Is statistical probablity accurate in Thailand?

[ATF]

I gave a simple example of how using the same methodology will lead to statistics that can not be compared.

Thailiketoo has not answered yet.

Typical.

He did the same thing in a different thread.

Just another statistic?

[thailiketoo]

I gave a simple example of how using the same methodology will lead to statistics that can not be compared.

Thailiketoo has not answered yet.

Typical.

He did the same thing in a different thread.

Just another statistic?

Below is the thread for unemployment. If I don't answer a poster it is because I didn't see him or he is on my ignore lists and I don't see him unless another poster refers to his posts.

http://www.thaivisa.com/forum/topic/737217-thailand-has-one-of-the-lowest-unemployment-rates-globally/page-8#entry8031810

[wolf5370]

I like to ridicule papists. I learnt that from my peers at school. But if I do it on the open forum, I get labelled a bigot.

SC

Your mistake is calling them papists. That's like Henry VIII stuff, no wonder. Try religion in general.

And I think there are some who would draw a distinction between religion and superstition. I don't but some might.

Don't forget the question was "I believe stereotyping is far more accurate than statistics" which is absolute balderdash and a flame before you began to join the pedantic Thai Visa group.

I've nothing against the "pedantic's for a brighter forum future" brigade but some might wonder if it was on topic.

Stereotypes are formed out of historic experience, and have often persisted centuries and held by many people - this will coalesce into a statistic phenomenon themselves (especially of one removes the extremes - and that which has been disproved, such as Blacks cannot learn to read, and so on). There are theories that suggest people from certain groups tend towards their stereotypes. The force of volume of belief can hold weight, but cannot be satisfactorily quantified. Statistical analysis always has the numbers to fall back on, beliefs change over time, circumstances also change over time (rendering both statistical and stereotypical conclusions obsolete), and belief is usually very subjective and easy to disprove as it usually deal with absolutes (i.e. fails the "Actori incumbit probatio" test) rather than odds. I think this shows quite readily that the mathematicians

have it over the bigots

[wolf5370]

Statistical probability would seem to work well where events are random - or at least we lack the ability to accurately model and come up with a reliable deterministic answer.

I can understand why another poster preferred to rely on stereotypes - he probably doesn't have to hand reliable data on which to make a sound statistical estimate of the risks to which he is exposed, and the stereotypes, which may be based on over-reporting of misfortunes and extreme outcomes, may help him err on the safe side.

Personally, I prefer to rely on my own judgement on a case-by-case basis, despite the obvious folly of such an approach.

SC

This is because you are not a computer - we do not work out the odds of everything (at least we don't formally and/or intentionally) at each step through life. We are, however, subconsciously doing just that - based on both what we know from experience, have "learned" through other's experience and media etc, and subliminal/paranoiac fears and discrimination of our own. One would not walk through a creepy forest road at night we do not know, or through he rough end of town just because we have not been murdered for the price of a bottle of gut rot; one does not leave their wallet and phone on the table in a bar when going to the toilet just because we haven't been robbed at that place before.

Believe me, your judgement is NOT on a case-by-case basis with all the will in the world - only babies and toddler can claim that.

[wolf5370]

I thought the "give a toss" was a witty reference to the reliance on coins by statistically minded tossers. Real life is not like tossing a coin. Real life is like calling a coin toss. Why do most people call heads? Should I call Heads as well?

Actually in some coins tails is the safer call - coins are often slightly heavier on the "head" side as there is less carving away of the surface than on the "tails" side. There was a report done for a English old ten pence piece I read years ago that showed there was a tiny, tiny difference and over many thousands of throws, tails should statistically appear slightly more often. This also precludes the possibility of it landing on its edge too (which actually favour the lighter side too - assuming the stamp on the floor solution is applied ).

[wolf5370]

You see I'm still waiting for ATF to come back and admit he was 100% wrong and did not have the vaguest idea what he was talking about.

Stereotypes exist because they are more often correct than not. Statistics try to rationalize the unpredictable. The OP's assumption that when you toss a coin it will usually be a 50-50 chance of being heads or tails is incorrect. Everyone knows there is a much greater chance of it coming up heads.

If statistics were highly accurate we could all make billions by gambling or playing the stock market.

I prefer to trust in facts and you are incorrect with US jobs data etc. because these figures are always revised a month or two after the quarterly figures have been announced.

Statistically speaking Thailand should descend into financial turmoil because there is a military junta running the Country, but we know that will not be the case because we've seen the same thing happen many times before.

I prefer to believe my own experiences to rationalize situations.

Oooh bad example - sorry, fell into the old lotto 6-49 trap. Simplifying back to a coin for a second (and assuming a perfect 1:1 result) the fact that the coin has been heads nine times in a row, does not affect the chance of the 10th toss - it still has two equal sides (in our perfect coin) and thus 50% chance (ignoring edges). Btw the Lotto 6-49 trap is that its is less likely that numbers 1 through 6 will be pulled than any separated and random 6 numbers between 1 and 49 - it seems fair on the surface, but is incorrect - each ball has the same odds as any other for each pulled ball (descending odds per ball of course as they are not returned).

The fact that something has occurred many times before does not mean it will not this time. However, something like national economics (or international economics) is far too complex a subject to accurately predict - that is why forex is still a punt and not a pure science - too many variables and that are ever changing.

Believing your own experiences IS statistics! Sure you may not quantify it formally - and it is weighted (i.e. it only takes one mugging to change your route home from a late night bar for example, but many "hello sexy man" come-ons from lady boys ) - but it is the strength of those experiences by volume multiplied by weighting factor by which you make those rationalizations

[wolf5370]

The TV jinx strikes yet again, anyway to answer post #67.

Ok, to answer your question, the laws of proability remain the same, wherever they are applied.

The exact same as the rules of algebra remain the same, I can give any student the following,

AB=CD, any student of alegbra can define A, B, C or D in algebraic terms.

Exactly as I mentioned before, the chances of an event happening are 1, the chances of it not happening are 0, these laws never change.

The only thing that changes is your sample data.

EG, are my chances of meeting a woman who has been married before and has kids more likely in Soi Cowboy or Thong Lor.

What are the statistical odds a go go dancer in NEP is more likely to be tattooed than a girl who works in the Asia Centre building on Silom.

Yes, that ois what he has been saying since the Op. The maths are immutable (at least until they are improved) - it is the data and representation of the results/conclusion that obfuscates - or the avoidance of following the mathematical process to begin with (e.g. making it up !).

[wolf5370]

QUOTE FROM THAILIKETOO:

Some people think that to get these figures on unemployment, the government uses the number of people collecting unemployment insurance (UI) benefits under state or federal government programs. But some people are still jobless when their benefits run out, and many more are not eligible at all or delay or never apply for benefits. So, quite clearly, UI information cannot be used as a source for complete information on the number of unemployed

In a previous thread, the OP explained how unemployment statistics in different countries are based on the same methodology, surveys, and therefore they can be compared.

Somchai from BKK and Fritz from Berlin are unemployed and surveyed.

Unemployment in Th: 1 /////// Unemployment in DE: 1

Somchai and Fritz decide to do something about their situation.

And what a coincidence, they both get surveyed again.

Fritz wanted to sell bratwurst in the streets of Berlin, but does not meet the 253 legal requirements, therefore unemployment in DE remains 1.

Somchai has set up a noodle shop, next to 253 other noodle shops, and manages to sell 3 portions of noodles each day, therefore unemployment in TH becomes 0.

To the OP: using the same methodology in completely different countries does NOT lead to comparable statistics.

In this particular case, statistics were wrong because the "invisible" or "hidden" unemployment is different in TH and DE.

This is all semantics! No the statistics are not wrong (if the mathematical process is adhered to), the statistical STUDY is non comparative. That is semantics, but it is true also. A report comparing or combining the results would be flawed and incorrect - but that does not mean the statistical methodology (the mathematic process) is also flawed.

[wolf5370]

I thought the "give a toss" was a witty reference to the reliance on coins by statistically minded tossers. Real life is not like tossing a coin. Real life is like calling a coin toss. Why do most people call heads? Should I call Heads as well?

You didn't like the seismic survey one eh? Or maybe you want another 100 examples of how you depend on statistics in your daily life?

What is the probability that many on here are involved in seismic data acquistion and are more likely to live in Pattaya or the ESB than say Issan or Surin?

Trying to distort and misdirect the thread? Why? It is an interesting topic. The great unwashed and uneducated among us know little of the important role statistics play in our daily lives. Perhaps someone would have learned something. ATF might learn how the unemployment rate is computed in the USA, Australia and Thailand. He might have learned how every medication he has ever taken has benefited by the statistical process. He might have learned that oil is found based on probabilities. Or maybe he will just close his eyes and say Statistics are dumb and I don't want to learn anything that i didn't know when I was 7 years old.

How about how every time we use an ATM card, Debit/Credit Card or Internet Banking or even log into our computers/email/use SSL - the encryption methods used rely on probability of hash collisions and brute force algorithm solutions - also pseudo random number generators used in ciphers (like block cyphers) are based directly on probability - Indeed the birthday problem mentioned earlier is actually the basis for the cypher methodology (to reduce distinguishable output revealing the algorithm over iterations/time - this is why we no longer use DES but AES and beyond - 64 bit to 128 to 256 etc as it is quadratically degradative). Too complex to go into here (see wiki http://en.wikipedia.org/wiki/Block_cipher if you are interested but be warned it get heavy).

//Edit: Typo

Edited June 29, 2014 by wolf5370

[wolf5370]

I don't know why some of you guys are giving the Op such a hard time. His truism is self evident: Mathematics does not lie, people manipulate the data or selectively bias the result to reach the desired conclusions.

1+1=2 regardless; we can corrupt that by saying such as a man and a women are unemployed, but are married to each other - therefore "we" can say the "wife" is a "house wife" (which is a job) and thus we up the unemployment figures by 1 not 2 - regardless of both of them claiming unemployment and/or other such benefits (this actually happened in the UK under New Labour in the early nineties to "reduce" the high unemployment rate!).

Pretty much everyone has agreed on that, so what's the problem?

[JLCrab]

Wikipedia -- brought to you by Jiminy Cricket & Thai - Like -Two W-I-K-I-P-E-D-I-A

http://www.youtube.com/watch?feature=player_detailpage&v=cy2jWJtO3lE#t=29

Edited June 29, 2014 by JLCrab

[thailiketoo]

No you are wrong, very wrong. Read.....Early each month, the Bureau of Labor Statistics (BLS) of the U.S. Department of Labor announces the total number of employed and unemployed people in the United States for the previous month, along with many characteristics about them. These figures, particularly the unemployment rate—which tells you the percentage of the labor force that is unemployed—receive wide coverage in the media.

It is impractical to count every unemployed person each month, the government conducts a monthly survey called the Current Population Survey (CPS) to measure the extent of unemployment in the country. The CPS has been conducted in the United States every month since 1940.

There are about 60,000 eligible households in the sample for this survey.

A sample is not a total count, and the survey may not produce the same results that would be obtained from interviewing the entire population. But the chances are 90 out of 100 that the monthly estimate of unemployment from the sample is within about 300,000 of the figure obtainable from a total census.

http://www.bls.gov/cps/cps_htgm.htm

I'm sorry you can't accept reality and see just how wrong you are.

Every facet of your life is controlled by a statistical process from the medications you take to the food you eat to the gas that powers your transportation. Sorry you choose not to see it. I can't help any more. I gave you the quotes from the US government and the link. If you choose to think they are lying to you so be it.

Every payout an unemployed person receives has to be accounted for. The only way to do this is with computers otherwise you could be claiming benefits in 50 different states. Plus the state and government has to know how much it is spending.

P.S. I never said statistics were dumb I inferred they were flawed and far from perfect.

So? The unemployment rate is not determined by the claimant method in the USA, Australia or Thailand. There in lies the problem. ATF will not believe this. ATF wrote, "In America the unemployment rate is easily determinable because they only need to count the people claiming benefit. All the rest will be working legally, illegally."

Well, No. America has not figured the unemployment rate using the claimant method since 1940.

ATF can't admit error. Easy to check. Look at the link below.

http://www.bls.gov/cps/cps_htgm.htm

Wolf 5370 wrote, "I don't know why some of you guys are giving the Op such a hard time. His truism is self evident: Mathematics does not lie, people manipulate the data or selectively bias the result to reach the desired conclusions."

ATF above is an example of the problem. Many people don't know just how many things are not counted anymore. Many, many things are done buy survey . So I have been given a hard time because I chose to dispel this myth of counting.

[ATF]

So? The unemployment rate is not determined by the claimant method in the USA, Australia or Thailand. There in lies the problem. ATF will not believe this. ATF wrote, "In America the unemployment rate is easily determinable because they only need to count the people claiming benefit. All the rest will be working legally, illegally."

Well, No. America has not figured the unemployment rate using the claimant method since 1940.

ATF can't admit error. Easy to check. Look at the link below.

http://www.bls.gov/cps/cps_htgm.htm

Wolf 5370 wrote, "I don't know why some of you guys are giving the Op such a hard time. His truism is self evident: Mathematics does not lie, people manipulate the data or selectively bias the result to reach the desired conclusions."

ATF above is an example of the problem. Many people don't know just how many things are not counted anymore. Many, many things are done buy survey . So I have been given a hard time because I chose to dispel this myth of counting.

TLT I will admit I was wrong because I studied the methods in detail last night and I must say that I was shocked to find out that such an antiquated method blatantly open to corruption and manipulation exists in the age of the computer.

I don't think anyone is giving the OP a hard time we are just having a normal debate.

[thailiketoo]

How can you tell if people lie? Sample a number of unrelated groups. In 2012 the US spent 50 million taking surveys about the election.

http://www.gallup.com/poll/9442/election-polls-accuracy-record-presidential-elections.aspx

Is it more accurate to count the unemployed or take a survey? It is more accurate to take a survey.

Not everyone will lie about the same things so sample more groups and the lie disappears.

How many people work on the farm and how many people work in a factory. Sample enough different groups and you have accuracy. Somchai may lie to me in one Soi of Surin but not the same lie with 10 Somchai's in 10 different Sois.

Survey researchers use specialized jargon to convey the reliability of survey results. Terms such as “statistically significant,” “margin of error,” and “confidence level” help describe and make inferences when analyzing data. These terms are used to extend limited poll results to the larger population.

The underlying principle which assures polling accuracy is that those interviewed are representative of the target population and that each individual in the population has an equal chance of being selected for interview. This is the science of sampling technique, and when using a scientific sample, a relatively small number of individuals can be used to project to the target population.

Sampling error” (or conversely, “sample precision”) refers to the amount of variation likely to exist between a sample result and the actual population. A stated “confidence level” qualifies a statistical statement by expressing the probability (usually 95-99%) that the observed result cannot be explained by sampling error alone.

http://ivn.us/2014/01/08/statistics-dont-lie-people-science-political-polling/