Thursday, December 08, 2011

D A T A - G A T H E R I N G

People don't know what to do with a lot of the data they are spoon fed by the right, if the right even has any data at all. First of all they don't isolate the independant variables. For instance people who smoke may well be the people who eat lots of fries, KFC and fried onion rings. Hence if the heart attack rate is higher you don't know which thing is causing it. If more blue collar workers smoke than white collar it may be the same group that breathes cold miner dust or asbestos or metal shavings or chemical toxins. So they have higher cancer rates. People in poor rural area may have lower cancer rates because they can't get to hospitals and may die from some flook accident like having a limb sheared off by farm equipment and bleeding to death. People who are rich are more likely to be married in a church and marry at a later age than poor people. Rich people have family members to bail them out at times of monetery extremity and are also helpful in getting an employment leg up. Contrary to the myth, I've heard rich people really ARE "happier" than poor people. Rich people who "fall" into sin are seen not as examples of shame, as are the poor, but as potential targets to recruit into your Church to increase your status. Rich people are better educated than poor people. If more fundamentalist Christians are from suburbia than other factors than Faith may shape what we know to be the ingrediants for a healthy family. (Selah) Well let's get generic for a bit and say that there are four "Movements" or of aspects of interest to statistical charts and we will go down them one by one and explain the relivence of then to you, the reader. First of all there is the Skew. I always used to spell it Skiew but now there is spell-check. Many people regard Skew is a synonym of bias, either psychological or political. It refers to an undue emphasis of certain data or other data to suit the psychological mindset of the one shelling out the statistics. To get strictly narrow and generic in our definition a skew refers to physical space. There is of course "psychological space" in our minds to store information. And we apportion more of our space to certain kinds of information and less significance or "space" in our minds to information we may deem irrelivent or not germain to the issue at hand. In a physical realm we portray it metaphorically to "slanting" the article or information either to the left or to the right. The idea is that even though the information contains a bias, given the proper "decoding" or weighting, the information can be restored to objectivity. The trouble is that the Tea Party dishes out information with such a minute Truth quotient that there is virtually NO truth in it. This information cannot be made safe for human consumption. As Ted Kennedy once said "They spout lie after lie after lie after lie". So were you to see it graphically it would ammount to no more than straight lines on a piece of paper, in other words- unintelligable. Bit skewing in statistics charts refers to a bias tword the left or the right of the chart where the graph is not symmetrical from the node or peak incidence point of the graph. Just to simplify this a bit, the node is usually close to the median. Where you may see skewing is on intelligence tests that do not test from a full spectrum of the population but where a certain segment, perhaps the dumb people, has been filtered out already. But more to the point, a skew takes place in such things is income of Americans. You can't be any poorer than no income yet to the right the chart can veer off into the millions and even billions. What you see in such a case is that the mean or agragate is skewed to the right, the way Republicans like it, from the median income or "average American household". This is the figure sociologists regard as the most telling because it means half the people are above it and half the people are below it. Most likely the median will be close to the node, or highest data incidence, and the mean will represent the AREA under the data line, which might be shaded, were one to divide it geometrically into equal area halves. Just so you know, all four terms are node, skew, standard deviation, and kurtosis. Let's move on to standard deviation. This is different with every data sample. It is NOT standardized, but rather a derived figure from the data at hand. It represents the center two thirds of the data by "area" under the data line. That a range is broad or narrow is not particularly significant. Certain types of data have more broadly occuring numbers than other realms of data.

The final thing is a statistic I have my own unique formula for. What it basically measures is now MUCH the data has dropped by the time it reaches the "hash marks" of the Standard Deviation zone. It's as though you squeezed it at this point. You do not lose any date but the data would go elsewhere, either to the top of the curve or off to the points. Keep this thought picture in mind as we explain further. We now move into the final and most contraversial category, which is kurtosis or "pointedness" of the chart or graph. Now I have a means of computing kurtosis which lands you so close to the "normal" point of three, to their far more complicated method, I consider mind valid because "It cannot be shown to be in error to any scientifically measurable significance". My reading varies from wolframalpha.com in that theirs can never fall below the square root of three. At this same figure my reading would be represented as a One. If you knew my method of computation you'd see why One is the only logical figure. Personally I would say that three as normal is too high. A kurtosis of two on my scale is perhaps too low. I'm pretty sure a two kurtosis on my scale can be derived simply by "quartering" the perimiter of a circle and rearranging the quarters so that they form a bell curve. My kurtosis reading is calculated simply by taking the ratio of the incidence of data at the node or top of the curve, and comparing it to the incidence of data at the SD line or edge of Standard Deviation. At the square root of three ratio- - mine would be three. The square root of two would rate mine a Two. A flat liner would rate mine a One. Conceivably it might be possible to dip below a one. A low kurtosis figure would denote an insufficient amount or not big enough sample of data. I would tollerate kurtosis in the two's because in my book that's how the data would pan out. It's generally understood that as the data sample rises, the kurtosis would rise and presumably even reach three. Perhaps my own reckoning is for sample size to not be overwhelming. This means the peak in the center would be lower and the edges would be lower and the data would assume more of a Square shape. A kurtosis too high would be one of two things as I see it. Either you are dealing with an overly filtered sample. Like the SD curve this might be a police department that routinely filtered people out if they were too tall or too short. But a more interesting explanation would be incorrect Testing procedure. If you tested for "too broad a band" of Intelligence, you would have a scenario where the population all tested the same because so many questions were aimed either at super geniouses, or else blithering idiots, with not that much definitive shading to short out the middle. Press one to hear message again.

HOW TO FILTER OUT THE CRAP YOU DON'T WANT OR NEED

The first part of this paragraph is a complicated graphic demonstration of a psychological truth. Skip over it if you must. You know it's an electronic reality as far as I can determine that FM signals won't show up on an AM tuner and AM signals won't show up on an FM tuner. To illustrate this if you go to the sound band on an analog tuner the screen will turn black. This is where the FM sound track is that AM can't see. On these pay TV stations they used to have, there IS no sound. TV tuners are such as they listen for FM sound. AM won't show up because AM carriers are not modulated to swing by frequency and FM carriers are not calibrated to register changes in amptitude. Your belief that the Sound track on Pay TV is AM because what was normally black is now seen as psychedelic waves. The picture lines are scrambled with the classic DES RHO pattern, which is a coding system I sometimes use to scramble words. If your name is Randy - - or Ron, you should be able to glean what I'm talking about. Also it is my contention that the "median" of the six M Hz band has been "shifted to the left" so that now the signal on the right side of said median how is seen in full and not surpressed by the TV. Analog TV unlike AM radio broadcasts in single sideband, which is kind of akin to dingle diode verses double diode or half wave or full wave DC wave restoration. When the two waves ever compete with each other you get whigged out shading as the two fight each other. The bottom line is in Pay TV you see a negative picture with a white line down the middle, with no sound.

WHICH BRINGS TO PSYCHOLOGICAL TRUTHS Many people are swayed by emotion and not their higher logic. Logic won't register with them but only Emotion will. For a Logical person like Spock, emotion will not register, only Logic will. Hence when a TV evangelist delivers his message, particularly is you are actually in the hall, then you are swayed by Emotion. To turn James Dobson's words back on him about "Your feelings will Lie to you" then when your head is spinning with a lot of new emotions this is a BAD TIME to make any major decisions about your life. The best thing you can do it give it a rest and let the emotions clear. The Evangelist knows if you think about things you'll never bite at the hook, so he will offer you every Lure not to postpone a decision. Now lets go to Obama's speech. Most of us on the left believe that people who work hard for a living ought to not pay more taxes than those who don't, but "wait around the pool for the dividend check to arrive". This whole thing about "rewarding hard work" and "giving people a chance to get out of the hole" seems to NOT resonate with those on the tea party right. These people on the right like to find somebody, anybody than they are better than. And if they can't fine anybody else they will say they are better than themselves, as they were before they got Saved. Starr Parker likes to brag about how she had five abortions and used to lay out on the beach waiting for the welfare check to arrive. By what stretch of logic is that supposed to make me she is a moral authority on Anything? I wouldn't doubt it if you said Tammy Bruce had become a Born Again Christian by now because she has moved so far to the right, politically. People like me see patterns in things and when something doesn't "fit", I for one don't rest till I figure out what is wrong - - - what the problem is. If more people on the right did this- - perhaps more of them would be willing to do more economic detective work. They seem to not want to even think about cause and effect. They would flunk out the first day of District Attorney school and wouldn't make it as police detective either. I myself faced a mental "cunundrum" a while back. We know if you have three dots that form a line that chances are that line goes out straight and will stay that way. Two points can mean anything but with three they form a pattern. So I was perplexed by my own, ONE, TWO, THREE rule of Sine waves. Perhaps I've mentioned it already. That is- - the Sine of 30 degrees is the half the square root of One, and the Sine of 45 degrees is half the square root of two. And I believed that the Sine of 60 was half the square root of three. But I was wrong. The TANGENT of 60 is the square root of three, and I know that now. I learned something. For a long time I resisted the idea that I needed a new computer hard drive telling people "No, it's memory and execution that is clogged up - not disk space". But then I was informed that a computer USES disk space for Swap files to swap in and out of memory and if it can't find a space on the hard drive quickly it will slow down the computer. It's these cause and effect things that people need to see. I will even concede a Small point to the Republicans where they Might be right. If we want Republicans to invest in business - - - a lower tax rate would encourage this and we could still be revenue Positive if we didn't lower it too far. We could lower the rate from 35 to 25 percent for example. But MORE important by FAR if we do this, is to raise the income and dividend tax rate much higher, like to at least 35 percent and more would be better. This would Discourage rich people from NOT investing, but instead squereling their money away in some account and pay only 15 percent on it. If we nabbed the rich on the loop holes and perk expenses they can write off - - -we would encourage a thriftier rich class, like Mr. Potter claims is a virtue, who didn't squander their money put put it into productive venues. We would raise money, and also encourage legitimate business investment, and just perhaps make us more competetive with foreign countries than we are now. Even a liberal can learn. If we are not blinded by our own biases we can make decisions that are better for ourselves and the people around us. Jesus said to take the sawdust out of your own eye before you venture to take the speck out of your brother's. But once you DO get rid of your own biases, if they are "no longer registering as relivent" then you CAN venture to help your brother and in the course both lives are improved.

No comments: