It looks like you're using an Ad Blocker.

Please white-list or disable AboveTopSecret.com in your ad-blocking tool.

Thank you.

 

Some features of ATS will be disabled while you continue to use an ad-blocker.

 

The Extraterrestrial Hypothesis and the null hypothesis

page: 30
8
<< 27  28  29    31  32  33 >>

log in

join
share:

posted on Apr, 19 2014 @ 12:47 AM
link   
Just to show it is the single observation that wasn't supposed to be there that makes the significance happen, I decided to create another example. This one is identical to the other for the (NN), (NY), (YN) cases but has 0 instead of 1 in the (YY) category.

You can see that neither the one-tailed nor the two-tailed alternatives are significant. Both say what was observed, (0 for (YY), is in complete agreement with that which would be expected under the null hypothesis of stalls being independent of lights overhead.

The reason it is significant in the other example is that the single (YY) observation was not supposed to happen if lights overhead is independent of stalls. We know this is the case because we had 2 out of a billion observations with lights overhead and 2 out of a billion observations with stalls. So the combination of a stall and lights overhead should show up in somewhere around 4 times per billion-billion observations. It was basically impossible to have one (YY) be seen in just a billion observations.

That's pretty easy to see if you understand how the analysis works. All analyses are done using the null hypothesis and all analyses use ALL the data. Anything else amounts to handwaiving and self-delusion. We've known this to be true for over 100 years now although good scientists have operated in this manner (their baseline is the null hypothesis) for far longer than the last 100 years. Only when science understands normal can we identify what is novel.



I know that focusing on the null isn't obvious to most people -- everyone wants to focus on the "interesting" cases. It's the wrong thing to do. This is partly why, if you think you see a cluster of cancers in an area, that the cancer cluster probably is normal and not something to get worried about. For things like cancer, we do routinely collect enough data in the US and in Europe to know what normal (the null hypothesis) looks like. Because of that, we know what a suspicious "cluster" looks like.

For UFOs, we don't have the data to tell us what the null hypothesis looks like. Scientists who are knowledgeable in data analysis do know to dismiss the focus on the "interesting" cases. But no one has taken the next step -- determining what the null hypothesis looks like -- at least publicly. That would be a huge effort because UFOs are more rare than cancer. And it would be incredibly expensive to understand the null in the general sense. If there are real aliens visiting, the cheaper thing to do is to wait for the evidence of aliens to "fall in your lap." Basically, go collect the proof at a crash site.

Some people believe we have. Some don't. The government is denying, but there is reason to believe they would deny in either case.
edit on 293am14America/Chicago58001kAmerica/Chicago by BayesLike because: (no reason given)




posted on Apr, 19 2014 @ 01:01 AM
link   
a reply to: Phage

First, another non scientific statement. You said:


Creating a null hypothesis does not make a hypothesis falsifiable. In order to be falsifiable it must be possible to prove the null. "No UFOs are controlled by extraterrestrials" cannot be proven.


Why do you need to prove the null hypothesis? In Science you falsify the null by showing correlation between 2 measurement. The null is assumed to be true. Here's more:


"Remember we can never prove the null hypothesis. All we can prove is that there is a relationship or effect (H1) between two or more variables."


Here's more:


The danger in thinking that null hypotheses are proven wrong if rejected is twofold. Both of these dangers are related to the word prove. Although this word has different meanings in different contexts (e.g., mathematics, printing, and cooking), most dictionaries indicate that we prove something when we establish its genuineness or authenticity. Proof, therefore, leaves no room for error. If you prove something, you and others can be 100% confident that your claim is true.


www.statisticalmisconceptions.com...

So you don't prove the null. That's just silly. Let's say you have this null:

"The drug doesn't help prevent heart attacks"

As a Scientist, you don't prove the drug doesn't help prevent heart attacks. You can't prove the null, you assume the null is true and falsify it by showing correlation between the drug and heart attack prevention.

Now to your example:


The hypothesis: "God made the Universe."
The null: "God did not make the Universe."
The hypothesis is not falsifiable because it cannot be proven that God did not make the Universe.


First, you don't have to prove the null.

A person can also falsify the null in this situation and there's been a lot of books written on this subject. I'm not trying to debate this issue here but let me play devil's advocate.

Right now a person making this argument can say that science has shown that God created the universe. Right now science needs parallel universes or some sort of multiverse to try and get around this. Right now, the initial conditions are right for life to exist and without things like the 10/500 vacua you see in String Theory, you have science showing a miracle occurred. A person has just shown correlation between science and the universe being created by a miracle.



posted on Apr, 19 2014 @ 01:15 AM
link   
a reply to: BayesLike

Before I respond and blow your post out of the water, what do you mean when you say this:


Not only have you and Neo failed to recognize that you have tossed out 99.9999999.... % of your data


Exactly what data are you talking about. Point me to this data.

And:


You compare to the expectation under the null hypothesis which means ALL the data has to be used.


Again, what data are you talking about when you say all the data. All what data?

You also said:


If there were only a few (NN) cases, we would fail to reject the null hypothesis.


When you say fail to reject the null hypothesis what do you mean. I haven't debated a debunker yet that understood the null.

So tell me, what data haven't I included? If you say 99.99999% of the data you must have a lot of data to show me, so let's see it.
edit on 19-4-2014 by neoholographic because: (no reason given)



posted on Apr, 19 2014 @ 01:17 AM
link   
a reply to: Phage

Ok you might have a point. But please then consider this. What if we instead turn it around and declare the ETH hypothesis as 'No ufocases are of ET source' then the null hypothesis would be 'there exist an ufocase that is of ET source'. We would then only have to prove one ufocase being ET, to refute the alternate hypothesis being no ufocases are ET source. Would that not indeed be a falsifiable hypothesis?



posted on Apr, 19 2014 @ 02:25 AM
link   

originally posted by: neoholographic
Before I respond and blow your post out of the water, what do you mean when you say this: ...........


Neo, just read the two examples. You have some very strange ideas about how hypotheses work and how to "prove" them. I assume you have the capacity to understand but for some reason prefer to remain blind and barefoot. I don't know if you came up with this pseudo knowledge all by yourself or you picked it up in a few popular (meaning non-academic) texts written by some whacko who doesn't have a clue but sounded convincing. However you arrived at the point at which you are today, your basic concepts are deeply flawed. Very deeply flawed.

So, if you want to "blow" my "post out of the water," don't bother with flawed rhetoric and hand waiving of deeply flawed concepts -- go do it with statistical software. I posted two examples with data, why don't you try to analyze the same data with your methods using real statistical software and show the rest of ATS how it is done. You get to work with data in just one cell, the (YY) cell. Collaborate with enPas if you want, you will need help I think!

Of course, you won't find any statistical software which analyzes data the way you think data should be analyzed. Those methods only exist in Neo's head -- the rest of the universe just hasn't caught up to your way of doing statistics yet. Something tells me the rest of the universe never will.

(enPas -- what a great name for the side you are on! Brilliant in fact! If you are actually a chess player, I'd suggest sticking to chess. If you aren't a master yet, I'd suggest learning 20 major openings and all defense variations to the appropriate depth for each.)



posted on Apr, 19 2014 @ 03:23 AM
link   
a reply to: BayesLike

You just don't get it. You cannot reduce this hypothesis to mere statistical analysis. These sighting are, by and large, singular events that by their very nature automatically exclude the mundane. Your constant adherence to statistics implies that the witnesses are mistaken when they say the events are of a kind that cannot be lumped in with mundane unknowns.

The issue here is that the evidence is so compelling that it renders statistical analysis redundant because it is by comparison so weak. If a person witnesses a crime involving a truck one does not compare the activities of the truck to everything trucks were doing that day and every other day and then say that there was an infinitesimal chance the truck was involved in the crime. The reason being that the SURROUNDING EVIDENCE renders statistical evidence irrelevant. Surrounding evidence in this case would be comprised of such things as some of the stolen goods being found in the abandoned truck, tyre marks leading up to the scene of the crime, witness testimony etc etc. The evidence completely overshadows any doubt that could arise from abstract argument about what COULD be the case. Essentially you are using the "It could be swamp gas" argument.

In the end you would be reduced to saying it could be anything and to do this would be to ignore the evidence.
edit on 19-4-2014 by EnPassant because: (no reason given)

edit on 19-4-2014 by EnPassant because: (no reason given)



posted on Apr, 19 2014 @ 03:56 AM
link   

originally posted by: EnPassant
The reason being that the SURROUNDING EVIDENCE renders statistical evidence irrelevant. Surrounding evidence in this case would be comprised of such things as some of the stolen goods being found in the abandoned truck, tyre marks leading up to the scene of the crime, witness testimony etc etc. The evidence completely overshadows any doubt that could arise from abstract argument about what COULD be the case. Essentially you are using the "It could be swamp gas" argument.


enPas, it's obvious you don't understand still and you can't address the example either. The reason we can dispense with a formal statistical analysis when stolen goods are found is the null hypothesis. Finding stolen goods is highly significant when the null is that none exist on the truck in question. Likewise with the other evidence.

This however is not what UFO evidence is like. Lights are seen overhead, cars stall, etc with and without claims of UFOs. We know trucks exist and so forth. We do not know UFOs exist as an object nor do we know aliens exit on Earth (or anywhere else). The two ideas are very different.



posted on Apr, 19 2014 @ 05:46 AM
link   
a reply to: neoholographic


So tell me, what data haven't I included?

What "data" do you think you included? I am being serious. How do you statistically analyze youtube videos and links to articles about stuff like crop circles? The only thing you are showing is that there are youtube videos and links to crop circles.



posted on Apr, 19 2014 @ 05:58 AM
link   
a reply to: EnPassant


You just don't get it. You cannot reduce this hypothesis to mere statistical analysis.


Null Hypothesis (wiki)

In statistical inference of observed data of a scientific experiment, the null hypothesis refers to...


Statistical inference

In statistics, statistical inference is the process of drawing conclusions from data that are subject to random variation, for example, observational errors or sampling variation.[1] More substantially, the terms statistical inference, statistical induction and inferential statistics are used to describe systems of procedures that can be used to draw conclusions from datasets arising from systems affected by random variation,[2] such as observational errors, random sampling, or random experimentation.[1] Initial requirements of such a system of procedures for inference and induction are that the system should produce reasonable answers when applied to well-defined situations and that it should be general enough to be applied across a range of situations. Inferential statistics are used to test hypotheses and make estimations using sample data.


There are exactly ZERO examples of doing what you are thinking in the real world. There is no exception for UFO cases because they are way cool.

edit on 19-4-2014 by ZetaRediculian because: (no reason given)



posted on Apr, 19 2014 @ 06:25 AM
link   
a reply to: Nixnada

We would then only have to prove one ufocase being ET


That works.



posted on Apr, 19 2014 @ 10:02 AM
link   
a reply to: neoholographic

I think I get it now.

X = ? "No X is X".

No U.F.O.'s are controlled by Extraterrestrials

is understood by most as :
None of the "things we cant identify"(UFOs) are due to "things we don't know exist"(ET).
OR
unidentified things are not due to unidentified things.

equally true but just as nonsensical:
unidentified things are due to unidentified things.

You translate as:
None of the "things we cant identify"(UFOs) are "IDENTIFIED as aliens"
OR
unidentified things are not due to identified things.

"Identified things" does not include aliens but you seem to think so because of your links. You are ASSerting the existence of aliens.

You are "proving" unidentified things are due to unknown things by asserting that unidentified things are identified.

So why do you need a hypothesis? Just make your assertion and post links to things you think might have proof of aliens buried in it. That is all you are doing anyway.










posted on Apr, 19 2014 @ 10:10 AM
link   

originally posted by: BayesLikeenPas, it's obvious you don't understand still and you can't address the example either. The reason we can dispense with a formal statistical analysis when stolen goods are found is the null hypothesis. Finding stolen goods is highly significant when the null is that none exist on the truck in question. Likewise with the other evidence.


That is my point, ETH has its equivalent of 'stolen goods' - evidence. That automatically excludes most other 'trucks'.


Not only have you and Neo failed to recognize that you have tossed out 99.9999999.... % of your data


Previously people were arguing that 99.99999...% would still not nullify the alternative hypothesis. Now you are saying it is relevant...

That fact is there is not any significant body of evidence when it comes to
1. Lights overhead.
2. Stalling engine
3. Encounter with beings
4. Headaches afterwards
5. Descriptions of the beings
6. Multiple witnesses
7. etc

If you were to find mundane examples that satisfy all the above, your sample would be so small as to be useless in nullifying, from a statistical point of view, the alternative hypothesis.
3.
edit on 19-4-2014 by EnPassant because: (no reason given)



posted on Apr, 19 2014 @ 10:15 AM
link   

originally posted by: ZetaRediculian
You are "proving" unidentified things are due to unknown things by asserting that unidentified things are identified.


Ergo: unidentified things cannot be identified. Nonsense.

Pulsars have been identified...



posted on Apr, 19 2014 @ 10:23 AM
link   
a reply to: BayesLike

You didn't answer any of my questions. You just bloviated about nothing. Tell me what data didn't I use? We're dealing with an aerial observed phenomena, so tell me what data didn't I use? Here's the questions again.

Exactly what data are you talking about. Point me to this data.

Again, what data are you talking about when you say all the data. All what data?

When you say fail to reject the null hypothesis what do you mean. I haven't debated a debunker yet that understood the null.

So tell me, what data haven't I included? If you say 99.99999% of the data you must have a lot of data to show me, so let's see it.

All you gave me was a bunch of nonsense.


Neo, just read the two examples. You have some very strange ideas about how hypotheses work and how to "prove" them. I assume you have the capacity to understand but for some reason prefer to remain blind and barefoot. I don't know if you came up with this pseudo knowledge all by yourself or you picked it up in a few popular (meaning non-academic) texts written by some whacko who doesn't have a clue but sounded convincing. However you arrived at the point at which you are today, your basic concepts are deeply flawed. Very deeply flawed.


This tell me you can't answer the question concerning the claims you made.



posted on Apr, 19 2014 @ 10:37 AM
link   

originally posted by: EnPassant

originally posted by: ZetaRediculian
You are "proving" unidentified things are due to unknown things by asserting that unidentified things are identified.


Ergo: unidentified things cannot be identified. Nonsense.

Pulsars have been identified...


Exactly, again people are even better observing unknown things because they're more alert when they see something. This is why scientist have to use eyewitnesses to help better understand how aerial phenomena behave.

So an eyewitness isn't trying to identify something that's unknown, there just telling you how something behaves and the data will begin to identify the aerial observed phenomena.

When dealing with an aerial observed phenomena it's important for researchers to know how the objects behaves. This is different from testing a drug. When testing a drug you're not looking into the sky to see how an object moves in the sky.

And there's data.

Radar reports - DATA

Trace evidence - DATA

Malfunctioning nukes and evading capture - DATA

Pictures and video - DATA

Eyewitness accounts from ASTUTE OBSERVERS - DATA

Close Encounters from ASTUTE OBSERVERS - DATA

These things are all data that can be used to build the ET hypothesis.

If there wasn't any DATA:

There wouldn't be any talk of U.F.O.'s

ATS might as well shut down the U.F.O. AND ALIEN forum

There wouldn't articles and news stories about U.F.O.'s if there wasn't any DATA that shows there's an aerial observed phenomena called U.F.O.'s.

If there wasn't any DATA for U.F.O.'s the debunkers wouldn't spend 99.99999% of their time worried about trying to debunk something that has no DATA to support it.



posted on Apr, 19 2014 @ 11:23 AM
link   

originally posted by: BayesLike
We know this is the case because we had 2 out of a billion observations with lights overhead and 2 out of a billion observations with stalls. So the combination of a stall and lights overhead should show up in somewhere around 4 times per billion-billion observations.


So how does 4 out of a billion billion add up to 99.99999...% ?



posted on Apr, 19 2014 @ 11:32 AM
link   

originally posted by: BayesLikeWe know this is the case because we had 2 out of a billion observations with lights overhead and 2 out of a billion observations with stalls.


You are dissecting the evidence and analysing it in pieces. It is not just lights and stalls, it is also missing time, encounters with beings etc. etc. etc. So dilute your data with billions and billions. You certainly won't get 99.999........%

You would be lucky to get 0.00000000000000000001%. This is why I say a COMMON SENSE appraisal of the evidence will dispense with the need to go around collecting every light in the sky. It is a waste of time because it doesn't recognize the particulars of the data for ETH nor does it include the corroborative links between various pieces of data.



posted on Apr, 19 2014 @ 12:11 PM
link   
a reply to: ZetaRediculian



That works

Yes but then lets elaborate on this. Wouldnt you say that is exactly the 'debunkers' hypothesis namely 'No ufocases are of ET source' and the believers are trying to refute this by showing the null hypothesis is true ie 'there exist at least one ufocase of ET source'?

If we agree this is a 'scientific' hypothesis regarding falsifiability we could argue that the original ETH in the OP is scientifically sound because proving this hypothesis false automatically proves the orginal ETH true...right?



posted on Apr, 19 2014 @ 12:17 PM
link   
a reply to: neoholographic

And there's data.

Radar reports - DATA

There are radar reports. How are you quantifying them other than say that there are reports?


Trace evidence - DATA

Which includes crop circle reports because there is no clear distinction.


Malfunctioning nukes and evading capture - DATA

How is "evading capture" data? How are you determining that?


Pictures and video - DATA

I can say for sure there isn't any.


Eyewitness accounts from ASTUTE OBSERVERS - DATA
Close Encounters from ASTUTE OBSERVERS - DATA

you have stories. How are you quantifying these stories so that they can be analyzed with inferential statistics?


These things are all data that can be used to build the ET hypothesis.

These things are links and youtube videos they are not "data". You need to define the data. the way you are presenting it is worthless and meaningless. The only thing to be said is that they are stories and links to all sorts of things that are not clearly defined.



If there wasn't any DATA:

If there was data, we would be able to inferential statistics on it as ALL your example demonstrate. How do you do statistics on youtube videos?

If you want to say that these are interesting stories that should be collated in a meaningful way to see what we can make of them, I would agree.



posted on Apr, 19 2014 @ 12:31 PM
link   
a reply to: neoholographic

Tell me what data didn't I use? We're dealing with an aerial observed phenomena, so tell me what data didn't I use? Here's the questions again.

ALL aerial observed phenomenon should be included. That especially includes cases that have been identified and particularly if it was initially thought to be a UFO and described as alien. This would also include known aircraft positively Identified as known aircraft. Another interesting comparison would be unknown aircraft positively identified as an unknown terrestrial aircraft.

You are not looking at data yet. You are looking at select stories of unidentified things that sound like they could be aliens to you. That's all that can be said about it.




top topics



 
8
<< 27  28  29    31  32  33 >>

log in

join