flaws of empirical data

[politicomind]

I recently debate without someone that relied excessively upon the need for empirical data to be convinced of even the most common sensical things:

Here are some instances when empirical data has been used fraudulently or erroneously:

1. In the two years prior to the Atlanta Olympic Games the police managed to erase 22,000 violent crimes and property crimes to make atlanta seem like a suitable place to hold the olympics.

2. In a town north of pittsburgh during their bid to get a walmart, the police were instructed not to arrest small drug offenses only large dealers, because ten arrests of one ounce is still only ten ounces, but one large dealer may be an arrest of fifty ounces. Yet the small arrests looks like there are ten times more drug offenses than the single large drug arrest because ounces are not counted in statistics. The small town north of pittsburgh finally got its walmart but when walmart needed to hire 200 people only 110 could pass the drug test. But walmart was already dug in so it was too late for them to back out of the deal. The walmart had such a problem finding and keeping people that it discontinued its drug testing as a requirement for employment and now only uses drug testing in case of an injury. This practice of only arresting the large dealers was recently pointed out on "the wire" but for a different motivation. There motivation was to not waste TIME on the small dealers.

3. "the unhappy housewives survey" of the seventies was ridiculed because it was a mail in survey which asked housewives to estimate the quality of the marriage. Yet, the outcome was the only unhappy women responded to the survey. Yet, the conductor of the survey didn't use common sense that a disproportionate amount of women who were in unhappy marriages would respond, while the happy women wouldn't waste their time because they would merely write: Yup, I'm happy and there would be nothing left to do with the survey which involved questions like, "how many times do you think your husband cheated on you." "how many times have you cheated on your husband." The survey creators should have realized that happy marriage people would answer zero to all of these and then would not want to waste time filling out the survey. IN the mid seventies the survey was broadcast on the news and then a few weeks later was ridiculed for having poor methodology.

4. A non-profit research group was slapped on the wrist for manufacturing surveys and results that they never conducted. They merely published the survey resutls as if they had asked a thousand people, but they didn't ask any one. It was the common sense suspicion of the people who read the results and thought "that doesn't sound right." When asked to get their data, the phoenix non-profit failed to provide any methodological data. Because it falls under freedom of speech, they could not be punished.

5. A girl that I went to college worked for a VERY REPUTABLE LARGE stock market research website. But the websited lost historical data on many of its largest companies. it was her job to go day by day week by week over a statistical span of twelve years, from 1990 to 2002 and merely make up the daily stock quote numbers of over thirty major top name stocks.

this is what she did all day monday stock high 19.5 stock low 18.7 tuesday stock high 20.1 stock low 19.3. PURE FABRICATION. She simply invented numbers from 1990-2002.

Now you can realize the damage that this could cause, because people are taught to look at a stocks past trends to see how it will fare in the future. So if a stock looked erratic in the past, it would be erratic in the future. But not if a college intern is making up the former stock quotes. I truly wonder how many people made bad decisions off of her fabricating.

6. The daily show in 2004 made a joking statement in response to finding out that almost 40% of people said they got there news from comedy news shows. and almost word for word john daily said: DON'T DO THAT PEOPLE. WE'RE MAKING THIS STUFF UP. In response to his book America which is full of mimic charts graphs and surveys like "Do you want to answer this survey" (YES NO NOT SURE) critics actually claimed that some of his surveys and data looked made up. John Daily responded, of course it is... its a comedy book. The point is that people are so anxious to accept data, that they even accept the joking data.

7. In Cuyahoga Falls, Ohio, 65,000 votes were counted, and of course they were in favor of GW, because .... only 55 000 people live in Cuyahoga Falls. This story was on the Keith Oberman and surprisingly nothing was done. Even election data is made up. 10 thousand more people vote, than are alive in cuyahoga falls. Not to name all of the deceased people that voted for good ole GW.

Ultimately there is danger in taking data at face value, especially when common sense, conviction, rational gesture indicates otherwise.

8. My school newspaper falsely reported that 32% of people in the US are paid at unemployment. In a hallway debate session a fellow student quoted the 32% as fact. I contended that the number was absurb. I have met thousands of people and a third of them were not working at minimum wage. When I contacted the paper they said it was a typo, while pressing the number two the writer also actually typed the three and no one caught it as a type.

So finally my point is that intelligent intuition, common sense, and rationality should prevail over simple data, which may have been deliberately frauded, accidentally mistyped, mistaken in their methodology or purposefully a joke. To demand data over common sense and the imaginative perseption of the truth is disobey einstein's principle that "Imagination is more important than knowledge.'' especially if that knowledge is wrong

 

[Kandahar]

No, data should be more closely respected than the thwarted view of one's independent common sense


And not to be a grammar nazi, but you didn't put a comma after the word "no," which completely changes the meaning of that sentence. Hehe. :lol:


Anyway, empirical data is better than common sense. "Common sense" is merely the collection of things you have been told your entire life...it doesn't mean that it's right. Empirical data deserves to be scrutinized to make sure that it isn't fabricated, and to make sure that there isn't any alternative explanation. But generally speaking, empirical data is MUCH more important than "common sense."

 

[Morrow]

My personal favorite is the 'research' used to support IQ. The guy who invented, whose names I don't care enough to look up the spelling on, was a master of statistical errors. He gave a bunch of people in different countries IQ tests, and concluded that Finland was the dumbest country on Earth. Despite the fact that he has given the test to people who weren't fluent in English, the language of the test.

He also gave IQ tests to people of different races and concluded that Black people were the dumbest race. Failing to account for any other variables, such as education level. Here is the worst part, in 2002 he was the most cited psychologist in the world. Oh My God!

Research is always accurate, numbers don't lie. That doesn't mean we know what they are measuring. Any and all scientific experiments are trying to eliminate errors, but that doesn't always mean it is possible, especially when it isn't something like physics. Numbers, research, and experiments only really have value to someone who is knowledgeable enough to interpret them. In and of themselves they are worthless.

 

[ashurbanipal]

I think people on the "empirical research" side of the fence would be absolutely amazed at how much inference (which might perhaps be akin to common sense) actually goes into their worldviews.

This is not to dis empirical data by any means--without at least an informal empiricism we'd be living quite primitively indeed. Our knowledge of the world would be, by definition, non-existent. But empirical observations can often only be interpretted in a context of a lot of other empirical data, some of which is assumed from the everyday experience of the interpretter. So common-sense is far from the janky step-child that it might initially appear to be.

 

[rathi]

What you described is not empirical data, but lies. Surveys can certainly be twisted to skew outcomes that do not reflect reality. However, it is possible to put a survey to scrutiny and determine whether the survey is valid or not. "common sense" is a just persons opinions and cannot be analyzed further.

I think you might be better off making a claim like "Surveys, especially politically charged ones, do not always reflect reality. Maintain healthy skepticism if the survey contradicts your common knowledge. Dig deeper to find the circumstances under which the statistics were gathered and look for similar surveys."

Statistics are full of crap, but thats nothing compared to what comes out of peoples mouths.

 

[Dezaad]

In general, I think we are better off with emperical data than without it.

The OP supplies many anecdotes that support his claim that ignoring common sense in favor of empirical data can lead to poor judgements. However the reverse is arguably more true:

Ignoring empirical data for the sake of accepting common sense often leads to poor judgements.

Now, what sort of evidence would the OP accept to verify this claim? The OP seems to implicitly advocate the acceptance of anecdotal evidence over empirical. I could supply anectodes equal in number to the ones found in the OP. Where would that leave us? Should we interpret this to mean that empirical evidence was equally as valuable as anecdotal evidence?

Suppose we gather a hundred stories of low wage workers who lost their jobs. The explanation given by their employers to them was that they could no longer afford as many workers, so they had to choose some to let go. This followed a regulatory raise in the minimum wage, in effect for a month. It seems at this point that common sense would tell us that regulating the minimum wage puts people out of jobs and is therefore bad policy. On the other hand, we have a statistical study by respected economists showing that there is most often a reduction in unemployment following a minimum wage increase. What should our judgement be? What happens if a month later, we read a new report that a different 100 people who have been out of work for 2 years or more have all found new jobs? I think we could soon get lost in an ocean of anecdotes. Should we total up the anecdotes and compare these totals against each other? How would we know that the number of stories we obtained were representative of the population at large?

Anecdotes are the source of common sense notions. Statistics are really just systematic summaries of anecdotes, where a scientific effort is made to have the group sampled representative of the population as a whole.

I will prefer the systematic approach to collecting information about anecdotes over the arbitrary one advocated by the OP. Its just common sense.

 

[AlbqOwl]

The old saw that there are lies, damn lies, and statistics is still true (if you are one prone to common sense anyway. )

Empirical data is a form of anecdotal evidence. It is useful as explanation for how we draw our personal conclusions, using common sense or otherwise, and it is useful to show something is possible. Those using common sense, however, would not be inclined to declare purely empirical data as the norm and certainly would not declare it to be an absolute. The possibility of one's own experience being an anomaly must always be a consideration, though common sense could also allow for the reality of some things If we touch a surface and find it to be hot, for instance, it's fairly safe to believe that the next person who touches it within a reasonable time frame will also find it to be hot. It would not necessarily follow, however, that the surface can be expected to be hot hours later or the next day.

Common sense is based on far more than what one is taught or what one has experienced. Common sense, outside of its immediate, more practical usages, is our God given ability to analyze the known data and consider possibilities that can exist in the unknown. Common sense tells me that the concept of all scientific research is born in common sense.

 

[niftydrifty]

I couldn't answer your poll because I feel every situation is different. I can't answer that question speaking in general terms.

 

[AlbqOwl]

The question in my mind is why did NiftyDrifty post that question on this thread? Common sense tells me it would be far more appropriate elsewhere.

 

[niftydrifty]

Common sense tells me that Niftydrifty didn't post a question. Common sense tells me that empirical data which supports a conclusion can come in a variety of forms. Common sense tells me that studies made which produce conclusions can be infinitely diverse. Common sense tells me that it is impossible to speak about the use of data, without knowing of a specific example of which I am speaking about. Common sense tells me not to make assumptions. Common sense tells me that everything I've just said (with the exception of pointing out that I didn't ask a question to the off-topic poster) lies at the very heart of this discussion.

 

[politicomind]

In the middle of sleeping, I woke up and realized what this relationship is called. It is the deduction to induction relationship of logical philosophy first introduced by Aristotle around 400 bc.

Okay, here's how it is explain. Deduction is a set of premises or principles or as we can say data for our example which leads to a conclusion. As one may say that if an increase in marijuana use is soon followed by a nationwide increase in cocaine use, then, the two seperate premises, yield a potential conclusion, that marijuana may be a gateway drug unto cocaine. Thus, the premises support the conclusion.

Whereas induction would look at the increase of marijuana alone, which would be a conclusion and then postulate as to why marijuana use rose in the first place, then seeks to gain understanding as to why marijuana use rose, for reasons such as the lower price of marijuana, or the practice of some law agency to not bust every drug dealer just the large drug dealers.

Deduction reaches a conclusion based off of the data.

Induction has a conclusion but is in search of the data to support it.

Thus, as Newton sat under his apple tree, his premise of the falling apple, is that all things fall toward the earth. That's the conclusion. He must use induction to seek out what it is that causes this. So he must develop the premise that gravity acts upon all things. That premise "Gravity acts on all things." leads to placing the Earth in as a thing. "Gravity acts on all things, even the Earth." So again he has a conclusion but must hypothesis as to what is acting on the Earth, he must use induction, and postulates that "The sun is acting upon the Earth."

Aristotle explained that it is like a racehorse going around the track, as one begins with premises (or data) he is lead to conclusions, where the half lap ends, the next half lap begins, and new presented with a conclusion he must use induction and returns back to finish the next half lap which opens up the next period of deduction, and endless process.

More simply, data and surveys lead to a generalization which summarizes the survey, but each generalization in itself can be compared against other generalizations and then needs more data to support the concept.

Example: The Data is marijuana and cocaine use are on the rise.
Conclusion: Marijuana use is causing the gateway to cocaine.
If one bad act can lead on to another bad act, can this same generalization be applied to other types of crime. The question, the moment of induction is an analogy, What if the murder of small things leads on to the murder of humans, a small crime on to a big crime. What if the theft of small amounts leads on to larger amounts of theft.

Thus, by getting a conclusion from marijuana were are left with a generialization, from the generalization we are then in search of other concrete data.

Data to platitude to analogy to the need for more data in other areas.

Drug Stats, A platitude that crime begets crime, to analogies to murder and theft, to the need to study murder and crime.


Now I am exhausted and I am going back to bed.

 

[AlbqOwl]

Common sense tells me that Niftydrifty didn't post a question. Common sense tells me that empirical data which supports a conclusion can come in a variety of forms. Common sense tells me that studies made which produce conclusions can be infinitely diverse. Common sense tells me that it is impossible to speak about the use of data, without knowing of a specific example of which I am speaking about. Common sense tells me not to make assumptions. Common sense tells me that everything I've just said (with the exception of pointing out that I didn't ask a question to the off-topic poster) lies at the very heart of this discussion.

Common sense tells me I need to read the post instead of the signature line before I pop off a criticism too. (I saw the sig line as his post.)

My apologies to NiftyDrifty and I shall be more careful in my observations. Now, carry on.

 

[Goobieman]

7. In Cuyahoga Falls, Ohio, 65,000 votes were counted, and of course they were in favor of GW, because .... only 55 000 people live in Cuyahoga Falls. This story was on the Keith Oberman and surprisingly nothing was done. Even election data is made up. 10 thousand more people vote, than are alive in cuyahoga falls. Not to name all of the deceased people that voted for good ole GW.

Cuyahoga Falls, Ohio has a population of 50494 (July 2005)
Cuyahoga Falls, Ohio (OH) Detailed Profile - relocation, real estate, travel, jobs, hotels, hospitals, schools, crime, news

Cuyahoga Falls OH, for the 2004 election, had 35369 registered voters
Summit County Board of Elections Akron, OH

In Cuyahoga falls OH, for the November 2004 general election, 26486 ballots were cast.
Summit County Board of Elections Akron, OH

Ultimately there is danger in taking data at face value, especially when common sense, conviction, rational gesture indicates otherwise

Hmm. Anyone else see the irony?
Anyone?

Anyone care to check the veracity of the other claims made by the OP?










.

 

[politicomind]

Cuyahoga Falls, Ohio has a population of 50494 (July 2005)
Cuyahoga Falls, Ohio (OH) Detailed Profile - relocation, real estate, travel, jobs, hotels, hospitals, schools, crime, news

Cuyahoga Falls OH, for the 2004 election, had 35369 registered voters
Summit County Board of Elections Akron, OH

In Cuyahoga falls OH, for the November 2004 general election, 26486 ballots were cast.
Summit County Board of Elections Akron, OH


Hmm. Anyone else see the irony?
Anyone?

Anyone care to check the veracity of the other claims made by the OP?



.

Listen for yourself, this is the report by keith olberman, previously there was an accompanying list of each city and its statistics, which is no longer available. Yet, I said 55,000 when it was actually more like 50,000. Listen, I never looked up the pop of cuyahoga falls, I was going from sheer memory from three years ago. I was off by 4 thousand on a rough estimate. The only way that I knew what even the approximate amount of pop was in cuyahoga falls was because there was a list with the results which showed cuyahoga falls having about 55 000 people but about 65 000 votes count. I was not aspiring for accuracy because my very effort was to exemplify that if either precise or not it was still a fabricated amount, a fraudulent vote. So enough of the splitting hairs .... get the point. At one part of Ohio, 93 000 votes turned up missing and then in another part of ohio an extra 92 000 votes magically appeared. And where they turned up missing, Bush won because of the shift in the vote due to the missing votes and where they turned up throughout cuyahoga county and surrounding areas the extra votes then leaned also in bush's favor. Get the point please that information is often fabricated for political agendas and for the objectives of the one tampering with or constructing the information.

To be blinded by this request to have precise empirical data has glared over your eyes from seeing the political significance of the estimation itself.

but if you would like to see the video which mentions the cuyahoga county area, please just listen to this video. I do note that it is 16 minutes long. Or if you demand more PRECISION it is 16 minutes and 18 seconds long to satisify your whimsical fancy for absolute precise amounts. Or should I note that it is 16 minutes and 18 seconds long as far as my real player displays.

November 2nd: Voter Fraud* and Homeland Security Terror Threat "Advisories" in Ohio and Florida

 

[niftydrifty]

Cuyahoga Falls, Ohio has a population of 50494 (July 2005)
Cuyahoga Falls, Ohio (OH) Detailed Profile - relocation, real estate, travel, jobs, hotels, hospitals, schools, crime, news

Cuyahoga Falls OH, for the 2004 election, had 35369 registered voters
Summit County Board of Elections Akron, OH

In Cuyahoga falls OH, for the November 2004 general election, 26486 ballots were cast.
Summit County Board of Elections Akron, OH


Hmm. Anyone else see the irony?
Anyone?

Anyone care to check the veracity of the other claims made by the OP?

.

Goobie, at the top of the list he says, "Here are some instances when empirical data has been used fraudulently or erroneously." And the example you cite is an example of Olbermann apparently doing just that. the irony is that politicomind doesn't realize it?

So finally my point is that intelligent intuition, common sense, and rationality should prevail over simple data, which may have been ... accidentally mistyped ...

I believe he's got you here, Goobie. However, I do think that empirical data should be provided which isn't accidently mistyped, a fraud or a joke.

We're striving for accuracy here, not assumptions.

 

[Goobieman]

Goobie, at the top of the list he says, "Here are some instances when empirical data has been used fraudulently or erroneously."

I didn't catch that part. Whoops.