Sunday, December 17, 2006

Oh, man. I really have to go, BAD!

 

By Joseph Walther

 

Dear Miss Kraby, Um, please excuse Joey from school yesterday. He had terrible awful cramps and a real bad case of diarea direa dyhrrea the shits! He’s been eating at Taco Bell a lot.

 

            Enough already about E. coli and Taco Bell! Yes, it’s a dangerous bacterium. Even though the average healthy adult can survive it easily, children and older folks can’t. I’m not trying to make light of this. Once medical experts establish a credible case, we should do whatever it takes to stem the spread. Implicated establishments should voluntarily close for health inspections. If they don’t do so, state authorities should force them to do so, with the added caveat of potential criminal action if anyone dies in the interim.

 

            Taco Bells in New Jersey, New York, Pennsylvania, and Delaware were all implicated. Seventy-one cases were reported before the outbreak was declared under control by the Center for Disease Control. The Taco Bells all closed voluntarily and remained closed while both food and employees underwent extensive testing.

 

            Delaware is the second smallest state in the country. We could place it inside Bronx County, New York and have room left over. Yet, in New Castle County—one of three in the state, the others being Kent and Sussex—there are 14-Taco Bells. This number includes the restaurants that sell only Mexican fare, as well as the combined Kentucky Fried Chicken (KFC)/Taco Bell units. While I have not been able to establish credible numbers for either Kent or Sussex counties, the 14-units in New Castle County cater to around 3.9 million customers a year. Only two, that’s TWO, cases of E coli were confirmed.

 

            Perspective seems to have been lost when it comes to news outlets: newspapers, TV, and radio, particularly the talk shows. It’s probably obvious, but incase you’ve missed it; I’m not big on radio talking heads, especially the local people. If the U. S. government taxed brains, all of these folks would receive large rebates and the majority of the callers would receive even larger ones.

 

            However, reason and logic do not sell newspapers or elevate TV and radio talk show ratings. Sensationalizing does, even when it’s uncalled for. A local radio talk show here in Delaware did it up big time. One of the hosts said that he would never eat at Taco Bell again. “If it HAS happened, it CAN happen,” he kept saying. He tries to be clever with these utterances. Of course, if this particular talk show host were a member of a group stranded on a tropical island, he’d be Gilligan, only much older and with a nasty disposition.

 

            I think a bit of perspective is in order regarding Taco Bell and our chances of contracting E coli by eating in one of their restaurants. While I’m at it, I’m going to include some perspective on inferential statistics in general.

 

            For the record, I do not eat Taco Bell food. I tried it once, a long time ago. I didn’t like it. I won’t go back. It isn’t because I’m skeptical, but because I’ve been to Mexico many times. I’ve tasted the real thing down there. It’s good stuff. People, who are able to make this kind of taste/quality comparison, tend to let out one of those big shivering EEEUWs whenever they think about the taste of Taco Bell food. In all, I don’t mind admitting that I’d rather snort Drano and eat nuclear waste than Taco Bell food.

 

            Food inspection in the United States, according to the local radio talkies, is dismal. They keep telling us that only two percent of the food coming into this country is inspected. What they don’t say—because they don’t know—is that if the inspection samples are random and truly representative of the food population, a two percent sample size is enough to give about a ± 2% margin of error on a 95% confidence interval. In other words, forecasts will fall within a 2% range of reality—up and down—and the forecasters will be correct about 95% of the time.

 

            This is good enough for me. Perhaps it’s not so for you. It’s OK, too, because we’re all entitled to an opinion. We could spend a lot more money and go to a 99.7% confidence interval for inspection forecasts. I think it would be a waste, though, because there isn’t much room for improvement on the margin of error. So, we’ll just spend larger amounts of money for the same margin of error. But, by God, we’ll be right about it a larger percentage of the time.

 

            Statistical Inference is the holy grail of hard science. I’ve spent my entire professional career involved with it, up to my eyeballs. Here’s an example, from first hand experience, of what I mean. It’s a true story. But, the thing to remember is that the same principle applies to all honest statistical measures.

 

            On January 31, 1971, NASA launched Apollo-14 with three astronauts onboard: Alan Shepard, Stu Roosa, and Ed Mitchell. There was nothing unique about the launch except that it followed on the heels of the Apollo 13 disaster. NASA, determined to avoid a repeat, achieved a statistical miracle. Apollo 14 launched with a safety reliability factor of 99.97%. Everyone at NASA, particularly at mission control, was justified in patting each other on the back. The astronauts were elated! Now, here’s how perspective and reality can kick us right in our righteous butts.

 

            Even with a 99.97% safety reliability factor, around 6,231 components could STILL fail. Which of these identical facts would YOU have preferred to hear, “We’re 99.97% sure that the ship will launch and return to earth without a hitch,” or, “We have great news, men, we’ve narrowed the things that could endanger the mission and your lives down to only 6,231 possibilities?” You see, it’s not what we say, but how we say it. Statistical truth, ya gotta love it.

 

            Here’s another statistical truth. Human beings screw things up. They’ve been doing it for a long time. Some people believe it all started with Adam and Eve. I don’t know about that, but I do know that whenever humans are involved in anything, crap happens. It’s never a question of “if” or “can”. It’s ALWAYS a question of “when.” Whether a mistake has already happened, as our local talk show host insisted, is irrelevant. Mr. Murphy—we all know who he is—has been fine tuning human screw-ups since the dawn of human civilization. He’ll continue doing so, too, only with more dangerous stuff. Think nukes!

 

            When I was a kid, I became “please let me die,” puking sick after eating Skippy

Chunky-style peanut butter. About six months later, the same thing happened after I ate some lemon meringue pie. It might have been E coli, but regardless, I have not eaten either since. Of course, in the case of the peanut butter, I always fail to mention that I ate the entire jar. Oh yes, I also ate the entire pie, not just one slice. In fact, my mother had warned me that if I ate one more slice of the pie, I was going to explode. So, I told her to give me another slice and get out of the way. Um, she didn’t like that; I almost didn’t live long enough to get sick from the pie!

 

            Back in my heyday, when I knew everything; before I became academically sophisticated; before I had kids of my own, I had a tendency to become fall-down, puking sick after drinking beer by the case. Did I blame this on “bad” beer? Did I look for a lawyer to sue the beer company? NO! Yes, I consumed massive quantities of it, but I was eating peanuts the whole time. Statistically, it had to be those damn peanuts. I’m just tickled to death that there was no lemon meringue pie on those bars. Otherwise, I might have died!

 

            So, if you enjoy Taco Bell food, have some. We’re all going to die eventually, anyway. Stay away from the chunky-style peanut butter, though. And, I wouldn’t advise taking any chances with the lemon meringue pie, either. Oh and here’s something interesting. Given similar exposure numbers, a person living in New Castle County, Delaware has a far greater chance of becoming an assault victim in the thriving metropolis of downtown Wilmington, than contracting E coli by eating at Taco Bell.

 

            Not that I think there’s a connection with Taco Bell, but I’ve established a statistical truth of my own. To get people to move out of your way, yell, “I’M GOING TO THROW UP,” instead of “Excuse me.” It’s over 99% more effective. For the statistical record, I base this assertion on a 5σ (pronounced as 5-sigma) confidence interval, which is 99.999994% with a ±.005% margin of error. I’m no cheapskate when it comes to establishing statistical truths!. Go ahead and try it if you don’t believe me. God, how I love statistics!

 

Have a great week. My study team did not make it to the prison in Smyrna, Delaware. The warden had to cancel for now. We will be going, though. It’ll be interesting!

 

Joseph Walther is a freelance writer and publisher of The True Facts. Copyright laws apply to all material on this site. Send your comments. Just click here.