The Skeptics Head Scratchers Reports.

[Guest Stan Norton]

You do not at all have to make that assumption. And for the last time, I am not trying to use anything to prove Bigfoot does not exist. I am saying that you cannot use weight of reports, especially in the numbers that currently exist, to argue that Bigfoot does exist. Can you not see the difference between the two? It is possible for information to not have any bearing on one side of an argument or another, while still being potentially useful information for other purposes.

My point is simply that somebody trotted out the old "Statistically speaking, the likelihood of the majority of reports being hoaxes or misidentifications is quite small. " line, which is not in any way the case.

And the same answer to you. I dismiss these reports in the context of being useful in proving Bigfoot's existence. They may well be useful to those who accept the existence of Bigfoot and wish to study it, but then it is up to such people to sort the wheat from the chaff.

Your bird analogy is flawed in that it is the interpretation of the data that is done by skilled individuals; I am not suggesting that knowledge of Bigfoot makes a report more credible. In a bird survey the starting point is that people are not lying about the birds they see; experience tells us (presumably) that they do so in insignificant numbers so the credibility of the witnesses is not the point at issue.

Well, if you do readily dismiss encounter reports (and without proferring any sustainable reason why) then you really have nothing to bring to the debate besides noises off. As someone who deals daily with the methodologies, results, evaluation and application of ecological surveys my professional and considered opinion is that such reports do have statistical validityalthough clearly we must attempt to standardise the way in which such data are collected and interpreted. Nothing controversial there, just science.

[LeafTalker]

Stan, the marriage of scientific skill and genuine curiosity and openness in you is a gift to this forum. Thank you for that.  

[Guest Stan Norton]

Stan, the marriage of scientific skill and genuine curiosity and openness in you is a gift to this forum. Thank you for that.

Well thanks, much appreciated! I can't see why any other approach is justifiable. This field is crazy enough (on both 'sides') without the (hopefully!) sensible among us failing to state our case. We need level heads!

[LeafTalker]

I actually think it's more than that.... I think we need level heads plus openness. You have that.  

 

(And Trogluddite does, too. And many others here, too. Phew!)

Edited December 29, 2013 by LeafTalker

[gigantor]

As someone who deals daily with the methodologies, results, evaluation and application of ecological surveys my professional and considered opinion is that such reports do have statistical validityalthough clearly we must attempt to standardise the way in which such data are collected and interpreted.

 

Stan,

 

Check out our amateur effort here, maybe you'd be interested in helping us improve things in your spare time. If so, please PM me and we can talk about it.

 

Thanks!

[Guest Llawgoch]

Well, if you do readily dismiss encounter reports (and without proferring any sustainable reason why) then you really have nothing to bring to the debate besides noises off. As someone who deals daily with the methodologies, results, evaluation and application of ecological surveys my professional and considered opinion is that such reports do have statistical validityalthough clearly we must attempt to standardise the way in which such data are collected and interpreted. Nothing controversial there, just science.

 

I've told you the reason, three times.  They are from a self-selecting sample which includes everyone willing to lie about Bigfoot and everyone likely to misidentify things as Bigfoot.  I don't see why this is hard to grasp.

Everyone in America who wants to file hoax Bigfoot reports  is included in the report database.  By definition.

 

People likely to misidentify things as Bigfoot are massively over represented in your sample compared to the general population.

 

Whether or not there are genuine reports in there, how can the above even be up for debate?

Edited December 30, 2013 by Llawgoch

[Trogluddite]

^^^^^

This is really not going to go any further as neither side is apparently communicating to the other side. 

 

You say "They are from a self-selecting sample which includes everyone willing to lie about Bigfoot and everyone likely to misidentify things as Bigfoot."  And they might very well include a self-selecting sample of people who have seen a bigfoot and elect to tell an outside group, but you consistently recognize that source of reports.  

 

I'm not (and I don't think anyone else is) saying because there are 1/100/1,000,000 reports, x% must be true.  Yet you are constantly stating that 10/50/99 % of reports in various databases are hoaxes or misidentifications w/no substantive support for your position other than "I said it, therefore, it must be true."  I have read some encounter reports and, for myself, concluded that they are probably misidentifications (based on limited duration of seeing the target, visibility of the target, poor light, etc.) or flat out frauds - I have read many, many more in which I cannot find rationale explanations that support misidentification.  While I can't rule out a liar, I don't simply assume that the individual giving the report is lying.       

 

And I'm not particularly clear as to why comparing those who report bigfoot encounters with "the general population" offers any insight. Relatively few people in "the general population" encounter murders or a life-threatening accident such as a housefire.  That doesn't stop police, psychologists, and others from studying murders and life-threatening accidents in order to learn what they can from past incidents. 

[Guest Stan Norton]

The self-selecting sample thing is a canard. Clearly the reports come from a non randomised sample of the population ie those who think they have encounteted sasquatch and are willing to say so. The same can be said for any sample you choose: those leaving a polling booth and bother to answer a question, those who see a hot air balloon on a Tuesday and tell someone...there is inhetent bias in any survey and no sample is truly random. My point was that, however biased in the statistical sense, the bigfoot reports database has value. Simple really. And scientific. Isn't that what everyone is after? How else can the wheat and chaff be separated?

[Guest Stan Norton]

Stan,

Check out our amateur effort here, maybe you'd be interested in helping us improve things in your spare time. If so, please PM me and we can talk about it.

Thanks!

Hi gigantor,

Many thanks...you and your colleagues have really put some great effort in so please keep it up!

I hope I haven't given the impression to anyone here that I'm a statistician because I most definitely am not! I curate, design and review ecological surveys for local government so I am dealing daily with field and desk based survey design, methods and interpretation...this does I think give me a useful take on how ecological data are collected, used and written about. Whether those skills enable me to be any more authoritative than others here is debatable, although I do believe I am well placed to comment on matters ecological due to my many years of experience. I am simply trying like so many others here to inject some scientific vigour to the debate.

[Guest Llawgoch]

^^^^^

This is really not going to go any further as neither side is apparently communicating to the other side. 

 

You say "They are from a self-selecting sample which includes everyone willing to lie about Bigfoot and everyone likely to misidentify things as Bigfoot."  And they might very well include a self-selecting sample of people who have seen a bigfoot and elect to tell an outside group, but you consistently recognize that source of reports.  

 

I'm not (and I don't think anyone else is) saying because there are 1/100/1,000,000 reports, x% must be true.  Yet you are constantly stating that 10/50/99 % of reports in various databases are hoaxes or misidentifications w/no substantive support for your position other than "I said it, therefore, it must be true."  I have read some encounter reports and, for myself, concluded that they are probably misidentifications (based on limited duration of seeing the target, visibility of the target, poor light, etc.) or flat out frauds - I have read many, many more in which I cannot find rationale explanations that support misidentification.  While I can't rule out a liar, I don't simply assume that the individual giving the report is lying.       

 

And I'm not particularly clear as to why comparing those who report bigfoot encounters with "the general population" offers any insight. Relatively few people in "the general population" encounter murders or a life-threatening accident such as a housefire.  That doesn't stop police, psychologists, and others from studying murders and life-threatening accidents in order to learn what they can from past incidents. 

 

 

This is exactly what people are saying when they say "statistically it is unlikely all are hoaxes or misidentifications".

 

Also please understand that I am not saying any percentage are hoaxes or misidentifications.  Simply that they COULD be.  This is a distinction that everyone seems very unwilling to recognise.  If a report could be a hoax, it is valueless as evidence of Bigfoot's existence even if it is actually true.  To those who believe Bigfoot exists, it may have tremendous value.

 

The self-selecting sample thing is a canard. Clearly the reports come from a non randomised sample of the population ie those who think they have encounteted sasquatch and are willing to say so. The same can be said for any sample you choose: those leaving a polling booth and bother to answer a question, those who see a hot air balloon on a Tuesday and tell someone...there is inhetent bias in any survey and no sample is truly random. My point was that, however biased in the statistical sense, the bigfoot reports database has value. Simple really. And scientific. Isn't that what everyone is after? How else can the wheat and chaff be separated?

 

There is inherent bias in any survey, but not nearly as much as there is this one.

 

This is a tiny sample compared to the general population and completely self-selected.

 

it's not a canard at all.  It's absolutely crucial to having a proper understanding of the statistical significance of Bigfoot reports.

 

Note that even if there were three reports, or no reports, Bigfoot still might exist.  The point is simply that having reports in the numbers we have have does not lend any weight to EITHER side of the argument for Bigfoot's existence.

Anyway, you can lead a horse to water but can't make it drink.  I have made my point and hopefully a few people understand it and have maybe been given something to think about.

Edited December 30, 2013 by Llawgoch

[Guest Stan Norton]

Ok this one has gone well off topic and it seems as though never the twain shall meet. But...sample size is perhaps the single most important factor in statistical integrity - it assists in assessing data normality. If three people say a bald eagle had a white head can we be sure it does? If 20,000 say so then the statistical significance is fat greater and we can have greater confidence. That is the value of increased sample size. The sasquatch database is large and so we are perfectly entitled to seek patterns in the data and attach statistical significance to them where valid to do so. No one has any evidence whatsoever that this or that proportion of the database is fake...we can only therefore apply standard statistical tests and see what comes out in the wash.

[Trogluddite]

"If a report could be a hoax, it is valueless as evidence of Bigfoot's existence even if it is actually true." 

 

 

PS - I'm prepared to call it a draw and move on .... please understand that I hold you in the highest regard. 

Edited December 31, 2013 by Trogluddite

[Guest Llawgoch]

Please don't assume that because you don't understand something, it doesn't make sense.  

By the way, I'm not sure you know what that word actually means.

[Guest DWA]

Ok this one has gone well off topic and it seems as though never the twain shall meet. But...sample size is perhaps the single most important factor in statistical integrity - it assists in assessing data normality. If three people say a bald eagle had a white head can we be sure it does? If 20,000 say so then the statistical significance is fat greater and we can have greater confidence. That is the value of increased sample size. The sasquatch database is large and so we are perfectly entitled to seek patterns in the data and attach statistical significance to them where valid to do so. No one has any evidence whatsoever that this or that proportion of the database is fake...we can only therefore apply standard statistical tests and see what comes out in the wash.

 

Exactly.  The 'skeptical' unfamiliarity with how a scientist deals with evidence laid bare, in a nutshell.  And why this is silly:

 

"If a report could be a hoax, it is valueless as evidence of Bigfoot's existence even if it is actually true."

 

That basically says:  until proof we can use no evidence, which any scientist on his meds will tell you is silly.  Following inconclusive evidence is how science is done.

[JDL]

The skeptical fallacy coupled with the stipulation that all reports are false by definition because bigfoot cannot, therefore, does not exist.  Classic circular reasoning that goes so far as to invalidate even the formulation of a hypothesis.

 

Worse, their position is based on nothing more than their own belief system.  They can't prove their own negative hypothesis, so they regress to attacking all evidence that does not support their own belief based, negative hypothesis.