Syria's Assad used chemical weapons on his own people

yes, plus the spss output please

(pretty please?)

  On 9/12/2013 at 3:28 AM, eugene said:

 

there's nothing smug about bringing some actual social science understandings to these discussions, that's what it's for. are sr4 or xxx being smug when they go in depth about history and chemistry/medicine respectively ? im not pretending to be some ultimate authority on those issues, i've merely done a b.a. in sociology and im currently in the middle of my studies for masters. it's just that some of this stuff that's brought in those political threads can be easily challenged even with very basic knowledge, like the fact that that 83% figure from the article is complete, meaningless bullshit if you consider the actual statistics. it's all just a mental exercise mostly but hopefully the byproduct will make watmm less fucktarded when it comes to such threads.

 

im not saying sociological stats or actual historical facts are useless, but i will say that both are very much open to interpretation, and both can be bent or placed into contexts of arguments in ways that may be less than honest to fit agendas. sociology is a soft science, if it's a science.

Anybody any idea what the US's goals are in Syria? It's energy-based geo-politics, but I'm not sure what Obama's ideal end result would be.

From what I've read the alliances in the area are:

US - Saudi Arabia

Russia - Syria

France - Qatar - Turkey

Opponents regarding geo-political influence:

Saudi Arabia - Qatar

(Presumably) Russia - US - France?

Opponents regarding energy supply to Europe:

Qatar - Russia

(Russia and Qatar are direct rivals of gas supply. Qatari success equals more influence for them in area and potentially decreases Saudi influence.)

France - Syria

So, what is the US strategy?

Side with Saudi Arabia and install a pro-Saudi Syrian regime that has no interest in piping Qatari gas to Europe.

Pro: Help their traditional allies to curb Qatari influence.

Con: Europe retains Russian energy dependency.

Side with France/Qatar and install a pro-Qatari pipeline regime.

Pro: Decreases European dependency on Russian energy.

Con: Would harm Saudi alliance and possibly destabilise House of Saud.

Something else?

What will happen to Syria in the aftermath of regime change, perpetual civil war, attracting fighters from all over the ME?

ok, i admit i my intuitive statistics skills failed miserably

it does looks statistically significant in spss when you run a logistic regression with a large sample for statistical power, i copy pasted the same 17 senators and their data 20 times so the sample size is 340.

the most important part of spss output looks like this:

Variables in the Equation
B S.E. Wald df Sig. Exp(B)
cash10k .231 .037 39.922 1 .000 1.260
Constant -.883 .213 17.260 1 .000 .414
a. Variable(s) entered on step 1: cash10k.

what it means is that for every 10k $ of defense donation the chance to vote yes (yes =1 in the data) increases by 26% (that exp(b) figure).

still the causation imo is completely opposite like i mentioned before, i.e donors are likely to donate to those who they believe will vote for pro-defense industry decisions. it's just much easier to explain than actual bribery that has to overcome many obstacles, i mean no one would disagree that different people hold different ideas and ideologies. but yeah, tables like this can really help to convince the already convinced.

statistics subforum noa! after approximately X runs, with a max error of 0.001.

  On 9/12/2013 at 12:55 PM, eugene said:

ok, i admit i my intuitive statistics skills failed miserably

it does looks statistically significant in spss when you run a logistic regression with a large sample for statistical power, i copy pasted the same 17 senators and their data 20 times so the sample size is 340.

 

the most important part of spss output looks like this:

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

cash10k .231 .037 39.922 1 .000 1.260

Constant -.883 .213 17.260 1 .000 .414

a. Variable(s) entered on step 1: cash10k.

 

what it means is that for every 10k $ of defense donation the chance to vote yes (yes =1 in the data) increases by 26% (that exp(b) figure).

still the causation imo is completely opposite like i mentioned before, i.e donors are likely to donate to those who they believe will vote for pro-defense industry decisions. it's just much easier to explain than actual bribery that has to overcome many obstacles, i mean no one would disagree that different people hold different ideas and ideologies. but yeah, tables like this can really help to convince the already convinced.

Neat - so even when your own data doesn't align with your belief you stick with your gut feeling.

/troll

On the reals though - I find it very interesting that people perceive the US and Canada as being relatively corruption free nations (for a full range of perceptions - visit transparency.org) when all we've done is label it as lobbying and adopted it as a norm. It's not bribery, it's the legitimized practice of giving money to influence votes.

Now what would be interesting, and I don't have the time to do this, is to see if a certain industry (agriculture, defense, pharmaceutical) gives more to politicians who have voted favourably for them in the past, or if they give more to those whose vote they are trying to sway.

the data doesn't align with anyone's belief actually, it can't show whether it's "pay for votes" or "support those who already vote in alignment with your interests". that's why it doesn't help much anyway and it's disingenuous to portray it as if it does show one scenario or another.

my assumption that it's the other way around is just more common sense grounded, i don't believe that usa is at such state of corruption that your default assumption should be that every politician takes bribes (it doesn't matter if it's sort of legitimized, people still realize when they're given money to do something very particular).

My take: if it can happen, it does happen. The connection between politics and money needs to be highly regulated. Simple as that. A democracy simply cannot function without regulating the money around political influence. There's no need for any statistics. It's a moral and ethical argument.

  On 9/12/2013 at 7:50 PM, eugene said:

the data doesn't align with anyone's belief actually, it can't show whether it's "pay for votes" or "support those who already vote in alignment with your interests". that's why it doesn't help much anyway and it's disingenuous to portray it as if it does show one scenario or another.

 

my assumption that it's the other way around is just more common sense grounded, i don't believe that usa is at such state of corruption that your default assumption should be that every politician takes bribes (it doesn't matter if it's sort of legitimized, people still realize when they're given money to do something very particular).

Well the easy way to find that out is to look how those particular candidates voted previously, and when the money started to flow.

Of course politicians don't take bribes, that would be unethical. They take lobbyist's money purely so they can continue to serve in the best interests of their constituents. Because without that money they would be forced to survive on their measly congressional salaries.

  On 9/12/2013 at 2:46 AM, eugene said:

it's not even statistically significant you dummy, i mean looks at the actual numbers: 86k, 62k, 59k - vote no, and then there are 19-60k'ers who voted yes, even if you multiply the sample while retaining the same data by 1000 you will get no significant effect of amount of defense cash on yes vote (or a significant effect which will be close to zero), it's obvious even without running the regression or whatever. i should have looked at the actual numbers and distribution before, would have saved some time.

  On 9/12/2013 at 2:55 AM, eugene said:

look brah, you're not familiar with stats and it shows, i have enough experience and intuition to know it's not gonna be significant in any way with such numbers, do you really want me to go full +31-on-autism-test and do the whole logistic regressions thing ?

  On 9/12/2013 at 3:12 AM, eugene said:

we're talking specifically about the article you brought here which can't tell anything like that at all, in fact it doesn't tell anything period, you're still not convinced of this ?

  On 9/12/2013 at 3:28 AM, eugene said:

it's just that some of this stuff that's brought in those political threads can be easily challenged even with very basic knowledge, like the fact that that 83% figure from the article is complete, meaningless bullshit if you consider the actual statistics. it's all just a mental exercise mostly but hopefully the byproduct will make watmm less fucktarded when it comes to such threads.

  On 9/12/2013 at 3:32 AM, eugene said:

yeah, which if you enter in spss, for example, to test whether amount of cash has an effect on vote via logistic regression (or any other method) will be not significant (and not because of small sample). just trust me on this, you didn't want me to paste the whole spps output here after all.

  On 9/12/2013 at 12:55 PM, eugene said:

ok, i admit i my intuitive statistics skills failed miserably

it does looks statistically significant in spss when you run a logistic regression with a large sample for statistical power, i copy pasted the same 17 senators and their data 20 times so the sample size is 340.

 

the most important part of spss output looks like this:

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

cash10k .231 .037 39.922 1 .000 1.260

Constant -.883 .213 17.260 1 .000 .414

a. Variable(s) entered on step 1: cash10k.

 

what it means is that for every 10k $ of defense donation the chance to vote yes (yes =1 in the data) increases by 26% (that exp(b) figure).

still the causation imo is completely opposite like i mentioned before, i.e donors are likely to donate to those who they believe will vote for pro-defense industry decisions. it's just much easier to explain than actual bribery that has to overcome many obstacles, i mean no one would disagree that different people hold different ideas and ideologies. but yeah, tables like this can really help to convince the already convinced.

lol that's pretty priceless

ouch.

but eugene would say in his defense that the stats don't indicate anything one way or the other? how does this work?

math has been involved in the argument, and thus I have no point of reference.

STATISTICS

  On 9/14/2013 at 4:56 AM, SR4 said:

ouch.

 

 

but eugene would say in his defense that the stats don't indicate anything one way or the other? how does this work?

 

 

math has been involved in the argument, and thus I have no point of reference.

Well he's saying that the causality is not clear. He's saying that perhaps the type of senators that would vote 'yes' on something like this are the type that would naturally attract more defense contributions in the first place.

so who brought up the whole idea of the statistics in the first place though? i dont feel like looking. but if someone brings up statistics, to say that they aren't significant and don't show or support the possibility of a specific connection between two things, then it turns out those stats are actually showing some kind of correlation or another... that then at least supports the possibility of the very connection that was claimed to be basically disproved by those statistics...

and thats what i would call an eggy face

but we all get egg and some of us possibly seed on our faces at some point

i havent been paying attention tho so maybe someone else brought up the stat-tit-dicks

i'll assume it may have been godel

found this on Annie Hall's HuffPo:

Aztecs Influence Public Affairs Committee

85% of Aztec Scienceologists deny Aztec influence is trying to buy creation of new Aztec state in Mexico

"some anti-Aztecites have claimed we want to turn Mexico into a new Aztec state, that we've been using money to get the USA to achieve our goal. these stupid fools keep saying it's the truth. BULLSHIT. Aztec scienceology proves otherwise. 95%. can't fuck with that number right there, son. it has a nine and a five, followed by a per cent sign. unfuckable. president enrique nieto must hand over the chemical weapons immediately or we'll make him to watch spy kids 3 on a DVD loop for 24 hours a day for 365 days in a row. putin beware!!!!!!"

  On 9/14/2013 at 5:55 AM, MisterE said:

so who brought up the whole idea of the statistics in the first place though? i dont feel like looking. but if someone brings up statistics, to say that they aren't significant and don't show or support the possibility of a specific connection between two things, then it turns out those stats are actually showing some kind of correlation or another... that then at least supports the possibility of the very connection that was claimed to be basically disproved by those statistics...

and thats what i would call an eggy face

but we all get egg and some of us possibly seed on our faces at some point

 

i havent been paying attention tho so maybe someone else brought up the stat-tit-dicks

i'll assume it may have been godel

Thanks sweetheart. I'd love to say that this time around it wasn't me being the stat-tit-dick. Although I must admit that, statistically speaking, your assumption wasn't that far of the mark. ;)

  On 9/14/2013 at 2:15 AM, luke viia said:

 

  On 9/12/2013 at 2:46 AM, eugene said:

it's not even statistically significant you dummy, i mean looks at the actual numbers: 86k, 62k, 59k - vote no, and then there are 19-60k'ers who voted yes, even if you multiply the sample while retaining the same data by 1000 you will get no significant effect of amount of defense cash on yes vote (or a significant effect which will be close to zero), it's obvious even without running the regression or whatever. i should have looked at the actual numbers and distribution before, would have saved some time.

 

 

  On 9/12/2013 at 2:55 AM, eugene said:

look brah, you're not familiar with stats and it shows, i have enough experience and intuition to know it's not gonna be significant in any way with such numbers, do you really want me to go full +31-on-autism-test and do the whole logistic regressions thing ?

 

 

  On 9/12/2013 at 3:12 AM, eugene said:

we're talking specifically about the article you brought here which can't tell anything like that at all, in fact it doesn't tell anything period, you're still not convinced of this ?

 

 

  On 9/12/2013 at 3:28 AM, eugene said:

it's just that some of this stuff that's brought in those political threads can be easily challenged even with very basic knowledge, like the fact that that 83% figure from the article is complete, meaningless bullshit if you consider the actual statistics. it's all just a mental exercise mostly but hopefully the byproduct will make watmm less fucktarded when it comes to such threads.

 

 

  On 9/12/2013 at 3:32 AM, eugene said:

yeah, which if you enter in spss, for example, to test whether amount of cash has an effect on vote via logistic regression (or any other method) will be not significant (and not because of small sample). just trust me on this, you didn't want me to paste the whole spps output here after all.

 

 

  On 9/12/2013 at 12:55 PM, eugene said:

ok, i admit i my intuitive statistics skills failed miserably

it does looks statistically significant in spss when you run a logistic regression with a large sample for statistical power, i copy pasted the same 17 senators and their data 20 times so the sample size is 340.

 

the most important part of spss output looks like this:

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

cash10k .231 .037 39.922 1 .000 1.260

Constant -.883 .213 17.260 1 .000 .414

a. Variable(s) entered on step 1: cash10k.

 

what it means is that for every 10k $ of defense donation the chance to vote yes (yes =1 in the data) increases by 26% (that exp(b) figure).

still the causation imo is completely opposite like i mentioned before, i.e donors are likely to donate to those who they believe will vote for pro-defense industry decisions. it's just much easier to explain than actual bribery that has to overcome many obstacles, i mean no one would disagree that different people hold different ideas and ideologies. but yeah, tables like this can really help to convince the already convinced.

 

lol that's pretty priceless

 

yeah it is a fail on my part but it's more of a fail on your part anyway because it looks like you still believe that this correlation (now that we're clear on its statistical significance) is actually a proof of your claim that "one dollar one vote".

  On 9/14/2013 at 4:56 AM, SR4 said:

ouch.

 

 

but eugene would say in his defense that the stats don't indicate anything one way or the other? how does this work?

 

 

math has been involved in the argument, and thus I have no point of reference.

the stats show that vote is correlated with defense donations, they can't show whether it's the money that causes vote to be "yes" or vote yes (or vote yes'ers) to attract more money

  On 9/14/2013 at 10:34 AM, eugene said:

 

  On 9/14/2013 at 4:56 AM, SR4 said:

ouch.

 

 

but eugene would say in his defense that the stats don't indicate anything one way or the other? how does this work?

 

 

math has been involved in the argument, and thus I have no point of reference.

the stats show that vote is correlated with defense donations, they can't show whether it's the money that causes vote to be "yes" or vote yes (or vote yes'ers) to attract more money

 

but the stats still show more of a possibility of the money causing yes vote than youd have a case for without the stats..

i dont know if that sentence was grammatically sound, but i think it shows why you don't bring stats into these things

because you make an istics out of u and me

i think i messed that up but, its pretty late and im drunk and on dope, so..

  On 9/14/2013 at 10:58 AM, MisterE said:

 

  On 9/14/2013 at 10:34 AM, eugene said:

 

  On 9/14/2013 at 4:56 AM, SR4 said:

ouch.

 

 

but eugene would say in his defense that the stats don't indicate anything one way or the other? how does this work?

 

 

math has been involved in the argument, and thus I have no point of reference.

the stats show that vote is correlated with defense donations, they can't show whether it's the money that causes vote to be "yes" or vote yes (or vote yes'ers) to attract more money

 

but the stats still show more of a possibility of the money causing yes vote than youd have a case for without the stats..

 

 

nope, not true. you might as well say that the strength of some senator's ideology about some issues are what causing more money to flow. there's no getting around that issue of causality with the data in that page.

i think misread your post actually, so yeah without that correlation both scenarios are impossible, with it you know that they're possible but don't know which is actually taking place.

You might want to get away from the causality/correlation discussion altogether.

Regardless of which scenario (power attracts money, or money attracts votes) is at play, there are a couple of things which could be argued as a given: money and power like to mingle; and they both (money; power) help each other. So whenever, in the game of politics, the money is unevenly distributed between the political players, the game becomes rigged. Whether it's the powerful players gaining more money, or the players who can be bought for certain votes, thing is, the game becomes inbalanced. When powerful players get more money, they become even more powerful (in a game which is close to zero-sum) with respect to the other players. And when votes are being bought, well, the game is not even played how it's supposed to at all (and it probably stimulates players to be bought more easily, because they make more money doing so).

Excuse me my simplistic views, but the whole argument about causality is, imo, not relevant. The fact that money is involved all by itself is problematic by definition. It distorts the political game. In the ideal political game, any voice would have the same weight. At the second money enters as a factor into the game, this ideal is basically lost.

well yeah, i guess ideally state would allocate equal sums of cash for campaign to each candidate and prohibit additional donations and usage of own capital and so on. but the question at hand is whether u.s. politicians are generally corrupt/bribable or not.

They certainly are generally corrupt. And not just in the US.

  On 9/14/2013 at 1:37 PM, Godwin Austen said:

They certainly are generally corrupt. And not just in the US.

what makes you think so ?