Empirical and rational reality

[HydrogenBond]

I am not a big fan of empiricm, although I can see its value in many walks of life, such as factories and quality control. I tried to think of an example of empiricsm leading to social conclusions out of touch reality. Here are two.

 

Picture the head of a youth organization wanting to buy shoes for the kids. So it gets a empirist to do this study by the book, so he can save money. There are twenty kids, 10 take size 10, and 10 take size 8. After the study is complete it is determined that the average size is 9+or-1. Based on this study the leader orders 10 size 9's and 5 size 10's and 5 size 8's.

 

What is sort of messed up is the study predicts an average of 9 even thought nobody wears a 9. Since the study was done with careful scientific scrutiny, the beaurocracy insists the study is correct and 10 students much wear size 9 or they will get no shoes.

 

Like all statistical studies the one-size fits all approach, works well to average large groups of people but runs short of reality when it come on a case by case basis. It makes things too tight for some and too loose for others.

 

It would have been better to go case by case to each youth and collect all their sizes and tailor ordering to what they actually need, instead of lump everyone into an average size that applies to the whole group but not to any particular person.

 

Here is an example of a good study with bad base assumptions. The rainfall in a new area to be monitered, goes through a type of 2-year cycle, 20inches one year and 40 inches the next year. The average is 30inches. The study indicates that for the past 10 years, the rain has either been 10in below average or 10 inches above average. That is sort of true if you meaure from Jan-Jan. But if we normalize it from June-June, it hits the average every single year. One is the land of gloom doom weather forcasting and the other is the land of farmer's delight. Empiricsm allows room for subjectivity.

[HydrogenBond]

The example of buying shoes shows how a statistical study can result in a response that may not reflect the person-by-person needs. In this case, a simpler rational strategy would be better, like doing a case-by-case head count.

 

If we extend this example, with our leader now needing to buy shoes for 100,000 children, the head count may be too cumbersome. This is where a good statistical study, using a more limited data set, would be much more cost affective. But there is no guarentee that the results will correctly fit all the 100,000 chidren. In other words, there are cases where statisitics do come in handy, but there is not guarentee it will hit the same results as doing a case-by-case study. It may order 50,000 size 9's, since this is the average, etc., One needs to take statistical results with a grain of salt inspite of their obvious cost savings when analyzing large numbers of things.

 

Another consideration when it comes to empiricsm is how the data is presented affects how it is interpretted subjectively. The easiest example to see are medical studies. For example, a study has determined that apples increase the rate of some forms of cancer by 3% (hypothetically). One can present this study as apples increasing the cancer risk factor by 3%. One can also present this study by saying, with 97% certainty, apples do not present significant risk of cancer to the population.

 

The first presentation of the empirican data not only raises awareness of potential problems for those at risk, but also increases the anxiety level in those who have no risk factor; statistics don't tell you who is who, so everyone much be aware. The second presentation, 97% risk free is an A+. This would make most people feel that apples do not present any significant problem. On the negative side, it may cause those at risk not to take the needed precautions.

 

One correlation that is well documented is that stress/anxiety increases the risk of health problems. The negative presention has a side affect of increasing the stress level in those not at risk and thereby can actually contribute to secondary negative health affects. The postive presentation of the study, creates a positive side affect. It may cause some not to heed the warning but it will not create the same level of negative side affects due to anxiety.

 

The medical community is not just an altruistic agency concerned with the health of the population. It is also big business. The negative approach is good for both altruism and business. It is a better way to guarenteed that the 3% percent at risk heed warning. It will also lure in hypocondriacs. It will also create global populaiton anxiety that will lead to secondary stress affects that will create additional business.

 

If the medical commnuity presented a positive spin to the data, it may treat less than the ideal 3% which would affect atruism. They would lose the hyponchonriac business and would not generate all the addtional revenue due to secondary anxiety affects. It may even lose business in other areas, due to the power of positive thinking.

 

This dual spin for presenting empirical data is sort of analogous to the example of rainfall in the first post. If you spin it one way, i.e., jan-jan, the rain is always too much or too little, making people worry for nothing. If you spin it from june-june, the rain is always reliable and much less of a concern. Statisitics allows for that wide range of subjective spin, and everything in the middle, since it is not fully rational. It is a precursor to reason and should be recognized as such.

[Erasmus00]

In both cases you are completely misusing statistics. Any competant "investigator" would NEVER use statistics in the case of the shoes. There aren't NEARLY enough data points for statistics to be valid. In the other case, no one competant would average across the cycle. These are ridiculous straw men arguements! You aren't arguing against empiricism, you are arguing against blatant misapplication of statistics. No one competant would make these errors.

-Will

[HydrogenBond]

I agree with you. The reason this wouldn't happen is that the power of reason would intercede before empiricsm was used inappropriately. But when things get too complicated for the power of reason, then statistics may become the only way to massage the data. But this does not guarentee the results will not come out off base.

 

One of the areas where statisitics comes in very handy in in the factory. If we are making widgets and need precise control to assure quality, statistics are very handy. In this case, one size does fit all since, since we are making only one reliable product. But humans are not widgets. Each is designed to be slightly different. To try to force fit them, so they are all the same human widget, is contrary to what they are meant to be.

 

Most studies on human health, treat humans like they are the same. They look for one size fits all, but end up with the reality of human diversity, i.e., exceptions and side affects. Then social policy ignors the human widget defects (better or worse) and tries to force fit the one size fits all approach, which stem from the medical empiricsm.

 

The idea that empiricsm can be presented with optimism or pessimism (glass is half full or the glass is half empty), show that the results of empirical science is vulnerable to human subjectivity. If something is rationally connected, there is no subjective steam. E=MC2; this is not bad or good, it just is. End of argument. If E=MC2 was stuck at empiricsm, people would line up and argue forever.

 

If we are stuck with medical empiricsm, why not give equal billing to the optimistic side of the data? The reason is, that would make it clear that the study is not a rational relationship. It is all the way one looks at it. If we just use the pessimistic approach, we can create the illusion of reason, where none really exists, since there is only one side; glass is half empty, end of story.