Tuesday, March 20, 2012

Scientific Method, Bias and Politics

Science is distinct from religion and philosophy.  In religion one begins from the position of received 'truth'.  God speaks to Moses and tells him to tell the Pharaoh, "let my people go".  Beginning with God's word the outcome is a product of the level of obedience to the projected plan by the individual.  In philosophy one arrives at truth through reasoning, predominantly deductive or inductive.
In science, truth is not known, but an hypothesis is assumed. An hypothesis is an unproven, but an experiment is devised to prove the hypothesis, thereby arriving at 'not absolute truth' but rather a "prevailing hypothesis".
In religion increasing truth about the world comes from God's sharing this information but in science it comes from human endeavour, that is the accumulation of 'proofs' regarding some aspect of creation. Whereas religion and philosophy delve into the greater meaning of the universe, the why's and wherefore's of creation, science is predominantly interested in the how.
So an hypothesis might well be that the light always returns after darkness.  A study is designed where two people wait through the night and sure enough the light comes with the sun in the morning.  Eventually the hypothesis that explains this and suggests this is prevailing experience is the astronomical recognition with aid of telescopes and other paraphernalia that the earth revolves around the sun.  Even watching the moon in its orbit around the earth one can see in eclipses the lightness and darkness confirming the original idea that the light follows the dark.
Experimentation is therefore central to science and why science students have 'lab' assignments in addition to regular class assignments.  In the arts a literature student doesn't require an extra lab class doing experiments with explosives or highly dangerous chemicals. A literature student will listen in class, hear ideas and go away, read books, discuss the ideas and write papers about the ideas.  In science classes the students are required to do the original critical experiments that defined the accumulated knowledge that forms the prevailing hypothesis in that field.
In scientific experimentation to better arrived at truth there are a number of variables which are identified and limitted.  These 'variables' , aspects of experimental design and work which can influence an experiment unfairly, are generally limitted in elegant experiment because they introduce bias into the experiment. Experiments are like a game in that there are usually two principal outcomes, either the experiment prooves your hypothesis right or it proves your hypothesis wrong. If you are right you get grant money and advancement whereas if you are wrong you can be advised to get an arts degree or perhaps physical education.
Scientists are competitive and science is a competition in this sense.
The first bias to consider is 'selection bias'. In social sciences if you wanted to say "all men are good dancers" you could unfairly prove your 'hypothesis' true by choosing to select only from Negro and Latino men preferably from US inner city. If on the other hand you chose 'white male judges' you'd find that your hypothesis was wrong since everyone knows 'white male judges' can't dance whereas Latino and Black men can.  Indeed, this fact alone, to scientists has often explained the 'legal' bias against Black and Latino men by the predominantly white judicial system. However without funding the experiments haven't been done to confirm this obvious example of how envy influences the judicial system scientifically.
Random selection is the way one avoids such obvious bias.  Therefore by random selection you stand on a street corner in New York and observe the rythym of every second male regardless of their race thereby removing the 'selection bias' component.
Commonly experiments have at least two study groups. Often in biological sciences the study group is considered the group that something is being done to whereas the 'control' group is the group that is being left wild.  In this way, if I wanted to study the sexual behaviour of women under the influence of alcohol, my hypothesis being that alcohol makes women women more sexual I'd have my control group drink coffee in a bar one night whereas I'd subject my study group of women to the influence of alcohol.   My experimental design might be the more sexual encounters in 24 hours is the criteria for proof in my experiment.
If I wanted to make the experiment even more 'stringent' I'd introduce "blind" factors. Blinding in scientific method is causing either experimenter or those in the experiment not to know what is being done. In 'double' blind experiments the subject and the experiment is unaware of who is the control and who is the experimental group.  In drug research commonly 'blinding' is done with 'placebo pills' such as sugar pills.  So one group of students is given LSD and another is given placebo. The experiment design is set up to study whether students on LSD act abnormally. The experimenters rating the behaviour don't know which group took the LSD and the students themselves don't.
Other biases in research include "effects", such as the "halo effects'.  In physics it was found that electrons being looked at acted differently from electrons not being looked at. This was the same in social sciences where people who are in an 'experiment' act differently than those not in an experiment.
Alot of 'field work' in anthropology and zoology was developed because in social sciences the very 'context' of the experiment had a biasing effect.  Observation in the 'wild' showed that animals for instance acted different than animals in the zoo the same way that humans in night clubs act differently from the same humans in church.
An example of this was a colleague who was so uptight he made a normal Austrian gestapo look left of Democrat. He wore his white lab coat pressed and cleaned each day and had his hair brushed exquisitely.  He carried a clip board and didn't talk to people unless he thought they were serious. No one knew this man to laugh and when his sense of humor was questioned he denied the claim insisting that he had a sense of humor most angrily.  He did experiments on students showing that LSD caused paranoia by sitting in a locked white room with student who had taken LSD and observing the student while avoiding interacting with him. The subjects became paranoid in his presence and he failed to consider that his own paranoia was the cause of their associative paranoia however his published results made him considerable amounts of money. Another researcher observed students at a frat party taking LSD and did notice that a subset of LSD users became paranoid while others became ecstatic.  When he reported his results the original researcher began a political campaign to have him discreditted.
Scientific method with experimentation is a means whereby a study design is set up much like a cookbook recipe with the outcome predicted based on the study. Using the rigors of scientific methods, a scientist knows that if he adds the right ingredients as say Newton or Einstein, he will most likely get chilli rather than curry  Subsequently the study is repeated enough times to establish it's 'reliability'. Then it's subjected to analysis to assess 'inter rater reliability'.  Over time a 'body' of evidence is achieved whereby a 'theory' becomes established.
Gravity is one such theory.
In social sciences the idea that the disparate complex nature of social behaviour can be simplified in some instances by the recognition that predominant motivational factors for men and women are sex and power.  With out this knowledge an arts student with a very beautiful wife and a whole lot of money might not understand the potential behaviour of his somewhat poor single ugly weazel like neighbour who works all day in his laboratory making a bomb.  The first man having some general knowledge of theories concerning human behaviour might well show some concern when the neighbour scientist invites him to see how his bomb works.
Bias is the inclination to hold a partial perspective at the expense of alternatives which might be equally valid is what science attempts to exclude to get a closer look at truth.
However, something as simple as 'funding bias' will politically grossly effect science. This is well understood in science and talked about all the time however in the courts even the suggestion of 'funding bias' can cause one to be held in 'contempt of court' since questioning 'funding bias' is indeed 'fighting city hall'.  In law one can't question funding bias but in politics it's what the media does however it was the 'funding bias' of the Nazis that was said to have affected the ethics of the Pope early in WWII when he didn't question the politics of Hitler.
Today if a person develops an experiment to 'disprove'  the 'prevailing hypothesis' that 'global warming is bad for the planet" and especially bad for "rich powerful humans living in previous temperate climates now faced with real estate price affecting climate change", since the universities are funded predominantly by people who have an invested interest in keeping their real estate prices high, they might well not get any funding for such an experiment.
This was seen all the time in research on marijuana when the US was on the  mission for police acquisition of tanks and helicopters called 'the war on drugs', all research that said 'marijuana might have beneficial medicinal effects' went unfunded. Indeed a researcher suggesting such a matter of research might well find himself in jail or at very least refused tenure and kept under surveillance for years. In contrast if a researcher found marijuana was unhealthy for budgigars, because the second hand smoke affected the quality of their bird song as judged by several choir masters and symphony conductors this researcher was likely get untold millions in grant money and likely become the next president of the university.
Politics, economics, and legal considerations are all such principal biases to day that grossly effect science.  Science was once a 'search for truth' in a rather intellectual and joyful pursuit principally by very wealthy bored white folk like Edison and Jefferson.  Increasingly  science seems  predominantly 'applied science' where the proposed use of the scientific product is already decided. Scientists are designated to 'reverse engineer' a product that arrives at the political or even religious or legal truth.  This is no longer science aimed at finding 'truth' but in fact science held hostage by bullies with guns in fact.  The question is whether this hasn't been the case most often.  Da Vinci is known for his air machines but not as well known as the man who developed the best armaments of the day. Indeed it seems that the day job of most scientists is working for the 'man' and only in the night job can they practice their 'science'. It's much like musicians who put out elevator music cover songs in their day job and write their own experimental music alone at night.
Some might indeed argue that all is politics. Then the field  "political science" would really just be a euphemism better named 'scientific politics' as in 'scientific warfare'.
As an example I would suggest that noone today is being funded to produce a gun that will make the shooter and the one shot fall in love and want to never kill humans again. No private industry or university is conducting experiments in this or funding researchers to make love guns.  Just an example. I'll believe science is once again free from politics, religion and law when I learn that a 'cupid gun' is in the works, especially if it is going to be provided as 'freeware'. But then I'd know that it must be something else, conspiracy theories being so seductive.
In reading research on the internet I'm amazed at how the untrained are routinely being duped by what is usually called 'pseudoscience'.
An example of questionable scientific  'validity' would  be  my saying 'buy black beans, I'm a scientist, and black beans are my preferred beans for quality farts'.  This is 'anecdotal evidence' and doesn't involve the traditional experimentation with groups and hypothesis and inter rater reliability. However because I am a 'scientist' and you don't know that 'black bean producers of Canada' is giving me awesome kick backs for my promotional books, then you will no doubt believe that there is 'scientific proof that' black beans are superior to brown beans or white beans in the production of 'quality' farts. All manner of 'celebrities' have been used to promote all manner of things when a celebrities usual only superior trait is the achievement of 'celebrity' and 'celebrity' status.
Indeed you might well not even know that the word 'quality' is a political and not a 'scientific' words. There are such unscientific 'words'.  I personally think my farts smell best so as a measurer of quality of product I will be 'biased' in this regard as each human might and my dog for sure is.
The study of science is the basis for living in what is today a scientific world.  Understanding the 'arguments' about products and people and ideas and political platforms is indeed difficult and precarious without  scientific training.  As one of countless examples marketting agencies have duped the world into believing that many products are 'organic' without realizing that an unholy alliance between marketting spin doctors lawyers and mass food producers came up with the most miniumum criteria for the labelling of a food 'organic'. Even the Canadian Food Certification system which is one of the world's best scientific studies of food source and provision isn't widely understood or known by the average shopper who is commonly duped by the word 'organic' when they should instead be looking for the Canadian Food Certification label.
Indeed if I were to produce a mass destructive instrument today and name it as was done in the 60's and 70's I would n't call it just the 'peace missile' . Instead I'd call it an "organic peace missile' and put some yogurt in it guaranteeing I'd get at least a million or more votes in favour of this overr my competitors 'killer flying bomb' applied science product. Indeed listening to my marketting agency I would have got the funding for the production of my 'organic peace missile' containing yoghurt by insisting that it would be 'guaranteed' to 'prevent global warming'.
It is therefore important to study 'biases' in 'scientific experiments', understand how science and religion and politics and philosophy are distinctive, understand that 'fallacies' in logic are similiar to the 'effects' that can influence experiments and really understand economics and funding to comprehend the limits of the telling of the 'truth' by 'science'.


No comments: