fact might not even qualify as "evidence" at all.
Another factor is the how well supported those alternatives are prior to considering the data in question. If some of the alternatives assume only things that are well established facts, then the data does little to increase the probability of some theory that relies upon yet unverified assumptions. For example, it is well established fact that people lie, they misperceive, they misremember events. Thus, if "evidence" for the theory that cows can jump over the moon consists of some people claiming they saw it happen, that is very weak evidence for the theory b/c there are multiple alternatives to explain people saying that which rely only upon known causal mechanisms. That related to a general principle of evidence based reasoning, which is that an event is more likely to be an instance of something that is already known to generally happen than to be an instance of something that has not been verified to ever happen. If you know there are deer in the area, then tracks that could be deer are more likely to be deer than to be from some hypothetical animal not yet verified to even exist. Same goes for anything that could be do to something in the known universe rather than God.

 

[steve_bank]

By definition facts are indisputable. A car is going 50 mph.

In general the problem is in proving something is indeed a fact and not supposition or conjecture...

I thought the problem was that 50 miles per hour is a disputable 'fact' because space/time is relative and not objectively observable from inside the uni/multiverse.

If you thought that then you would be wrong. Someone in a different reference frame would still agree that the car was traveling at 50mph in the reference frame in which that measurement was made.

A true fact is by definition a fact. Running a red light caught on traffic cam. Eyewitness to a murder.

Newton's Laws Of Motion, within understood limits. From any courtroom drama, a preponderance of subjective evidence leading to a conclusion of guilt when the actual truth is not known.

Good evidence is like good pizza, you know it when you taste it.

Good evidence is that which passes open scrutiny, but that is not always true. 'Truth' and evidence of truth is a complex issue.

 

[Brian63]

Thus, if "evidence" for the theory that cows can jump over the moon consists of some people claiming they saw it happen, that is very weak evidence for the theory b/c there are multiple alternatives to explain people saying that which rely only upon known causal mechanisms.

Is it more appropriate to still consider that "weak evidence" or should we simply dismiss it as being any evidence at all? What still has me confused is whether that is a meaningful and also objective distinction.



https://www.youtube.com/watch?v=gU3uIibaDb0

In rewatching the above link from my call to The Hang Up podcast, Matt Dillahunty makes a remark at about 1:04:00 that evidence becomes sufficient when it narrows to one answer only. If that is the case though, then I do not understand what distinction there is between "evidence" and "proof" for a hypothesis/theory. At the time of the call I had not thought to raise that question, unfortunately. Any other helpers out there? Is he right or wrong, in your view? I would think that some newfound data could elevate the likelihood of multiple hypotheses simultaneously. Maybe just adding the qualifier "sufficient" is what he was emphasizing though.


He immediately follows by listing how "some people" are deluded, irrational, etc. That claim is true, but also seems to defeat his own argument because it applies to himself too. All of us humans have minds that evolved to be biased and we will hold irrational, deluded, biased views throughout our lives on a variety of issues. None of us are going to be purely unbiased in evaluation of data on those matters. Any evaluation we make will be made inside of our own subjective preexisting biases. Good use of the scientific method and peer review would help tremendously in reducing the influence of our biases, but they cannot eliminate them. So the views we hold, and what we think constitute "sufficient evidence" are influenced by our biases. Even when an expert consensus is formed about a scientific view, that is an intersubjective opinion. It is still the result of the personal biased preferences of various individuals, though more carefully processed than gut guesses from random ignoramuses on the street.



Also at about 1:05:20 he describes a scenario where he believes he has an experience that nobody else in the large surrounding group has, and how in that case it would be rational for him to conclude that he is being delusional. I responded by asking if that was actually an argumentum ad populum fallacy. He responded back saying it was not that fallacy because others were merely claiming they did not observe the same phenomenon he did. It would be a fallacy if they take the further step and said it was not real because they did not observe it. However, mere moments earlier Matt did actually form an argument, drawing a conclusion that he was delusional on the basis that others did not observe the same thing. I still do not understand how that is not a fallacy. Similarly, if 99.9999% of living people claimed to have an experience with a god or to have observed a god, and Matt did not have the same, would he also conclude that he was likely delusional then? If it was a unicorn instead of a god, would he think he was the delusional one? If it was a fish in a pond that everybody else saw but him (even when everyone was looking), would he consider himself delusional and to have simply missed observing something that is actually real?



Later, the cohost Phil seems to give advice to not take a strong atheist stance, moreso for tactical reasons. In a debate, it would shift a burden of proof onto me. However, my sincere view is that there are no gods, and I am not going to pretend otherwise. Also, I think there is a strong justification for that view. So no, I am not going to shy away from expressing my honest opinion just because I want to avoid bearing a burden of proof, especially one that I think can be met. I wish he did not warn against taking a strong atheist view on those weak grounds, but at the time I did not think to challenge him on it. Matt, who also believes it is likely there are no gods, agreed with Phil and in doing so did not see how it contradicted his own views too.



Days after that show I recall sending an email to them with these followups, but never received a response.

 

[Lumpenproletariat]

Not all "good evidence" is equal.

Is “good evidence” a subjective or objective determination?

Objective. But that doesn't mean everyone agrees in all cases whether the evidence is good. Perhaps such disagreement could be resolved if enough time or effort is spent trying to resolve it, but that might be impractical so that there'd never be agreement that it is or isn't good evidence. But even so each person's judgment whether it's good evidence is an objective determination, and this could be mistaken but is more reliable than a subjective determination.

Different groups of people can look at the exact same data and infer different conclusions from it. In religion, a Christian can point to nature and sincerely believe it is “good evidence” for God. I can sincerely believe it is not.

That could probably be resolved, with both sides finally agreeing that we don't know, or that it can't be proved, or that maybe "God" is some kind of real possibility which cannot be proved with certainty. It's reasonable as a possibility, but it cannot be proved -- or there's reason to believe it but also doubt it, or it's a reasonable hope, because of the evidence, which is better than if there were no evidence at all, and one can reasonably believe or disbelieve, based on the same limited evidence (believe because there's evidence, but disbelieve because the evidence is limited or not "good" enough).

Is that disagreement because what constitutes sufficient evidence or good evidence is subjective to each individual? Or is there an objective standard that neither of us are able to explicitly state?

The latter, usually. And you could explicitly state the objective standard, and agree, if you'd both argue the question long enough (which is unlikely). To reach this agreement probably both sides would have to modify slightly but end up figuring that their original belief was approximately true anyway, but made better with the modification.

In our current culture there's usually not enough spare time for disputants to stay at it long enough to find a reasonable resolution. And also at some point they decide that it no longer matters anyway.

Different members of a jury can each subjectively agree that the evidence is good or not good enough to warrant convicting the defendant of the crime.

Maybe it's not the evidence they disagree about, but the meaning of "beyond a reasonable doubt" or other standard required to convict. But if it's the evidence -- whether it's "objective" or good evidence -- and they have all the information, then they should finally be able to resolve whether it's good evidence or not. But probably they are not willing to go at it long enough to resolve it. Probably there's a personal animosity preventing them from arguing long enough to resolve it.

The vast majority of legal cases are much more easily resolved than this kind where the jurors cannot agree on the merit of the evidence. Most cases are resolved using "good evidence" and without even going to court.

However, are there ever cases where what is “good evidence” is indeed an objective matter?

Yes. Probably most.

Or is it inherently subjective?

No. Usually objective.

It may be intersubjective. A group of medical experts can come to agreement that a certain drug will be effective on 90% of patients. Whether 90% constitutes “good evidence” that the drug works is a subjective determination still.

Maybe, but that it's better evidence than only 80% is an OBJECTIVE determination (if the 80% or 90% is established).

Is this right?

What's right is that a higher probability is better evidence and makes a stronger case, and stronger does matter. Even if "good evidence" is problematic, drawing an artificial line between "good" and "not good," still "better evidence" is not problematic and is also important in truth-seeking.

We can objectively assess a 51% probability for a certain event to happen, but to take the next step and say that the probability is a “good probability” or “good evidence” is where it becomes subjective.

No, because "good" probably means better enough to decide the issue. But 90% is preferable to 51%. So one could reasonably judge that 51% isn't good enough.

Some legitimate decisions by those responsible could be based on only 51% or even less probability.

You could say there is subjectivity involved, but still it's objectively the case that a higher probability is "better evidence" than a lower probability. And "better" does matter and can decide an important issue -- even the most important issues.

What is an important issue or judgment which cannot be decided based on "better evidence"?

When there's an "impasse" it usually could be resolved. Today's communicating mechanisms to resolve it may not be sufficient yet, but they could be improved, and today's mechanisms are better than those 1000 years ago.

It is seldom that an important issue has to be left unresolved simply because it's "subjective" -- rather, the reason it can't be resolved is that it would take too long, and of course the usual lack of information might leave too much uncertainty. It's usually not because the opposing sides have contradictory "feelings" or "vibes" about it or are "wired" differently.

 

[steve_bank]

It is subjective and contextual as usual, Perhaps weak evidence versus strong evidence.

Killing someone on camera in front of a cop would ne strong evidence.

A cop turns a corned and sees somebody running from a shooting scene. Suspicious but not good evidence of commitingg a crime.

 

[Marvin Edwards]

A fact is objective evidence if it is confirmed by multiple unbiased observers and can be reconfirmed over time by every skeptical observer who comes along later to take a look.

That's probably as close as we subjective humans will ever get to objective knowledge of reality.

Each of us has multiple senses that can be used to check on each other to help insure accuracy. We can see an apple in a bowl of fruit. But when we pick it up it is too light. And when we smell it, it has no apple scent. And when we thump it, it sounds hollow. So, it is objectively a fake apple.

 

[steve_bank]

Multiple unbiased observers? Many people hear god and Jesus speaking to them. Is that objective evidence of god?

It is well established t in law enforcement that multiple witnesses to an event relate it differently. We humans naturally fill in the blanks without thinking. This is demonstrated experimentally.v

Have a group of people read na involbed text missing some details. Ask questions about the text and people will extrapolate and interpret ansers not exlicyt in the text.

Again from Popper, the only thing we can reasonably call objective is an experiment that can be repeated.

Before sciince many thought the sun went around the Earth based on common observation of unbiased observers....

 

[Marvin Edwards]

Multiple unbiased observers? Many people hear god and Jesus speaking to them. Is that objective evidence of god?

It is well established t in law enforcement that multiple witnesses to an event relate it differently. We humans naturally fill in the blanks without thinking. This is demonstrated experimentally.v

Have a group of people read na involbed text missing some details. Ask questions about the text and people will extrapolate and interpret ansers not exlicyt in the text.

Again from Popper, the only thing we can reasonably call objective is an experiment that can be repeated.

Before sciince many thought the sun went around the Earth based on common observation of unbiased observers....

Well, when people independently report the same observation, then that is functionally an experiment that is being repeated.

We might add that the observed phenomenon is subject to multiple explanations, which we would whittle down using Occam's razor, and coming up with additional tests to pin down as best we can what is going on.

Oh, and the Sun does continue to rise in the East and set in the West. That's still a fact that is helpful if we become lost in the woods. But perhaps less helpful when landing a rocket on the Moon.

 

[Swammerdami]

I approach such questions from the standpoint of an analyst applying Bayes' Theorem. If this is too remote from the philosophical topic, please ignore this post!

The questions just asked shriek out for Bayes! Eyewitnesses are unreliable? We want to know HOW unreliable they are, statistically. In fact we might improve that probability estimate by considering age, sex or other facts about the witness. We want to use a scientific experiment? If we lack the wherewithal to study the experiment personally, we'll want to know the probability that some particular scientist is reliable. ("Peer-reviewed" is a bugaboo, BTW. Bad scientists publish in bad journals reviewed by bad peers!)

Eyewitness evidence can be horrid. But sometimes it can be better than "scientific" or "authoritative" evidence. (It would be too much of a digression to give examples, but some are trivial.) So discussions of "objective" vs "subjective" must end up inconclusive. All we have are probability estimates. But those estimates can be hard to come by! I just watched a YouTube by Richard Carrier — a fan of Bayes! — who declared that the odds against Jesus' historicity are 2-to-1 with favorable prior estimates, or 12,000-to-1 against with unfavorable prior estimates.

("Good evidence" for X can be viewed as new facts supporting X which substantially increase the probability that X is true. If I think the probability that ghosts are real is 0.0000000001% but I get evidence which changes my probability estimate to 0.01%, that can be called "good evidence" — the probability is now much larger, though still tiny. But this quibble is probably way off-topic.)

Probability is defined by Ignorance. Here's a simple example. Mary places $1000 on the Pass line at a Las Vegas Craps table and rolls two dice, one red and one green. What is the probability her pass-line bet will win?
(1) Adam sees a Five on the green die but can't see the red die from his vantage. But he knows 2 or 6 will win, and she has no really bad numbers. 60.7% he calculates.
(2) Beverley sees only a One on the red die. A crap roll is likely. 36.5% she calculates as Mary's chance.
(3) Charlie sees both the One and the Five. 45.5%
(4) Dan is too busy staring at Mary's cleavage to look at either die. He goes with the 49.3% a priori odds.
(5) Elaine doesn't watch the dice; she studies people's faces. Mary's eyes have lit up! She's probably got a natural. 80%.
(6) Fran is like Elaine, but noticed that Mary placed a $75 side-bet on Snake-Eyes: that's probably why her eyes lit up. Pass chance? Just 30%.
(7) George knows something: The dice are rigged and the box-man, who hardly ever lets anyone win a $1000 bet, controls the magnet. 10%.
(8) Harry knows what George knows, but also knows something only he and the box-man know: Mary is a favored mistress of mob boss Tony "Mad Dog" Gargotta. If anything the box-man will help Mary win.
(9) And so on.

And they're all right! Probability is not an absolute; it's a calculation based on one's particular information and ignorances.

Oh, and the Sun does continue to rise in the East and set in the West. That's still a fact that is helpful if we become lost in the woods. But perhaps less helpful when landing a rocket on the Moon.

The great mathematician Godfrey Hardy was a betting man, quite familiar with probabilities. He seems to have lost three half-pence on long-shot bets against the Sun:

Feb 6th 1923 ‘The Subwarden bets Prof Hardy his fortune till death to ½ d that the sun will rose tomorrow, 7th Feb 1923.’
...
Nov 24th 1923 ‘Dr Manning bets Professor Hardy 1/12 his fortune till death to 1d that the sun will rise tomorrow’

After the Sun extended its winning streak by another 291 days, its chances went down? Did they think its luck was running out?

 

[steve_bank]

Bayesian techniques can be used to prove anything. It depends on your assumptions. I am familiar with it from my time as a reliability engineer.


There is no substitute fpr hard data.

Multiple peole may experince the same event, not all neccesarily 'see; te same thing.

Multiple peple seeing someone shoot someone is convincing evidence in court. That woud be one extreme/

People witness a shooting and a person runnng away. Some say they saw a black man, others say Latino, othrs say white. Some say tall, some say short. The oposite extreme.

Whter obseved actions are accurate and sufficnet is itself a subjective jusgement.

Demonstrating Newtonian gravity is an objective fact regardless of who observes the expermet.

More than once multiple people assured me something works, when after investigation it cod not possibly have ever worked.

 

[Marvin Edwards]

...
Demonstrating Newtonian gravity is an objective fact regardless of who observes the expermet.
...

A priest, sitting under another apple tree, says, "Thank you God for this apple you have provided".

 

[Swammerdami]

Just to be clear, Bayesian analysis is a matter of simple mathematical fact, like 2+2=4. Sure, "garbage in—garbage out" and 4 will be the wrong answer if the 2 and 2 were the wrong numbers to start with. Bayesian arithmetic is just one step in the complex problem of probability estimation. But it is a key part. Don't blame Bayes' Theorem if your model is flawed.

 

[Marvin Edwards]

Just to be clear, Bayesian analysis is a matter of simple mathematical fact, like 2+2=4. Sure, "garbage in—garbage out" and 4 will be the wrong answer if the 2 and 2 were the wrong numbers to start with. Bayesian arithmetic is just one step in the complex problem of probability estimation. But it is a key part. Don't blame Bayes' Theorem if your model is flawed.

If I understand correctly, Bayesian analysis adds missing factors to get a more realistic probability. For example, if 1% of people get prostate cancer, that would actually be 2% of men.