Statistically Speaking, Of Course...
Mike Williams
February 23, 2006
I thought this would be a good week to talk about numbers. Not the CBS show “Numbers” but rather the various numbers we get exposed to on a daily basis. These are the polls, studies, and dollar amounts that permeate our society. Some politicians live by them while insurance companies determine how much your premium should be based on even more numbers. I really wanted to talk about the now infamous “port deal” but as of now, I’m not sure what to think about that.
I’ve written about a new tax being proposed in Pennsylvania this week. This tax would affect those businesses and individuals who buy products or services outside of the commonwealth. As it stands now, the PA tax office estimates that it’s losing $228 million dollars because residents and business owners are shopping online without paying Pennsylvania sales tax. While I have an unfavorable opinion on any “new” taxes, what caught my attention was the amount of money the state claims they are losing. Where did they come up with $228 million?
To arrive at any type of valuable estimate, you would need to have access to various data. Tax records would be helpful but since these residents aren’t paying state taxes, it stands to reason that the exact number is a complete unknown. You could conduct a poll. A poll would be a limited number of people (usually 1000) and again, that would vary tremendously as people in rural areas may tend to buy more items on-line due to their location. You could look at credit card receipts but then, what about companies like “Green Zap” or “PayPal?” How about shipping records? Well, you’d have to look at every company that ships to Pennsylvania and what about those folks living in another state who may simply buy something for a family member within the state? In other words, the PA resident sends a family member money and the family member buys and sends them a brand of potato chips they like. If you ask UPS, they could tell you how many packages they handle but then, how do you know the value of what’s in the box?
As you can quickly surmise, there really is no intelligent way of arriving at any monetary figure because you simply don’t have enough information.
I look at polling data the same way. Using simple logic, pollsters expect us to believe that 1000 people can determine how millions feel about a particular issue and go on to present the polling data as if, it carries certainty. During the 2004 Presidential election, Zogby’s exit polling data showed that John F. Kerry was clearly the frontrunner. Reality showed us something quite different. Polls make an assumption that everyone who takes them is telling the truth. Various radio talk shows, including Glenn Beck, featured voters who admitted on the air that they had lied to pollsters. The bad numbers may very well have ignited the Republicans who may have simply “stayed at home” otherwise, to get out and “join the fight.” I don’t know that for sure. It’s pure speculation.
I wrote a column a few years ago about the now famous “1 in 5” statistic that seemed to be the poster child for statistics in general. I focused on the Ad Council and particularly the area of on-line sexual predators. I believe the ad featured Jamie Lee Curtis. The ad cites says that by the age of 10, “1 in 5” kids have been sexual solicited on-line. I found that number to be incredible. Actually, it was too incredible to believe. I conducted my own study with a slightly smaller sampling but my numbers fell closer to the statistic being “1 in 400.” I could do the same survey and come up with “1 in 50” or “1 in 102” or “1 in 10000.” It all depends on where I sample, how many kids I ask, and how willing they are to talk to me about this. There is absolutely no way to nail down a number on this subject. I realize that we have an awful lot of “kooks” out there but there are so many variables and parameters that we can never fully know.
Taking what I’ve written so far and applying it to the world we live in becomes a tad bit frightening. Based on information we really don’t know, we modify our behavior, invent new government programs, and change our eating habits. We thought that changing the foam insulation on the space shuttle’s booster rocket was a good idea because we “believe” that the components that make up the foam may harm the environment. NASA looked at studies and various reports comprised of limited scope and made the change to “environmentally safe foam.” The end result was the shuttle Columbia breaking up upon re-entry killing the entire crew. We “think” we made the environment safer and killed a few people in the process.
One of my personal favorites is health studies and how they seem to change results every few years. Coffee is good. Coffee is bad. Well, coffee is good again. Eat meat. Don’t eat meat. Eating oats will lower cholesterol. No wait, they don’t. Butter is bad. No wait, our substitute is worse. What’s a trans-fatty acid anyway? The ulcer rate is up because we’ve stopped eating butter and butter naturally coats the stomach. A low fat diet helps prevent cancer, heart disease, and stroke. The latest $415 million dollar government study says “no it doesn’t.”
Perhaps, the biggest lie of all is life expectancy. In my lifetime, the number seems to go up every year. I happen to believe it’s going down. The people who live to be over a hundred today didn’t get there by eating low fat anything. There simply was no sugar substitute when they were kids. If you wanted a sugar substitute, you used honey. In other words, they ate “al natural” because they didn’t have a choice. Unfortunately, the life expectancy figure is based on statistics. We also do not have a model to suggest that anything they come up with today will serve to prolong your life. There is no one person who grew up on natural substitutes that we can look at from birth until death and say, “See, this guy lived to be 150!” Today, the 75 year old can simply claim that the substitutes haven’t killed them yet. I’ve never seen a fat 100 year old either.
I’ll leave you with some great words impressed upon me as a younger man in college in Great Lakes, Illinois. I had a few professors who repeated the same thing to us over and over again. At the time, I figured that these words were specific to engineers but I suppose they also apply to just about everything in life. Especially, today.
Question everything.
I’ve written about a new tax being proposed in Pennsylvania this week. This tax would affect those businesses and individuals who buy products or services outside of the commonwealth. As it stands now, the PA tax office estimates that it’s losing $228 million dollars because residents and business owners are shopping online without paying Pennsylvania sales tax. While I have an unfavorable opinion on any “new” taxes, what caught my attention was the amount of money the state claims they are losing. Where did they come up with $228 million?
To arrive at any type of valuable estimate, you would need to have access to various data. Tax records would be helpful but since these residents aren’t paying state taxes, it stands to reason that the exact number is a complete unknown. You could conduct a poll. A poll would be a limited number of people (usually 1000) and again, that would vary tremendously as people in rural areas may tend to buy more items on-line due to their location. You could look at credit card receipts but then, what about companies like “Green Zap” or “PayPal?” How about shipping records? Well, you’d have to look at every company that ships to Pennsylvania and what about those folks living in another state who may simply buy something for a family member within the state? In other words, the PA resident sends a family member money and the family member buys and sends them a brand of potato chips they like. If you ask UPS, they could tell you how many packages they handle but then, how do you know the value of what’s in the box?
As you can quickly surmise, there really is no intelligent way of arriving at any monetary figure because you simply don’t have enough information.
I look at polling data the same way. Using simple logic, pollsters expect us to believe that 1000 people can determine how millions feel about a particular issue and go on to present the polling data as if, it carries certainty. During the 2004 Presidential election, Zogby’s exit polling data showed that John F. Kerry was clearly the frontrunner. Reality showed us something quite different. Polls make an assumption that everyone who takes them is telling the truth. Various radio talk shows, including Glenn Beck, featured voters who admitted on the air that they had lied to pollsters. The bad numbers may very well have ignited the Republicans who may have simply “stayed at home” otherwise, to get out and “join the fight.” I don’t know that for sure. It’s pure speculation.
I wrote a column a few years ago about the now famous “1 in 5” statistic that seemed to be the poster child for statistics in general. I focused on the Ad Council and particularly the area of on-line sexual predators. I believe the ad featured Jamie Lee Curtis. The ad cites says that by the age of 10, “1 in 5” kids have been sexual solicited on-line. I found that number to be incredible. Actually, it was too incredible to believe. I conducted my own study with a slightly smaller sampling but my numbers fell closer to the statistic being “1 in 400.” I could do the same survey and come up with “1 in 50” or “1 in 102” or “1 in 10000.” It all depends on where I sample, how many kids I ask, and how willing they are to talk to me about this. There is absolutely no way to nail down a number on this subject. I realize that we have an awful lot of “kooks” out there but there are so many variables and parameters that we can never fully know.
Taking what I’ve written so far and applying it to the world we live in becomes a tad bit frightening. Based on information we really don’t know, we modify our behavior, invent new government programs, and change our eating habits. We thought that changing the foam insulation on the space shuttle’s booster rocket was a good idea because we “believe” that the components that make up the foam may harm the environment. NASA looked at studies and various reports comprised of limited scope and made the change to “environmentally safe foam.” The end result was the shuttle Columbia breaking up upon re-entry killing the entire crew. We “think” we made the environment safer and killed a few people in the process.
One of my personal favorites is health studies and how they seem to change results every few years. Coffee is good. Coffee is bad. Well, coffee is good again. Eat meat. Don’t eat meat. Eating oats will lower cholesterol. No wait, they don’t. Butter is bad. No wait, our substitute is worse. What’s a trans-fatty acid anyway? The ulcer rate is up because we’ve stopped eating butter and butter naturally coats the stomach. A low fat diet helps prevent cancer, heart disease, and stroke. The latest $415 million dollar government study says “no it doesn’t.”
Perhaps, the biggest lie of all is life expectancy. In my lifetime, the number seems to go up every year. I happen to believe it’s going down. The people who live to be over a hundred today didn’t get there by eating low fat anything. There simply was no sugar substitute when they were kids. If you wanted a sugar substitute, you used honey. In other words, they ate “al natural” because they didn’t have a choice. Unfortunately, the life expectancy figure is based on statistics. We also do not have a model to suggest that anything they come up with today will serve to prolong your life. There is no one person who grew up on natural substitutes that we can look at from birth until death and say, “See, this guy lived to be 150!” Today, the 75 year old can simply claim that the substitutes haven’t killed them yet. I’ve never seen a fat 100 year old either.
I’ll leave you with some great words impressed upon me as a younger man in college in Great Lakes, Illinois. I had a few professors who repeated the same thing to us over and over again. At the time, I figured that these words were specific to engineers but I suppose they also apply to just about everything in life. Especially, today.
Question everything.
Print
Email
