Instead, they used bad math, with no evidence & no context, to transform Mr. Kelley’s pollution cleanup efforts with the “C” word… as posted @ Phsyic’s Today (here):

…his hometown… was troubled by the pollution—the city and surrounding county had one of the highest levels of air pollution in the US, and residents suffered from cancer rates that were 23% higher than the state average, according to the Texas Cancer Registry.

The problem is this statement seems to imply that this particular “pollution” led to higher cancer rates, yet no evidence is presented.

Not only is no evidence presented, but additionally statistics teaches us that in any broad distribution, pockets within that distribution will contain statistical anomalies.

As a thought experiment – try this analogous (numbers based upon TX population versus Port Arthur), yet meaningless statement:

“when testing the 50/50 proposition of coin flips, out of 24 million total flips, in a group of 55K of those flips, there as a 23% higher incidence rate of ‘heads’ than in the total.”

Infinite Monkey Theorems

Headlines 

Worth Reading 

….or at least pondering and forgetting….. 

From the First Amendment Center, the new Alabama governor displays amazing religious intolerance and arrogance.  I thought this was 2011….. (whole thing here): 

BIRMINGHAM, Ala. — Alabama Gov. Robert Bentley told a church crowd just moments into his new administration that those who have not accepted Jesus as their savior are not his brothers and sisters, shocking some critics who questioned yesterday whether he could be fair to non-Christians. 

“Anybody here today who has not accepted Jesus Christ as their savior, I’m telling you, you’re not my brother and you’re not my sister, and I want to be your brother,” Bentley said Jan. 17, his inauguration day, according to The Birmingham News. 

From the Obama Administration: Remember Obamacare?  Which was going to add 39 million previously uninsured people to the status of insured (here via CBO)?  

Well, here we are in 2011 – a time when Obamacare is *not* implemented and the provisions that have gone into effect only went into affect on January 1, 2011. 

Apparently that’s a very long time though….. as according to the WhiteHouse via the Department of Health and Human Services, repealing Obamacare will put 129 million insured at risk (here via HealthCare.gov). 

Seriously?  I wonder if DHHS is still accepting information on those (here via DA) dealing in misinformation with regards to Obamacare? 

From Wired, a meaningless, and based upon presented evidence, a false headline [emphasis added] Supreme Court Upholds Intrusive Government Background Checks.   

The actual article?  

The Supreme Court ruled that private contractors working for the government cannot be shielded from background investigations based upon a right to privacy.  That government contractors can in fact, by virtue that they are basically government employees, be treated just as any other federal employee. 

Maybe it’s just me, but subjecting yourself to a background check that resembles the exact same background check of others you work with doesn’t seem to be intrusive.  

Which is irregardless for Wired anyway, as even *if* this decision could be argued logically as intrusion, the article doesn’t even attempt to offer proof of such an assertion. 

From eScience News, US Office of Naval Research announces big news on the “Cool Things That Kill” front (here): 

Scientists at Los Alamos National Lab, N.M., have achieved a remarkable breakthrough with the Office of Naval Research’s Free Electron Laser (FEL) program, demonstrating an injector capable of producing the electrons needed to generate megawatt-class laser beams for the Navy’s next-generation weapon system. 

Artist's Rendering "PHALANX WITH LASER CANNON" Source: Raytheon

To put a little context into what megawatt means (1,000 kilowatts), Scientific American reports in July 2010 (here): 

In a grainy, black-and-white video that looks like a home movie of a UFO attack a sleek aircraft streaks through the sky one minute, only to burst into flames the next and plummet into the sea…. 

Using a 32-kilowatt laser (article cont’d): 

The defense contractor says it depicts part of a test conducted in May during which the U.S. Navy used a solid-state laser to shoot down unmanned aerial vehicles over the Pacific Ocean…. 

& Lastly – ESO’s Hidden Treasure Contest reveals winner (here): 

M78 for ESO Processing contest. WFI camera on 2.2m telescope

Hidden Treasures gave amateur astronomers the opportunity to search ESO’s vast archives of astronomical data for a well-hidden cosmic gem. Astronomy enthusiast Igor Chekalin from Russia won the first prize in this difficult but rewarding challenge…. 

More amazing astronomical artwork here: Top 100 from ESO

Wired reports on Darpa – that agency which built the internet, now wants a new mathematical language to describe everything (here):

The very first step? Create a unified mathematical language for everything the military sees or hears.

The armed forces are overwhelmed by all the data its various sensors are sniffing out. They want a single data stream that combines drone video feeds, cell phone intercepts, and targeting radar. Darpa’s solution, found in the brand-new Mathematics of Sensing, Exploitation, and Execution program is to design an algorithm that teaches the sensors how to interpret the world — how to think, how to learn and what data, accordingly to collect.

The Economist debates: This house believes that restricting the growth of cities will improve quality of life (here).  An interesting topic, with the debate revolving around whether size is a problem and if so, forcing a certain size or giving individuals freedom to choose.  Research, not (yet at least) discussed,  has been attempted in the past to find the perfect size for a city; meaning how large does a city get before standard city services such as garbage collection or policing become less effective with the addition or each new citizen.

Either way, I’m still for free choice.

SCOTUS Blog on an upcoming Supreme Court arguments about corporate “person hood” (here):

At 10 a.m. Tuesday, the Court will hear one hour of oral argument on a government appeal arguing that business corporations do not have a right of of “personal privacy” that shields from compelled public disclosure the records they turn over to federal agencies.

From Stratfor, especially needed in light of gun control regulation based upon a single incident, Separating Terror from Terrorism. The piece concludes with this:

Recognizing that terrorist attacks, like car crashes and cancer and natural disasters, are a part of the human condition permits individuals and families to practice situational awareness and take prudent measures to prepare for such contingencies without becoming vicarious victims. This separation will help deny the practitioners of terrorism and terror the ability to magnify their reach and power.

]]> http://detailedabstractions.com/2011/01/19/infinite-monkey-theorems/feed/ 0 http://detailedabstractions.com/2010/12/16/update-economists-language-debate/?utm_source=rss&utm_medium=rss&utm_campaign=update-economists-language-debate http://detailedabstractions.com/2010/12/16/update-economists-language-debate/#comments Thu, 16 Dec 2010 14:55:11 +0000 Michael S. Langston http://detailedabstractions.com/?p=1607 Yesterday, DA posted an article on the Economist’s debate over language (here) with the following [emphasis added]:

Irregardless, she goes forward to talk about more interesting research, including ethnic bias, gender specific nouns, and even points to a study which attempts to show Hebrew-Arabic bilinguals speakers seem to show more favorable attitudes towards Jews when tested in Hebrew than Arabic.  (I find that claim dubious, but without more access to the research, I’ll leave that point for now.)

My initial concerns with this research were how it was conducted and what assumptions were made.  After some searching, I’m unable to retrieve the actual study, but did find some underlying issues worth nothing.

The study Ms. Boroditsky cites above, used a test known as Implicit Association Test (IAT) (here via Association for Psychological Science – APS):

The study used a computer test known as the Implicit Association Test, which is often used to study bias….

The question at hand then, is whether IAT can actually show true bias.  The way it works (APS continued):

…Words flash on the computer screen, and subjects have to categorize them by pressing two keys on the keyboard as quickly as possible. It’s a nearly automatic task, with no time to think about the answers. The trick is, the subjects are classifying two different kinds of words: words describing positive and negative traits and, in this case, names – Arab names like Ahmed and Samir and Jewish names like Avi and Ronen. For example, they might be told to press “M” when they saw an Arab name or a word with a good meaning, or “X” when they saw a Jewish name and a word with a bad meaning. In this example, if people automatically associate “good” words with Arabs and “bad” with Jews, they’ll be able to do the classifications faster than if their automatic association between the words is the other way around. In different sections of the test, different sets of words are paired….

 The idea being that a quick test might help to show hidden biases, even for those who are actively attempting to prevent such displays.

Intuitively however, this test seems unlikely to prove bias for a few reasons. 

The first of which is our knowledge that human’s subconscious isn’t conscious.  By its very definition, we don’t know what it is and we still have little idea of how, or even whether, these hidden thoughts interact with our conscious mind or thinking.

The second is our knowledge of overall human development.   Take a child raised and consistently indoctrinated with racist ideals.  While not simple, some of these children do grow up and by themselves learn the truth:  hating others based upon their race, religion, or other superficial factors, is not just stupid, but also marks one of the lowest or immoral thoughts.

But take this person… Any guesses on how this person might do on this test, even if they now live a life of a non-biased person?  Speculation  for sure; but I think ultimately logical.

Which dovetails to the third basic critique, brought to us by basic ideas in critical thinking. 

Nominally, one of the things a critical thinker must do in order to minimize bad decisions, is to understand their own biases.

& in this case, I don’t mean to limit this to racial biases brought about through family or societal pressures, but consider the word bias as more encompassing to include any consistent mode of thinking which negatively affects decisions.

From that point of view, the biases one should look for in critical thinking outside personal life, are biases shared by the entire human population.  Brought about through evolutionary pressures over roughly 250K years, these biases include:

our pattern seeking behavior (good skill, but dangerous when over used)

our filtering mechanisms (necessary, but poorly understood)

our ability to pick even bad answers to open questions just because good answers don’t exist (still prevalent today)

group think (tribal mentality), and on and on and on

The point here isn’t to hash out all possible biases, but to explain that knowledge of such biases is only a first step to mitigating them.  Knowing this information isn’t a cure; it only gives one the best chance in mitigation (assuming objective introspection). 

Outside of these critiques, several studies demonstrate problems with IAT’s usefulness. 

One study questioning IAT found (here via University of Virginia Law School):

We reanalyze data from two influential studies—McConnell & Leibold (2001) and Ziegert & Hanges (2005)—that explore links between implicit bias and discriminatory behavior and that have been invoked to support strong claims about the predictive validity of the Implicit Association Test (IAT). In both of these studies, the inclusion of race IAT scores in regression models reduced prediction errors by only tiny amounts and IAT scores did not permit prediction of individual-level behaviors. Furthermore, the results were not robust when the impact of rater reliability, statistical specifications and/or outliers were taken into account, and reanalysis of McConnell & Leibold (2001) revealed a pattern of behavior consistent with a pro-Black behavioral bias, rather than the anti-Black bias suggested in the original study….

Reworded means the study failed to predict individual behavior, even though it’s theoretically proving individual biases.  Additionally, when they evaluated the tests with reference to the raters (rater reliability) they found the degree of subjectivity to be so high that a simple change in raters could result in the complete opposite conclusion.

The second study found questioning the specific results of an IAT test with reference to physicians’ biases noted (here via SpringerLink):

 …the main study results were based on an interpretation of crosssectional data as if the data were longitudinal (see Fig. 1) [D]; 3) although  andomization was performed by allocating a white or black face randomly with each scenario, a non-randomized variable (IAT score) was interpreted as if it had been the unit of randomization [E]; 4) Green et al. conflated measurement issues with interpretation; the progression from the Introduction through the Methods to the Results section of the terms “increasing time for association,” “racial preference,” “racial bias,” to “pro-white/pro-black scores” does not represent a sequence of synonyms…

 And lastly, a professor with interest on this topic points to several studies questioning IAT’s effectiveness, but specifically this one which shows people encountering unfamiliar stimuli can test as biased (here via University of Washington):

…Brendl, Markman, and Messner (2001) is an example of a study that used totally unfamiliar stimuli, intending these to represent pseudo-categories that should lack associations with other categories. Greenwald and Nosek (2001) concluded that such uses of totally unfamiliar stimuli would yield problematic IAT findings and should be avoided….

So oddly enough, what started as an article to point people towards what is still likely to be an informative debate, has turned into an example of bias in science. 

This wasn’t the initial point, but given a society prone to instantly believe things which are thought to have consensus among scientists, a very necessary point to make.

& for those in the sciences, or like me, like to study human behavior, it’s a nice reminder that we should always be careful how much faith we put into any “consensus”, whether it’s a religious/moral consensus, or one by scientists.

A must read:  Cargo Cult Science (via Lock Haven University): Richard Feynman’s Caltech commencement address given in 1974, expressing this same philosophical idea, but with more eloquence and brevity than I have.

The paper extends research on fixed-pie perceptions by suggesting that disputants may prefer proposals that are perceived to be equally attractive to both parties (i.e., balanced) rather than one-sided, because balanced agreements are seen as more likely to be successfully implemented. We test our predictions using data on Israeli support for the Geneva Accords, an agreement for a two state solution negotiated by unofficial delegations of Israel and the Palestinian Authority in 2003. The results demonstrate that Israelis are more likely to support agreements that are seen favorably by other Israelis, but — contrary to fixed-pie predictions — Israeli support for the accords does not diminish simply because a majority of Palestinians favors (rather than opposes) the accords. We show that implementation concerns create a demand among Israelis for balance in the degree to which each side favors (or opposes) the agreement. The effect of balance is noteworthy in that it creates considerable support for proposals even when a majority of Israelis and Palestinians OPPOSE the deal.

Or restated, it seems our sense of justice prefers balanced deals over seemingly one-sided deals, even if the balanced deal is opposed by the majorities of both interest groups.  Seems intuitive.