Top 5 Skeptical Fallacies

[Drewbot]

My thread synopsis:

Dinwar : Science stuff Blah Blah Sciencey.

Emilys Cat: But this science stuff here.

Me: Where is the Bigfoot crap?

 

[John Nowak]

Agreed. Some conclusions really did come out of thin air (see Aristotle's work on physics, and the assumptions therein!). But many--I would say most--have SOME basis in reality. The difference between me and a Geocentrist isn't that I rely on data and they do not; the difference is which data we both use.

I think there's a lot to that, especially in timeframes where the theories are genuinely contesting. I understand Brahe's model of the solar system was better than Copernicus'.

I'm not sure if I'd say the same of a modern geocentrist, however.

 

[Emily's Cat]

My thread synopsis:

Dinwar : Science stuff Blah Blah Sciencey.

Emilys Cat: But this different science stuff here.

Me: Where is the Bigfoot crap?

Minor addition, but yeah, pretty spot on

 

[jt512]

You cannot say "lack of evidence is evidence of lack" as a generality.

Yes you can, if you state it rigorously: If hypothesis H1 predicts that evidence E should have been observed, and if the complementary hypothesis, H0, predicts that the absence of evidence E should have been observed, then observing the absence of evidence E is evidence against hypothesis H1 (and in favor of H0).

That's never what H0 means. You're assuming a perfectly bifurcated solution set... which would need evidence to support your assumption that only two possible outcomes are possible.

In fact, usually we do test exactly two mutually exclusive and exhaustive hypothesis: a hypothesis H0, that states that something doesn't exist, and the complementary hypothesis, H1, that states that something does exist. Examples of such pairs of hypotheses would include:

H0: There is no difference in effect between Drug A and placebo.
H1: There is a difference in effect between Drug A and placebo.

H0: Women are more likely to wear red clothing during the peak fertility portion of their menstrual cycle than at other times.
H1: Women are not more likely to wear red clothing during the peak fertility portion of their menstrual cycle than at other times.

H0: Bigfoot exists.
H1: Bigfoot does not exist.

H0 is always "we can't tell".

Where are you getting this from? H0 is (usually) the hypothesis of no difference, non-existence, no effect, etc. It's the hypothesis with the "no" in it. That's where the term null in "null hypothesis" comes from.

Your specific H1 then is "If we observe E then we can reject the null - we will be able to tell for sure".

That's not H1 or any hypothesis. You're stating the test itself, the criterion, or decision rule, for rejecting H0. H0 and H1 are assertions about the part of the world we want to test. For example, the assertion (H0) that bigfoot exists, or the assertion (H1) that bigfoot does not exist. You are conflating the hypotheses with the decision rules to be used to accept or reject them.

H0 is NEVER a falsifiable hypothesis.

Exactly backwards. H0 is always a falsifiable hypothesis. For example a if a clinical trial found that Drug A worked 10 times as well as a placebo, with a p-value of 10^-6, we might consider that sufficient evidence to have falsified H0 that the drug is no better than placebo. As another example, a single sighting of a claimed bigfoot that was scientifically confirmed to be a previously unknown species would falsify H0 that bigfoot does not exist."

That's why it's a null hypothesis.

I explained above where the term "null hypothesis" came from, and it's not what you claim.

You can reject the null hypothesis if you H1 pans out...

But you said we can never falsify the null hypothesis. H1 "panning out" would falsify H0.

...but you can never accept the null.

Classical statisticians tend to say that you can never accept the null, because they can't collect evidence in favor of the null, only against it. In Bayesian statistics, however, evidence can be collected in favor of the null, and if enough such evidence is collected, the null can be accepted. This is the crux of the argument. If evidence that should have been observed if H0 is false (ie, H1 is true) is found to be lacking, then that is evidence in favor of H0. A consistent lack of such evidence can be reason to accept that H0 is true.

That would require you to have exhausted every possible Hn that could possibly be imagined ever in the entirety of time .

Like I said, most tests in the real world are designed to discriminate between exactly two hypotheses, H0 and H1, which comprise the entire hypothesis space.

In the context of bigfeet, H0 is "bigfoot is indistinguishable from imagination".

At least now you've actually stated a hypothesis, albeit a tortuously worded one. If bigfoot is "indistinguishable from imagination" then bigfoot is imaginary. Thus we can state H0 simply as "bigfoot does not exist.

All of your set of Hns then are cases of "Evidence En exists to show that bigfoot is not imaginary".

Before, you conflated the hypothesis with the decision rule. Now you are conflating the hypothesis with the evidence. Hypotheses are not evidence. Evidence is something we collect to support or refute a hypothesis.

We have a repeated set of Hns that fail to reject the null. It is reasonable to conclude that bigfoot is indistinguishable from imagination, and assume that such a creature does not exist. But you cannot conclude that bigfoot does not exist and simply shut the book on it - because you cannot prove nonexistence of a thing.

You have described the fundamental fallacy of classical skepticism, or as scientific skeptics appropriately enough call it, "bigfoot skepticism." This fundamental fallacy is the failure to recognize that we can collect enough evidence against a purported phenomenon that the probability that the phenomenon could still be true is so infinitesimal that we have for all practical purposes shown it to be false. Classical skeptics like you say, "sorry, you can't prove the null," and so classical skepticism goes on forever investigating the same old tired phenomena—bigfoot, psi, dowsing, UFOs, etc. Most purported phenomena that draw the interest of classical skeptics had little to no plausibility in the first place, and every investigation has failed to find supporting evidence (or has found disconfirming evidence). At this time, the posterior probability of any of these purported phenomena is infinitesimal. Yet, because classical skeptics assert "you can never disprove a thing," they go on beating dead horse after dead horse.

We tried many different approaches to find evidence of the Higgs. We looked quite hard, we simply didn't have any luck with those approaches. That did NOT lead us to conclude that the Higgs did not exist. All it led to is that we could not tell whether it existed or not.

That sounds like something you just made up. My understanding was physicists have always understood that there would be no way to even look for the Higgs boson, until a collider with sufficient energy to detect it could be built. The Americans were going to build one a couple decades ago, but the anti-science party killed the project, and so it was not until CERN completed their Large Hadron Collider that efforts could be undertaken to detect the Higgs boson.

 

[Emily's Cat]

In fact, usually we do test exactly two mutually exclusive and exhaustive hypothesis: a hypothesis H0, that states that something doesn't exist, and the complementary hypothesis, H1, that states that something does exist. Examples of such pairs of hypotheses would include:

H0: There is no difference in effect between Drug A and placebo.
H1: There is a difference in effect between Drug A and placebo.

H0: Women are more likely to wear red clothing during the peak fertility portion of their menstrual cycle than at other times.
H1: Women are not more likely to wear red clothing during the peak fertility portion of their menstrual cycle than at other times.

H0: Bigfoot exists.
H1: Bigfoot does not exist.



Where are you getting this crap from? H0 is (usually) the hypothesis of no difference, non-existence, no effect, etc. It's the hypothesis with the "no" in it. That's where the term null in "null hypothesis" comes from.

Just for curiosity... are you just not paying attention to which element you're assigning to null above?

You can't prove that something doesn't exist - it's unfalsifiable. The null is that big foot does not exist. The null is unfalsifiable.

You test whether you can distinguish between the unfalsifiable null hypothesis and the falsifiable alternative hypothesis. If you can distinguish between them, then you can reject the null. You can show that the null hypothesis is errant.

If you can collect enough evidence through enough experimentation, you can conclude that the null is very highly likely to be correct - there's your bayesian conclusion. You can then deduce that the null is true. But you can't actually accept the null hypothesis through direct testing, nor through formal logic.

It's mostly semantics, but to me it's important semantics.

 

[jt512]

Just for curiosity... are you just not paying attention to which element you're assigning to null above?

Apparently not. I inadvertently switched H0 and H1 in one example. As I said, H0 is generally the hypothesis with "not" or "no" in it.

You can't prove that something doesn't exist - it's unfalsifiable. The null is that big foot does not exist. The null is unfalsifiable.

So, in addition to not knowing what a hypothesis is, you also don't know what "unfalsifiable" means; you are confusing "unfalsifiable" with "unprovable," which means exactly the opposite.

You test whether you can distinguish between the unfalsifiable null hypothesis and the falsifiable alternative hypothesis. If you can distinguish between them, then you can reject the null. You can show that the null hypothesis is errant.

In other words, you can falsify the null: you can show that it is false. You cannot (in a certain sense) falsify the alternative hypothesis. Falsifying the alternative would be affirming the null, which would be "proving" non-existence, the thing you claim cannot be done.

If you can collect enough evidence through enough experimentation, you can conclude that the null is very highly likely to be correct - there's your bayesian conclusion. You can then deduce that the null is true.

Good. Stop right there.

It's mostly semantics, but to me it's important semantics.

If it's just a matter of semantics, then why do skeptics keep investigating nonsense such as paranormal claims, supernatural claims, claims about UFO sightings, claims that people can detect underground water with a stick, claims that people can bend spoons with their minds, or can talk to dead relatives or see ghosts—etc, etc, ad nauseum? I think skeptics keep an "open mind" about these psuedo-phenomena precisely because they believe that "you can't prove non-existence," and therefore every new idiotic claim about any worn-out nonsensical hypothesis has to be dealt with on its own merits.

 

[Emily's Cat]

Apparently not. I inadvertently switched H0 and H1 in one example. As I said, H0 is generally the hypothesis with "not" or "no" in it.

[pedant] You reversed it in two of your three examples [/pedant]... But I know what you mean so it's not really a big deal.

So, in addition to not knowing what a hypothesis is, you also don't know what "unfalsifiable" means; you are confusing "unfalsifiable" with "unprovable," which means exactly the opposite.

In other words, you can falsify the null: you can show that it is false. You cannot (in a certain sense) falsify the alternative hypothesis. Falsifying the alternative would be affirming the null, which would be "proving" non-existence, the thing you claim cannot be done.

I know what it means, I'm just using it completely wrong. You are correct - the null is unprovable, not unfalsifiable. I apologize for that.

Good. Stop right there.

[pedant] deduce <> prove [/quote]

That's kind of my point - you can deduce that it's extremely unlikely to be true... which is not the same thing as proving that it's false. Saying that big foot is very, very, extraordinarily likely to not exist is NOT the same as proving that big foot doesn't exist.

Deductive reasoning is about likelihoods. Formal Logic is about proof.

It's extraordinarily unlikely that I will win a $500M lottery at noon on my next birthday. That doesn't mean it's impossible. It does, however, mean that I shouldn't plan on it .

If it's just a matter of semantics, then why do skeptics keep investigating nonsense such as paranormal claims, supernatural claims, claims about UFO sightings, claims that people can detect underground water with a stick, claims that people can bend spoons with their minds, or can talk to dead relatives or see ghosts—etc, etc, ad nauseum? I think skeptics keep an "open mind" about these psuedo-phenomena precisely because they believe that "you can't prove non-existence," and therefore every new idiotic claim about any worn-out nonsensical hypothesis has to be dealt with on its own merits.

I don't think it's because they believe you can't prove non-existence. I don't think it's anywhere near as cogent as that. Some people need to believe in something. They need magic, wonder, a promise of more than just what their senses perceive. They need an answer to "why?", and they're not comfortable with null as the answer. They've been brought up in a culture where god is a thing, and they've been brought up to believe in miracles, magic, angels, demons, etc. Is it little wonder that they fill the gaps with other magical answers?

In my opinion, people cling to bigfoot and fairies and ghosts for the same reason that they cling to religion. It fills the chasm where "I don't know" and "there's no reason" would otherwise terrify them. They're a tiny little creature on a tiny little planet circling a tiny little sun in a tiny little galaxy, cast adrift in the vastness of space. They need a rudder so they don't feel lost, so they feel that there's a direction, a path, a purpose to their existence.

All of the rest, all of the argument, is simply defense of that belief. Sure, they may be tilting at windmills... but it's worth it if they don't feel powerless in the face of imagined dragons.

 

[Garrette]

If it's just a matter of semantics, then why do skeptics keep investigating nonsense such as paranormal claims, supernatural claims, claims about UFO sightings, claims that people can detect underground water with a stick, claims that people can bend spoons with their minds, or can talk to dead relatives or see ghosts—etc, etc, ad nauseum? I think skeptics keep an "open mind" about these psuedo-phenomena precisely because they believe that "you can't prove non-existence," and therefore every new idiotic claim about any worn-out nonsensical hypothesis has to be dealt with on its own merits.

I don't think it's because they believe you can't prove non-existence. I don't think it's anywhere near as cogent as that. Some people need to believe in something. They need magic, wonder, a promise of more than just what their senses perceive. They need an answer to "why?", and they're not comfortable with null as the answer. They've been brought up in a culture where god is a thing, and they've been brought up to believe in miracles, magic, angels, demons, etc. Is it little wonder that they fill the gaps with other magical answers?

In my opinion, people cling to bigfoot and fairies and ghosts for the same reason that they cling to religion. It fills the chasm where "I don't know" and "there's no reason" would otherwise terrify them. They're a tiny little creature on a tiny little planet circling a tiny little sun in a tiny little galaxy, cast adrift in the vastness of space. They need a rudder so they don't feel lost, so they feel that there's a direction, a path, a purpose to their existence.

All of the rest, all of the argument, is simply defense of that belief. Sure, they may be tilting at windmills... but it's worth it if they don't feel powerless in the face of imagined dragons.

Emily’s Cat, I think you’re answering the wrong question: He asked “why do skeptics keep investigating nonsense…” while your answer addressed why believers continue to believe in the face of evidence to the contrary.

The answer to that, I think, is that most skeptics are not investigating such things in order to see if they are true after all – and in fact, most skeptics here at the ISF are not doing any investigating at all, at least not to a great extent.

Rather, most skeptics either investigate, or engage in discussions about, for a variety of reasons mostly unrelated to an expectation of or hope for proof. The reasons are neither consistent between skeptics nor necessarily consistent within one skeptic’s set of motivations.

I engage in such things for a variety of reasons depending on my mood. Sometimes it’s an appealing way to pass the time. Other times I feel like I can learn something. Other times I feel like I can impart some learning, Other times I am concerned about lurkers. Other times I want to enjoy the witty repartee (not being facetious here).

And there are undoubtedly other reasons. But I imagine that the set of those skeptics who investigate these claims and are expecting or hoping for proof is a very small set indeed.

 

[Emily's Cat]

Garrette, you're right. I answered a question that wasn't asked! I really should stop trying to multitask

 

[Garrette]

Garrette, you're right. I answered a question that wasn't asked! I really should stop trying to multitask

Woo hoo! I was right! Finally!

This day shall receive two stars in my diary.

 

[jt512]

[pedant] You reversed it in two of your three examples [/pedant]... But I know what you mean so it's not really a big deal.

Guilty with an excuse: I had just returned from a 6-mile run in the mountains with a 3300-foot elevation gain and a maximum elevation of 13,200 feet. I'm amazed I didn't get it wrong all three times. (And my excuse today, should I need it, is that I'm recovering from yesterday.)

[pedant] deduce <> prove

Actually, "deduce" means to reach a conclusion by deductive reasoning, which is the type of reasoning you are using when you say that you can't "prove" non-existence. So, the way you used "prove" was synonymous with "deduce." On the other hand, when you were drawing a conclusion from evidence (or the systematic lack of it), that is, by inductive reasoning, you should have you used the word "infer" rather than "deduce."

That's kind of my point - you can deduce that it's extremely unlikely to be true... which is not the same thing as proving that it's false. Saying that big foot is very, very, extraordinarily likely to not exist is NOT the same as proving that big foot doesn't exist.

But one can often establish the truth of a hypothesis up to such a high probability that it becomes ridiculous, not to mention cumbersome, to say "very, very, extraordinarily likely..." every time you refer to the thing. When the probability of a hypothesis about the world is that high, then the hypothesis becomes an empirical fact. Evolution by natural selection is a fact, even though there is a tiny probability that all the evidence for it is mistaken. I'm (thankfully) no expert, but I would say that the non-existence of bigfoot is such a fact as well.

Deductive reasoning is about likelihoods. Formal Logic is about proof.

Again, deductive reasoning is about formal logic, or "proof." Inductive reasoning is about drawing conclusions from evidence, or "likelihoods" as you put it.

I don't think it's because they believe you can't prove non-existence. I don't think it's anywhere near as cogent as that. Some people need to believe in something. They need magic, wonder, a promise of more than just what their senses perceive. They need an answer to "why?", and they're not comfortable with null as the answer. They've been brought up in a culture where god is a thing, and they've been brought up to believe in miracles, magic, angels, demons, etc. Is it little wonder that they fill the gaps with other magical answers?

In my opinion, people cling to bigfoot and fairies and ghosts for the same reason that they cling to religion. It fills the chasm where "I don't know" and "there's no reason" would otherwise terrify them. They're a tiny little creature on a tiny little planet circling a tiny little sun in a tiny little galaxy, cast adrift in the vastness of space. They need a rudder so they don't feel lost, so they feel that there's a direction, a path, a purpose to their existence.

I think that's very well put. But my question is why do skeptics bother investigating reports of such phenomena when science informs us that none of it is true. I think, based on the arguments that I've head skeptics make, that part of the reason is that some skeptics really take to heart that you can't "prove" non-existence. Therefore, in their minds, new reports of bigfoot or paranormality should be taken seriously.

 

[jt512]

But I imagine that the set of those skeptics who investigate these claims and are expecting or hoping for proof is a very small set indeed.

I agree, but I think there is a substantial set of skeptics who think such investigations are logically necessary because "you can't prove non-existence."

 

[abaddon]

Woo hoo! I was right! Finally!

This day shall receive two stars in my diary.

Are not so called "diaries" a myth? I have never seen one.

 

[Garrette]

Are not so called "diaries" a myth? I have never seen one.

Then where shall I put my stars?

 

[Garrette]

I agree, but I think there is a substantial set of skeptics who think such investigations are logically necessary because "you can't prove non-existence."

I think any disagreement we have is on the size of that set. I would not call it substantial, but I have only my personal interactions on which to base that.

 

[Emily's Cat]

Guilty with an excuse: I had just returned from a 6-mile run in the mountains with a 3300-foot elevation gain and a maximum elevation of 13,200 feet. I'm amazed I didn't get it wrong all three times. (And my excuse today, should I need it, is that I'm recovering from yesterday.)

Actually, "deduce" means to reach a conclusion by deductive reasoning, which is the type of reasoning you are using when you say that you can't "prove" non-existence. So, the way you used "prove" was synonymous with "deduce." On the other hand, when you were drawing a conclusion from evidence (or the systematic lack of it), that is, by inductive reasoning, you should have you used the word "infer" rather than "deduce."

...

Again, deductive reasoning is about formal logic, or "proof." Inductive reasoning is about drawing conclusions from evidence, or "likelihoods" as you put it.

Mental note: INDUCTIVE reasoning. Thank you.

But one can often establish the truth of a hypothesis up to such a high probability that it becomes ridiculous, not to mention cumbersome, to say "very, very, extraordinarily likely..." every time you refer to the thing. When the probability of a hypothesis about the world is that high, then the hypothesis becomes an empirical fact. Evolution by natural selection is a fact, even though there is a tiny probability that all the evidence for it is mistaken. I'm (thankfully) no expert, but I would say that the non-existence of bigfoot is such a fact as well.

Hey, you get no argument from me that the probability is infinitesimally small. It's sufficiently small that it's perfectly reasonable to treat it as a hard fact.

{arms crossed} But that doesn't make it impossible. Yes, I'm quite pedantic on this... but it's the difference between vanishingly small and actual zero. It's possible for me to win a $500M lottery at exactly noon on my next birthday. Not likely, in fact so unlikely as to be laughable if I were trying to get a loan against that future amount... but still technically possible.

I think that's very well put. But my question is why do skeptics bother investigating reports of such phenomena when science informs us that none of it is true. I think, based on the arguments that I've head skeptics make, that part of the reason is that some skeptics really take to heart that you can't "prove" non-existence. Therefore, in their minds, new reports of bigfoot or paranormality should be taken seriously.

I think it's twofold (now that I am addressing the right question). On the one hand, I think that some skeptics investigate those claims simply so that they can tell true believers how wrong they are yet again. A bit of a Nelsonesque approach of pointing and laughing.

On the other hand, I think some people really do take it to heart. New information is exactly that - new information. I mean, they did find that prehistoric fish thing that everyone thought was extinct. Weird things happen. I don't think that there's necessarily any expectation that *this time* the evidence will be compelling, so much as a belief that it is their duty to consider each piece of information on its own merits.

 

[Donn]

Interested in analogies and metaphors to explain why:

If it's possible for you to win the lottery, why do you not seriously make plans on it?

 

[John Nowak]

If it's possible for you to win the lottery, why do you not seriously make plans on it?

The chances of winning are so low that there's not much point in planning.

The only rational argument in favor of buying a lottery ticket is when the investment is negligible to you.

 

[Donn]

The chances of winning are so low that there's not much point in planning.

The only rational argument in favor of buying a lottery ticket is when the investment is negligible to you.

I get that, but I don't feel it. Do you have any metaphor, or analogy, to give it punch?

It's the same situation with Ghostfoot: the chances he floats out there are so low, it's crazy to seek him.

Why is it actually crazy? Many seem not to think so. They think that tiny possibility is so large they spend an entire life therein.

 

[Garrette]

The chances of winning are so low that there's not much point in planning.

The only rational argument in favor of buying a lottery ticket is when the investment is negligible to you.

Precisely, which is how I do it. For 1 dollar I can get several days worth of day dreams. For $30 I could get two hours of a mediocre movie and popcorn. Very few argue that the movie is a tax on the math-ignorant.

I get that, but I don't feel it. Do you have any metaphor, or analogy, to give it punch?

It's the same situation with Ghostfoot: the chances he floats out there are so low, it's crazy to seek him.

Why is it actually crazy? Many seem not to think so. They think that tiny possibility is so large they spend an entire life therein.

Besides everything that has been discussed in this many threads? And besides those like our participants here who are in it not for Bigfoot but for the manipulative thrill? I'll put on my armchair psychologist hat:

There are two main sets of people outside the liars and hoaxers. First are the true believers who have -- for various reasons -- never been exposed to actual evaluation of the purported evidence or never learned how to properly evaluate such. I'll set them aside.

The second set is the fun seeker. I didn't really believe in the Wendigo in Boy Scouts but when -- during my initial camp outing with my troop I was the target of the stories meant to make the night spooky and me enjoyably nervous, I played along. I don't believe in all the monsters and ghosties on Supernatural, but it is one of my favorite shows. Bigfootery for some is a willful suspension of disbelief in an attempt to become part of their own real life Supernatural. A very rough analogy may be drawn with the SCA in which no one really believes they are peasants or nobles or mercenaries or what have you; they are simply more open about the make believe aspect.