If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(The Planetary Society)   April 1: Bill Nye convinces Congress that Climate Change and evolution are realities. Not April 1: Bill Nye the Science Guy convinces Congress to restore funding for planetary exploration. Hero trumps cool   (planetary.org) divider line 58
    More: Hero, Timeline of Solar System exploration, evolution, U.S. Congress, climate change, Planetary Society, union dues, Planetary Science, NASA  
•       •       •

3284 clicks; posted to Geek » on 02 Apr 2013 at 9:41 AM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



58 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread

First | « | 1 | 2 | » | Last | Show all
 
2013-04-02 09:15:56 AM  
planetary.s3.amazonaws.com

Bowties are cool.
 
2013-04-02 09:26:28 AM  
Bill Nye, kicking ass and chewing bubblegum.
 
2013-04-02 09:52:03 AM  
Hey now. That money could purchase like 1.5 F-35s. What about those manufacturing jobs he so triumphantly pissed away with his little faux "science" tirade?? Typical elitist liberal intellectual America hating Marxist job destroying tax monger. Where will Bill Nye be when North Korea invades Portland?
 
2013-04-02 10:17:35 AM  

Oakenshield: Hey now. That money could purchase like 1.5 F-35s. What about those manufacturing jobs he so triumphantly pissed away with his little faux "science" tirade?? Typical elitist liberal intellectual America hating Marxist job destroying tax monger. Where will Bill Nye be when North Korea invades Portland?


I know you're kidding but it's more like ~94% of one at the current costs.
http://en.wikipedia.org/wiki/Lockheed_Martin_F-35_Lightning_II
 
2013-04-02 10:21:17 AM  

Oakenshield: Hey now. That money could purchase like 1.5 F-35s. What about those manufacturing jobs he so triumphantly pissed away with his little faux "science" tirade?? Typical elitist liberal intellectual America hating Marxist job destroying tax monger. Where will Bill Nye be when North Korea invades Portland?


He'll be running his weather machine from his secret lair atop Bald Peak in an attempt to set the very planet against them, silly.
 
2013-04-02 10:37:03 AM  
 
2013-04-02 10:45:57 AM  

Oakenshield: Hey now. That money could purchase like 1.5 F-35s. What about those manufacturing jobs he so triumphantly pissed away with his little faux "science" tirade?? Typical elitist liberal intellectual America hating Marxist job destroying tax monger. Where will Bill Nye be when North Korea invades Portland?


He'll actually be on his space station nuking them from orbit.
 
2013-04-02 10:54:52 AM  

God Is My Co-Pirate: Bill Nye, kicking ass and chewing bubblegum.


And he's all outta gum.
 
2013-04-02 11:08:49 AM  
This could...dare I say it...CHANGE THE WORLD!
 
2013-04-02 11:18:28 AM  

kudayta: This could...dare I say it...CHANGE THE WORLD!


http://www.youtube.com/watch?v=HHP9Rh-ooh0

for your viewing pleasure
 
2013-04-02 11:24:36 AM  
More reality:

www.woodfortrees.org
But hey, any set of data can have an "off" 15+ years, amirite?
 
2013-04-02 11:35:29 AM  

justtray: God Is My Co-Pirate: Bill Nye, kicking ass and chewing bubblegum.

And he's all outta gum.


He finished his ass kicking for NASA, and went by the minimart.  Now, it's the gum's turn to be afraid.
 
2013-04-02 12:01:04 PM  

SevenizGud: More reality:

[www.woodfortrees.org image 640x480]
But hey, any set of data can have an "off" 15+ years, amirite?


Oh sweet, the green line hovers at just over 0.4?  And the red line is all wiggly?!  I AM DRAWING SO MANY CONCLUSIONS RIGHT NOW

/label yo' axes foo!
 
2013-04-02 12:03:24 PM  

SevenizGud: More reality:

[www.woodfortrees.org image 640x480]
But hey, any set of data can have an "off" 15+ years, amirite?


i25.photobucket.com
 
2013-04-02 12:30:53 PM  

SevenizGud: More reality:

[www.woodfortrees.org image 640x480]
But hey, any set of data can have an "off" 15+ years, amirite?


It's so cute how you always start that graph on the 1998 outlier to deliberately skew the trend.

Oh wait, I did not mean cute, I mean "dishonest and stupid".
 
2013-04-02 12:40:28 PM  
Now, we just need the shuttle or an alternative again...
 
2013-04-02 12:51:39 PM  

rogue49: Now, we just need the shuttle or an alternative again...


We're almost there...
0.tqn.com
 
2013-04-02 12:56:42 PM  

KiltedBastich: SevenizGud: More reality:

[www.woodfortrees.org image 640x480]
But hey, any set of data can have an "off" 15+ years, amirite?

It's so cute how you always start that graph on the 1998 outlier to deliberately skew the trend.

Oh wait, I did not mean cute, I mean "dishonest and stupid".


What do you mean? His method of selecting and analyzing data is perfectly legitimate. In fact, when we apply Sevenizgud's method to the 2012 election results, we can clearly see that Mitt Romney won every state, and is now our president.

s24.postimg.org
 
2013-04-02 01:57:14 PM  
at this point I am starting to think that Bill is the Doctor.

He is just keeping it on the downlow.

/so who is his companion?
 
2013-04-02 03:07:37 PM  
Climate change isn't what is up for debate.  It is whether it is caused by man or not.  Anyone with any education knows that is impossible but like most people that have been brainwashed, the cult of AGW is here to stay as there is way too much money in it for them.
 
2013-04-02 05:09:26 PM  

KiltedBastich: It's so cute how you always start that graph on the 1998 outlier to deliberately skew the trend.


Global warming is so bad that the temperature from 15 years ago is STILL considered an outlier.

In other news, the data start in 1997, Einstain.
 
2013-04-02 05:20:04 PM  

curriemaster: /label yo' axes foo!


The axes are the same as they always are for HADCRUT3VGL plots, Genius.

Lemme guess, you had NO IDEA that 2002 meant the year.

And that is the intelligence of your typical Chicken Little, folks.
 
2013-04-02 06:04:45 PM  

SevenizGud: More reality:

[www.woodfortrees.org image 640x480]
But hey, any set of data can have an "off" 15+ years, amirite?


www.skepticalscience.net
 
2013-04-02 06:15:52 PM  

SevenizGud: curriemaster: /label yo' axes foo!

The axes are the same as they always are for HADCRUT3VGL plots, Genius.

Lemme guess, you had NO IDEA that 2002 meant the year.

And that is the intelligence of your typical Chicken Little, folks.



Weird that you would use HadCRUT3 since it was superceded by HadCRUT4...
www.woodfortrees.org
Oh, that's why.

Well, I suppose if you're going to cherry-pick a period of time to look at, you might as well cherry-pick a data set to use as well.
 
2013-04-02 07:56:40 PM  
 
2013-04-02 08:04:44 PM  

Katolu: AiryAnne: If they find a sandwich, let me know.

Whut?


I think the point there is why spend more money on science when people like that are all over the place.  I say, why not do something about both?  Critizing one thing for getting attention and saying it is at the expense of another is a bit of a BS argument.  That line of thought just perpetuates the race to the bottom where nothing is done about anything.
 
2013-04-02 08:07:54 PM  

lordaction: Anyone with any education knows that is impossible


Citation needed, argument from omniscience, etc.


/Go away, dumb dumb
 
2013-04-02 08:22:39 PM  

Damnhippyfreak: Weird that you would use HadCRUT3 since it was superceded by HadCRUT4...

Well, I suppose if you're going to cherry-pick a period of time to look at, you might as well cherry-pick a data set to use as well.


Or, it could be that, you know, I realize that HadCRUT4 is simply a sham operated so as to use the smoothing fudge-factor/lie "correction" to the data to spread out anomalously high readings over a greater surface area, rather than normalize the spuriously high temperature readings.

But hey, it's okay to cherry-pick that correction factor when it is being done by climate "scientists", amirite?
 
2013-04-02 08:37:45 PM  

SevenizGud: But hey, it's okay to cherry-pick that correction factor when it is being done by climate "scientists", amirite?

 
2013-04-02 10:20:12 PM  

SevenizGud: Damnhippyfreak: Weird that you would use HadCRUT3 since it was superceded by HadCRUT4...

Well, I suppose if you're going to cherry-pick a period of time to look at, you might as well cherry-pick a data set to use as well.

Or, it could be that, you know, I realize that HadCRUT4 is simply a sham operated so as to use the smoothing fudge-factor/lie "correction" to the data to spread out anomalously high readings over a greater surface area, rather than normalize the spuriously high temperature readings.

But hey, it's okay to cherry-pick that correction factor when it is being done by climate "scientists", amirite?


Besides being a bit absurd (there has been no change to the spatial weighting, just the addition of more data covering more surface area), there's a few extreme claims in there sprinkled with some conspiracy theory thrown in. Care to back up any of this?
 
2013-04-02 10:58:51 PM  

Damnhippyfreak: Besides being a bit absurd (there has been no change to the spatial weighting, just the addition of more data covering more surface area)


I should be clear here that adding more data and more coverage does not somehow "spread out anomalously high readings over a greater surface area" as the grid size used in spatial weighting has not changed. It appears that SevenizGud may be unclear how spatial weighting works, in addition to some aspects ofstatistics ("cherry-pick that correction factor" makes no sense in this context).
 
2013-04-03 01:50:18 AM  

Unknown_Poltroon: Oh, look what happens when you don't cherry pick the data from 1998 on??  Lying fark.


That's cute.

Similarly small date.

Various graphs, last 213 years


www.woodfortrees.org

www.woodfortrees.org


Those are "bad" because they're land only?  Here's some Sea only.

www.woodfortrees.org

www.woodfortrees.org

Boy, that looks almost cyclicial...

How about some sea ice?

www.woodfortrees.org
www.woodfortrees.org


But my favorite:

www.woodfortrees.org

An additional 1/3 of previous levels.
No correlated changes in any of the other data.  Shades of a degree, sure, on charts that also place as high or warmer 200 years ago before we started "murdering" the planet.

/all dates were as far back as they had data for.
 
2013-04-03 01:55:16 AM  

Damnhippyfreak: It appears that SevenizGud may be unclear how spatial weighting works


It appears that Stinkyhippy doesn't understand how fudge-factoring works.

Doesn't it bother you to just lie like that?
 
2013-04-03 02:56:39 AM  

SevenizGud: Damnhippyfreak: It appears that SevenizGud may be unclear how spatial weighting works

It appears that Stinkyhippy doesn't understand how fudge-factoring works.

Doesn't it bother you to just lie like that?



You may not be on the winning side of the argument when you have to ignore what the other poster says and just call them names.
 
2013-04-03 03:00:24 AM  

SevenizGud: Unknown_Poltroon: Everyone always points out how you're lying.

The folks at Hadley will be happy to learn that the last 15 years of their data is "a lie".


Cherry-picking a subset of the data relative to the phenomenon of interest could be considered dishonest, especially as you're already aware of the problems with doing so:

SevenizGud: Damnhippyfreak: [socratic] Again, since we're interested in why whether "The earth is not PRESENTLY warming" or not, 4 years would be preferable to 10 or 15 years, yes? [/socratic]
Quite a departure from the Hansen standard of 8 years. I like to be more robust in the analysis, to, you know, take out the variability. That's why 15 years. You know, more scientific. Because global warming is all about the underlying science, and not political footballing and shading the data.


So we know you are very much aware that a short term period (relative to variability) can be misleading. You contend that this is similar to what James Hansen used (supposedly only 8 years) in past congressional testimony. This is not the case, as his testimony and the papers it was based on used a longer period of time than that and did not solely rely on some sort of simple linear regression or simple correlation.
 
2013-04-03 04:18:32 AM  

omeganuepsilon: Unknown_Poltroon: Oh, look what happens when you don't cherry pick the data from 1998 on??  Lying fark.

That's cute.

Similarly small date.

Various graphs, last 213 years

[BEST]
[BEST upper 95% CI]


As long as we're messing with the woodfortrees graph generator, trends are a bit easier to see if we apply a filter (30 years in this case):

woodfortrees.org
Note that the current warming trend is somewhat pronounced.

That aside, keep in mind the second graph you posted is of the upper 95% CI. Be careful in that the inferences you can draw from such are very limited considering that the CI changes over time.


omeganuepsilon: Those are "bad" because they're land only?


The land-only record isn't bad, you just have to keep in mind its limitations.


omeganuepsilon: Here's some Sea only.

[PDO]

[AMO]

Boy, that looks almost cyclicial...


Unsurprisingly, they are indeed cyclical (it's right in the name, after all). PDO and AMO indexes are calculated by first removing other factors (such as anthropogenic climate change) - it's essentially a residual. Not sure what you were trying to get at by posting those.

Anyway, if you wanted to plot sea surface temperature over time (as a counterpart to the BEST data set), HadSST2 is also an option with the woodfortrees graph generator:

www.woodfortrees.org
Again, the current warming trend is evident. If we combine what we see here with the AMO and PDO indexes you posted, you will note that since they are cyclical, they cannot be responsible for the current warming trend.


omeganuepsilon: How about some sea ice?


[SH]
[NH]


Possible trends would be much easier to see if we were to filter out that strong intra-annual variability:

www.woodfortrees.org

www.woodfortrees.org

This is nothing new. There's been some interesting discussion about why Antarctic sea ice is increasing - reasons for which aren't incompatible with anthropogenic climate change. I was just reading an explanation that was novel to me just yesterday - let me see if I can find it. Here, with the press release with more context here. We can discuss this more if you wish.

omeganuepsilon: But my favorite:

[CO2concentration]

An additional 1/3 of previous levels.
No correlated changes in any of the other data.


Except that there are, in the global land and SST data sets. You also have to keep in mind that simple correlation isn't all that good of a heuristic given the conflation of the underlying mechanisms and processes that affect temperature simultaneously - the PDO and AMO indexes you posted are a good example of this. Let me point you to a post I made in the other thread that address some of this in regards to a similar problem; we can discuss it more if you wish.


omeganuepsilon: Shades of a degree, sure, on charts that also place as high or warmer 200 years ago before we started "murdering" the planet.


Keep in mind that this probably isn't true.  Variability increases the further back in the record as uncertainty increases. Put another way, noise increases the further one looks back. Therefore extreme values further in the past are more likely to be spurious rather than representative - one of the reasons why a filter (as a moving average) is commonly used.
 
2013-04-03 04:20:11 AM  

Damnhippyfreak: Here, with the press release with more context here.


Whoops. Forgot that Fark doesn't like Nature links. The paper can be found here:

http://www.nature.com/ngeo/journal/vaop/ncurrent/full/ngeo1767.html
 
2013-04-03 07:56:42 AM  

Damnhippyfreak: omeganuepsilon: Unknown_Poltroon: Oh, look what happens when you don't cherry pick the data from 1998 on??  Lying fark.

That's cute.

Similarly small date.

Various graphs, last 213 years

[BEST]
[BEST upper 95% CI]

As long as we're messing with the woodfortrees graph generator [snipped]


Omega posts a bunch of graphs without actually understanding what they mean or how to apply the information and gets calmly rebuffed.  Beautiful.

But don't call him a denier!  He's just an skeptic trying to keep everyone honest even though every point he attempts to make goes against ACC/AGW.

Damnhippyfreak: Unsurprisingly, they are indeed cyclical (it's right in the name, after all). PDO and AMO indexes are calculated by first removing other factors (such as anthropogenic climate change) - it's essentially a residual. Not sure what you were trying to get at by posting those.


It's pretty obvious that he
A) Had no idea the O stood for oscillation
B) Thought he was making some amazing critique of global warming

It's doubly funny to see him referencing data prior to roughly 1900 as in one of the other recent climate change threads he was complaining that the error bars on non modern records are too large to be trusted.
 
2013-04-03 10:08:51 AM  

Damnhippyfreak: As long as we're messing with the woodfortrees graph generator, trends are a bit easier to see if we apply a filter (30 years in this case):


Still selective use of data every bit as disengenuous as that of others that get called out for being deniers.

Real science doesn't need to adjust and filter and discard data in such ways.


http://en.wikipedia.org/wiki/Law_of_large_numbers

Alarmists math says a rolled dice should have an average of 3, and gets all freaked out when it's actually 3.5.

http://www.probabilitytheory.info/content/item/6-the-law-of-large-nu mb ers-/-the-law-of-averages

It seems that a large number of human minds are just simply unable to cope with the often seemingly contradictory laws of probability. If only they had listened to their maths teacher.
An understanding of the law of the large numbers leads to a realisation that what appear to be fantastic improbabilities are not remarkable at all but, merely to be expected.

How poignant.
 
2013-04-03 12:29:04 PM  

omeganuepsilon: Damnhippyfreak: As long as we're messing with the woodfortrees graph generator, trends are a bit easier to see if we apply a filter (30 years in this case):

Still selective use of data every bit as disengenuous as that of others that get called out for being deniers.

Real science doesn't need to adjust and filter and discard data in such ways.


http://en.wikipedia.org/wiki/Law_of_large_numbers

Alarmists math says a rolled dice should have an average of 3, and gets all freaked out when it's actually 3.5.



Heh. I'm not sure you realize that your example here strongly undermines your own point as it relies on that the idea that an empirically-derived  average is (unsurprisingly) necessary in order to determine the average value of a rolled dice. According your own flawed reasoning, determining that the average of a rolled dice is 3.5 isn't "real science" as it filters and discards data.

Averages aren't somehow the tool of the Devil, you know. You yourself relied on it here for good reason. Think about this for a second - how would you make any inferences about any of the graphs you yourself posted without implicitly or explicitly using averages? I mean, if we go by your current argument, all the data  you yourself previously posted and attempted to make inferences about isn't derived from "real science" as they represent averages.


omeganuepsilon: http://www.probabilitytheory.info/content/item/6-the-law-of-large-nu mb ers-/-the-law-of-averages

It seems that a large number of human minds are just simply unable to cope with the often seemingly contradictory laws of probability. If only they had listened to their maths teacher.


LOL. You should probably read and understand the reasons why that article characterizes the laws of probabalilty as seemingly contradictory, as you're more likely the target of that cautionary statement. The idea that article presents is that rejecting what averages tells us (as you tend to do) can be misleading. It is in fact dismissing the average value (like what you're pretty much doing) that provides the basis for the gambler's fallacy that your own link highlights.  What extreme values tell us us  seemingly contradicts what averages tell us - however it is the former, based on "human instinct" that can be misleading. In short, dismissing what averages tell us (as you're doing) can be misleading.The idea that averages reduce the contribution of extreme values isn't some little-known pitfall, but one of the advantages of it being used, as the article you yourself posted relies on.

I highly suggest that in rejecting the idea of averages it is you who should have "listened to their maths teacher". Come on now. You've presented two bits of evidence that instead make the opposite case of what you're trying to argue. It's probably time to re-evaluate your position here as it's untenable.
 
2013-04-03 12:36:31 PM  

omeganuepsilon: Real science doesn't need to adjust and filter and discard data in such ways.



To bring it home, let's consider the sentences in that article just prior to the ones you quoted:

The number 53 in an Italian lottery had failed to appear for some time and this lead to an obsession with the public to bet ever larger amounts on the number. People staked so much on this corrective that the failure of the number 53 to occur for two years was blamed for several deaths and bankruptcies.

One of the false rationalizations that one of these misguided people would provide for ignoring what averages tell us could be that "real science doesn't need to adjust and filter and discard data in such ways".
 
2013-04-03 12:43:42 PM  

SevenizGud: Global warming is so bad that the temperature from 15 years ago is STILL considered an outlier.

In other news, the data start in 1997, Einstain.


Uh huh. So somehow starting the data set one point BEFORE the huge outlier is suddenly not going to skew the whole data set? You really don't know anything at all about statistics, do you?

omeganuepsilon: Real science doesn't need to adjust and filter and discard data in such ways.


And apparently neither do you.
 
2013-04-03 09:02:58 PM  

Damnhippyfreak: According your own flawed reasoning, determining that the average of a rolled dice is 3.5 isn't "real science" as it filters and discards data.


No, it uses every data piece, discards nothing or tampers with the results in any way.

Unlike alarmist math.

Damnhippyfreak: Think about this for a second - how would you make any inferences about any of the graphs you yourself posted without implicitly or explicitly using averages?


My inferences(and if I made any, it's that anyone can pull graphs from that site and have say a few words to make them look reasonable to laymen, and that the guy who I called out cherry picked just as bad as he was claiming the other guy did.[typical alarmist tactic, do as I say not as I do]
In other words.
Those graphs do not serve as citations, merely talking points about how the references can say whatever the poster wants, kind of like alarmism in general, like FoxNEWS but a different political party.

 =/= The extent alarmists take their Beliefs and what they call proof.

The funny part, is that I don't even deny global warming.  You all just like to pretend I do so that you have someone to argue against.

Really, I'm apathetic to this.  I'm skeptical of the reasoning behind the interpretation of so much data as a single number, supposedly representative of weeks/months/years worth of highly variable real world weather.  Process the data so much and it loses any precision(propagation of uncertainty).

That different organizations get such highly variable graphs, sometimes within the same organization, and it still garners the faith of the masses is sort of sad.  Not surprising in a world where many think vaccines cause autism, or that the earth is 6000 years old, or that Jersey Shore was entertaining, but still, just sad.

I don't argue against what you say, but how you say it, the reasoning that is supposedly proof.  If a religious person goes on and on about god, sure, annoying, but I really can't say it's wrong.  It's when he starts in about how the earth is 6000 years old and how jesus had pet dinosaurs, or the rain is god's tears, or god is love, that kind of idiocy needs to be argued against.  That's the human analog of error propagation   Call it what you will, misinformation, lying through your teeth, or extreme willful ignorance of the faithful.

Einstein could say "1+1=2 Because Unicorns ejaculated into my left ear."
And yes, rational people will have a problem with that, as it indeed does have issues, because the days that it is not a unicorn and instead a 350lb furry dressed like a tiger, his argument is faulty.
 
2013-04-03 09:26:14 PM  

KiltedBastich: omeganuepsilon: Real science doesn't need to adjust and filter and discard data in such ways.

And apparently neither do you.


Point?

A confounding variable.  What of it?  Did you even read the page?  That's an accusation I may toss out there as food for thought.
A perceived relationship between an independent variable and a dependent variable that has been misestimated due to the failure to account for a confounding factor is termed a spurious relationship, and the presence of misestimation for this reason is termed omitted-variable bias.

Sounds rather exactly like something I may say.

Yes, outliers are removed, as they are not "data".  A sensor that reads 500 degrees is broken.  90 degrees in January in south dakota?  Yeah, not happening either.

But some of the organizations go above and beyond that, and is evident when you really read their methodology reports, they're quite blatant about it. Berkeley has attempted to not do it so harshly(read: obviously), but they still do so.  They all dampen their references for the floating average, for example, and attempt to extrapolate (read: guess) at missing data.  sure, some guesses are more reasonable than others, but none are certain by any means, so that only adds another error to propagate.
 
2013-04-03 09:28:02 PM  

KiltedBastich: you.


And that link was relevant how?

You're going to have to explain that one, because, in your very limited context, you may as well have said "math".
 
2013-04-03 10:47:48 PM  

omeganuepsilon: Damnhippyfreak: According your own flawed reasoning, determining that the average of a rolled dice is 3.5 isn't "real science" as it filters and discards data.

No, it uses every data piece, discards nothing or tampers with the results in any way.

Unlike alarmist math.


Note that the average you used discarded the range of possible values and their distribution ;)

Again, averages aren't some sort of inherently bad thing - you just have to realize it's limitations as well as strengths (that you yourself are relying on).


omeganuepsilon: Damnhippyfreak: Think about this for a second - how would you make any inferences about any of the graphs you yourself posted without implicitly or explicitly using averages?

My inferences(and if I made any, it's that anyone can pull graphs from that site and have say a few words to make them look reasonable to laymen, and that the guy who I called out cherry picked just as bad as he was claiming the other guy did.[typical alarmist tactic, do as I say not as I do]
In other words.
Those graphs do not serve as citations, merely talking points about how the references can say whatever the poster wants, kind of like alarmism in general, like FoxNEWS but a different political party.


Not quite - while I'm sure one can use data to say a variety of things, there's only so far it can stretch. For instance, the specific claims that you made that are somewhat false (and you've notably failed to address). I mean, you can try to have data "say whatever the poster wants", but that does not mean you're correct.


omeganuepsilon: =/= The extent alarmists take their Beliefs and what they call proof.

The funny part, is that I don't even deny global warming.  You all just like to pretend I do so that you have someone to argue against.

Really, I'm apathetic to this.  I'm skeptical of the reasoning behind the interpretation of so much data as a single number, supposedly representative of weeks/months/years worth of highly variable real world weather.  Process the data so much and it loses any precision(propagation of uncertainty).


Being skeptical is fine - but you have to realize the limitations of the knowledge you're basing such skepticism on. From what I've seen here, your knowledge of this topic is very superficial, which is just fine, but puts limits on how accurate your perception is. I suggest you put some of the energy you're expending here into actually learning about the topic - you're certainly actively attempting to discuss this for someone who is "apathetic".


omeganuepsilon: That different organizations get such highly variable graphs, sometimes within the same organization, and it still garners the faith of the masses is sort of sad.  Not surprising in a world where many think vaccines cause autism, or that the earth is 6000 years old, or that Jersey Shore was entertaining, but still, just sad.


Unfortunately, that's the nature of real-world data - variation is the norm, not the exception. That such variation exists does not mean that the information gathered isn't useful. Think about this for a second -the knee-jerk reaction you're having to using averages would also discount the range (and, yes, variable) estimations of the age of the Earth. That variation exists does not somehow mean that the resulting information has to be taken on faith.


omeganuepsilon: I don't argue against what you say, but how you say it, the reasoning that is supposedly proof.  If a religious person goes on and on about god, sure, annoying, but I really can't say it's wrong.  It's when he starts in about how the earth is 6000 years old and how jesus had pet dinosaurs, or the rain is god's tears, or god is love, that kind of idiocy needs to be argued against.  That's the human analog of error propagation   Call it what you will, misinformation, lying through your teeth, or extreme willful ignorance of the faithful.

Einstein could say "1+1=2 Because Unicorns ejaculated into my left ear."
And yes, rational people will have a problem with that, as it indeed does have issues, because the days that it is not a unicorn and instead a 350lb furry dressed like a tiger, his argument is faulty.


Let me just first say that a good way to combat what you're talking about is to rely on evidence and solid reasoning.

With that out of the way, I get what you're saying here, but I'm sorry, from what I've seen here you're not practicing what you're preaching. You've misused data and retreated behind the excuse that you weren't serious and just demonstrating "that anyone can pull graphs from that site and have say a few words to make them look reasonable to laymen" (as if saying that you were arguing in bad faith is any better). You've repeatedly posed a somewhat absurd critique of averages then presented links and examples that argue against your own claim. You ignore yourself acknowledge that you're "apathetic" to the issue but you still put some effort into expressing your uninformed opinion on the topic. Worse yet, you largely ignore the argumentation where the above is rationally reasoned out to you.

Come on now - if you really believe what you say, start walking the talk yourself.
 
2013-04-03 11:28:20 PM  

omeganuepsilon: KiltedBastich: omeganuepsilon: Real science doesn't need to adjust and filter and discard data in such ways.

And apparently neither do you.

Point?

A confounding variable.  What of it?  Did you even read the page?  That's an accusation I may toss out there as food for thought.
A perceived relationship between an independent variable and a dependent variable that has been misestimated due to the failure to account for a confounding factor is termed a spurious relationship, and the presence of misestimation for this reason is termed omitted-variable bias.

Sounds rather exactly like something I may say.


But apparently not understanding, as this would strongly argue against the approach you've been using thus far.  I'll get into it more further down, but one of the ways one accounts for a confounding factor is to minimize its influence through averages. This attempting to "account for a confounding factor" is the reason one uses averages in the first place (and something you seem hell-bent on arguing against).


omeganuepsilon: Yes, outliers are removed, as they are not "data".  A sensor that reads 500 degrees is broken.  90 degrees in January in south dakota?  Yeah, not happening either.


You're missing the point. The idea is that such outliers are a case of processes that we aren't interested in - something like the working condition of a sensor is in fact a confounding variable. There are other outliers (or less extreme, yet unwanted variation) that are due to other confounding variables that we aren't interested in - just like we aren't interested in the variation a broken sensor provides. Let's go back to some of those graphs we posted earlier in order to provide an example. You posted this graph:

www.woodfortrees.org

If we're interested in the change in Arctic sea ice extent over the course of long periods of time, we may consider using simple linear regression to get an idea of the rate. We would most likely be unable to do so as the amount of variation is so high. When we look at the graph, we notice that the high amount of variation is due to a confounding variable that we aren't interested in - in this case intra-annual change due to seasonality. So how would we account for the confounding variable that seasonal variation provides? One way would be to filter it out using yearly averages, which is exactly what I did:

www.woodfortrees.org

There we go. It is now much more likely that we would be able to get a valid linear regression and an idea of the rate of change.

So what this has demonstrated is how using an average is an example of accounting for confounding variables that make analyzing a perceived relationship much more difficult if not impossible. We can run through more if you're still fuzzy on the idea. I hope you're starting to get the idea that averages aren't somehow the Bad Thing you're making it out to be.


omeganuepsilon: But some of the organizations go above and beyond that, and is evident when you really read their methodology reports, they're quite blatant about it. Berkeley has attempted to not do it so harshly(read: obviously), but they still do so.  They all dampen their references for the floating average, for example


What you have to realize is that these methods are meant to reduce error - as in an attempt to get a more accurate picture of temperature. The example I just ran through is similar to the use of a floating average. We are interested in long-term changes due to climate, not short-term ones, such as ENSO. So in this way, short-term changes are confounding variables - unwanted variation that makes analyzing a perceived relationship much more difficult. One way to account for such short-term variation (that, again, we aren't interested in) is to filter it out using an average - very much like in the example I ran through. Again, averages are useful in this context.

omeganuepsilon: and attempt to extrapolate (read: guess) at missing data.  sure, some guesses are more reasonable than others, but none are certain by any means, so that only adds another error to propagate.


Extrapolation isn't a guess since one uses existing data. It's of course imperfect, but it's sometimes better than no information at all. Again, this is meant to get a more accurate picture of temperature.

I can understand one being skeptical of such methods, but you have to attempt to know the reasons for doing so in order for such skepticism to be more than an argument from incredulity.
 
2013-04-03 11:46:26 PM  

omeganuepsilon: KiltedBastich: you.

And that link was relevant how?

You're going to have to explain that one, because, in your very limited context, you may as well have said "math".



I don't wish to speak for KiltedBastich, but I can provide additional information. Multivariate statistics (as well as statistics in general) are highly dependent on the use of averages or techinques that "adjust and filter and discard data". To highlight examples, the wiki page that was linked to provides a list of some of the common techniques:

MANOVA - relies on averages during calculation and even in the end spits out whether the means are significantly different between variables

Multivariate regression analysis - relies on averages and discards information about variation in order to generalize a relationship - in fact, the fun stat fact of the day is that any least-squares linear regression generates a line that by necessity includes through the mean of both variables. [the more you know.jpg]

PCA - even worse than just averages discards the variables themselves (we can get into this if you're curious)

etc...

The general idea is that if you're going to exclaim that

omeganuepsilon: Real science doesn't need to adjust and filter and discard data in such ways.


You have a problem with a heck of a lot of science, and statistics, and (as I pointed out in an earlier thread) - pretty much any mathematical operation as they all lose information. Again, I urge you abandon your increasingly irrational Jihad against averages.
 
2013-04-04 12:24:46 AM  
(This comment has been removed)
 
2013-04-04 02:31:37 AM  
http://www.savoryinstitute.com/desertification/

Done in one, Or are you guys still wasting time trying to eliminate petrol?

While your getting up to speed on why there is so much carbon running around you should stop sending fake scientists with honorary degrees to congress just because they are celebrities.

Who is next?  Stantz Spengler and Venkman?
 
Displayed 50 of 58 comments

First | « | 1 | 2 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »






Report