# Statistical howlers

**URL:** <https://boards.straightdope.com/t/statistical-howlers/132578>\
**Category:** The BBQ Pit\
**Created:** [October 16, 2002, 9:40pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578 "2002-10-16T21:40:53Z")\
**Posts on this page:** 17\
**Page:** 3

<div class="post-metadata">

**Author:** ![Rick](https://avatars.discourse-cdn.com/v4/letter/r/9f8e36/32.png) [@Rick](https://boards.straightdope.com/u/Rick)\
**Post date:** [October 19, 2002, 8:42pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/41 "2002-10-19T20:42:52Z")

</div>

> [@](#):
>
> \*Originally posted by astorian \*  
> \*\*Along the same lines, I don’t know how many times I saw the late Ann Landers express dismay that (her words, I’m not making this up!), “50% of the doctors in this country graduated in the bottom half of their class at medical school.”
> 
> I guess if Ann had her way, medical schools would tighten up their standards so that EVERY doctor would graduate at the top of his/her class! \*\*

[hijack]  
I think you missed Ann’s point.  
Just because someone spent however many years in medical school, that does not make them caring, or for that matter competent.  
One time I was referred to a surgeon to have some severe lacerations on my fingers sewn up. Now for a living I am an automotive instructor and I watch auto mechanics work for a living. This guy that was working on me, I discovered, did not have the manual dexterity required of an auto mechanic. If I had a student show such poor dexterity, I would suggest to them that they consider a different line of work. This guy may have been Marcus Welby in other aspects of medicine, but as a surgeon he was worthless.

BTW, do you know the technical term for the guy that finishes dead last in his class at medical school?  
They call him doctor.  
[/hijack]

---

<div class="post-metadata">

**Author:** ![Fang](https://avatars.discourse-cdn.com/v4/letter/f/96bed5/32.png) [@Fang](https://boards.straightdope.com/u/Fang)\
**Post date:** [October 19, 2002, 9:14pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/42 "2002-10-19T21:14:59Z")

</div>

But did any of you know that 50% of Americans are below average? We must correct this problem immediately!

---

<div class="post-metadata">

**Author:** ![matt\_mcl](https://avatars.discourse-cdn.com/v4/letter/m/7ab992/32.png) [@matt\_mcl](https://boards.straightdope.com/u/matt_mcl)\
**Post date:** [October 19, 2002, 9:40pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/43 "2002-10-19T21:40:14Z")

</div>

No, 50% of Americans are below the _median_, not the average.

---

<div class="post-metadata">

**Author:** ![UncleBill](https://avatars.discourse-cdn.com/v4/letter/u/d26b3c/32.png) [@UncleBill](https://boards.straightdope.com/u/UncleBill)\
**Post date:** [October 20, 2002, 2:11am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/44 "2002-10-20T02:11:05Z")

</div>

Without the bottom half of each class there would BE no top 50% of the class! Think of the tragedy!

---

<div class="post-metadata">

**Author:** ![Kaitlyn](https://avatars.discourse-cdn.com/v4/letter/k/73ab20/32.png) [@Kaitlyn](https://boards.straightdope.com/u/Kaitlyn)\
**Post date:** [October 20, 2002, 7:10am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/45 "2002-10-20T07:10:39Z")

</div>

Back to the OP: On a recent episode of Enterprise, Dr. Phlox reveals he is part of a marriage with three husbands and three wives. He goes on to say that this makes for 720 possible relationships.

There are 720 possible permutations of six people taken six at a time, but that’s irreleveant to the number of relationships. The number of relationships between two people (as implied by Dr. Phlox) is figured by finding the number of combinations (not permutations) of six people taken two at a time, and that number is 15.

If you want to include relationships with more than two people in them as seperate relationships (not my inference about Phlox’s meaning), there are 20 combinations of six people taken three at at time, 15 combinations of six people taken four at time, six combinations of six people five at a time, and only 1 for six taken six at a time, for a totatl of 57 possible relationships. Nowhere near 720.

Someone was using the wrong formula and doesn’t understand the difference between combinations (order doesn’t matter) and permutations (order does matter).

And on the Ann Landers/doctor thing, there are so many problems with that that I don’t know where to begin. First, medical school graduates are not doctors until they have completed a residency and passed a pretty stringent set of tests, which likely eliminates a larger portion of that bottom half than the top.

Second, not all Medical schools are created equal. The bottom half of Harvard Medical is likely more highly qualified than most of the graduates of Podunk U.

Third, as a general rule, American medical schools (here I am excepting diploma mills) only get the cream of the crop in the first place, and tend to have high standards for even completing the program, so that lowest graduate from medical school is much more likely to be competent than the lowest graduate from most less stringent courses.

---

<div class="post-metadata">

**Author:** ![Koxinga](https://avatars.discourse-cdn.com/v4/letter/k/4af34b/32.png) [@Koxinga](https://boards.straightdope.com/u/Koxinga)\
**Post date:** [October 20, 2002, 10:37am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/46 "2002-10-20T10:37:53Z")

</div>

> [@](#):
>
> \*Originally posted by Number Six \*  
> Back to the OP: On a recent episode of Enterprise, Dr. Phlox reveals he is part of a marriage with three husbands and three wives. He goes on to say that this makes for 720 possible relationships . . .
> 
> Someone was using the wrong formula and doesn’t understand the difference between combinations (order doesn’t matter) and permutations (order does matter).

Maybe his definition of a relationship includes who gets to be on top.

---

<div class="post-metadata">

**Author:** ![sturmhauke](https://avatars.discourse-cdn.com/v4/letter/s/e47c2d/32.png) [@sturmhauke](https://boards.straightdope.com/u/sturmhauke)\
**Post date:** [October 21, 2002, 8:12am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/47 "2002-10-21T08:12:28Z")

</div>

> [@](#):
>
> \*Originally posted by matt\_mcl \*  
> \*\*No, 50% of Americans are below the _median_, not the average. \*\*

While that is technically correct, for large populations they tend to be very close.

---

<div class="post-metadata">

**Author:** ![RickJay](https://avatars.discourse-cdn.com/v4/letter/r/bb73d2/32.png) [@RickJay](https://boards.straightdope.com/u/RickJay)\
**Post date:** [October 21, 2002, 3:23pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/48 "2002-10-21T15:23:25Z")

</div>

> [@](#):
>
> \*Originally posted by matt\_mcl \*  
> \*\*No, 50% of Americans are below the _median_, not the average. \*\*

The word “average” is a perfectly acceptable substitute for “median.” You’d be correct in stating that 50% of all people are below the median and not the MEAN… but the non-technical term “average” can be used to describe either. (Or it can be used to describe the mode, the most commonly occurring specific value in a set.) From Merriam-Webster:

AVERAGE: 1 a : a single value (as a mean, mode, or median) that summarizes or represents the general significance of a set of unequal values.

---

<div class="post-metadata">

**Author:** ![Tretiak](https://avatars.discourse-cdn.com/v4/letter/t/a88e57/32.png) [@Tretiak](https://boards.straightdope.com/u/Tretiak)\
**Post date:** [October 21, 2002, 3:40pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/49 "2002-10-21T15:40:20Z")

</div>

Boy you guys are tough 🙂 First I am not here to support the public official in the OP. But, using only 3 eggs does reduce the probabilty of getting salmonella 75% IF the 1 out of 4 stat is true, which it probably isn’t. Of course, then using 2 egg would further reduce the chances 75%, etc… And throwing three eggs out of the initial dozen, but still using four eggs in the original does not have the same effect.

Still it is very silly.

---

<div class="post-metadata">

**Author:** ![Miss\_Bianca](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Miss\_Bianca](https://boards.straightdope.com/u/Miss_Bianca)\
**Post date:** [October 21, 2002, 4:22pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/50 "2002-10-21T16:22:02Z")

</div>

> [@](#):
>
> \*Originally posted by sturmhauke \*  
> \*\*  
> While that is technically correct, for large populations they tend to be very close. \*\*

Regardless of the size of the population, the mean and median will only be “very close” if the data are normally distributed. For characteristics that are not normally distributed (e.g., income), the mean and median can be quite different. In those cases, the median is the preferred statistic because the mean is unduly influenced by extreme scores (e.g., [insert name of famous rich person]'s income).

And **RickJay** , point taken, but I for one applaud **matt\_mcl** ’s precision!

---

<div class="post-metadata">

**Author:** ![Sofa\_King](https://avatars.discourse-cdn.com/v4/letter/s/58956e/32.png) [@Sofa\_King](https://boards.straightdope.com/u/Sofa_King)\
**Post date:** [October 21, 2002, 5:30pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/51 "2002-10-21T17:30:32Z")

</div>

Well, since we are in the firey Pit, I suppose we could drag this decomposing horse out of the meatlocker again:

[How long would your state run with ANWR’s oil](http://www.doi.gov/news/anwr8.htm)?

And here’s the [accompanying fluff](http://www.doi.gov/news/020312.html). Perhaps most amusing, that first link implies that the entire ANWR reserve could be slurped up by the state of Texas alone in less time than it can be pumped out of the ground.

Sometimes it seems as if statistics juggling is the last refuge of the devious; other times I think it comes from genuine cluelessness. I suppose it’s possible that the Department of the Interior has both of those qualities.

---

<div class="post-metadata">

**Author:** ![december](https://avatars.discourse-cdn.com/v4/letter/d/838e76/32.png) [@december](https://boards.straightdope.com/u/december)\
**Post date:** [October 21, 2002, 11:07pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/52 "2002-10-21T23:07:43Z")

</div>

> [@](#):
>
> It’s a fact. Interest rates are at their lowest point this year. But the likelihood of them going up in the near future is a very real possibility.

[Ditech.com](http://Ditech.com) commercial

---

<div class="post-metadata">

**Author:** ![Kaitlyn](https://avatars.discourse-cdn.com/v4/letter/k/73ab20/32.png) [@Kaitlyn](https://boards.straightdope.com/u/Kaitlyn)\
**Post date:** [October 22, 2002, 2:45am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/53 "2002-10-22T02:45:51Z")

</div>

> [@](#):
>
> Boy you guys are tough First I am not here to support the public official in the OP. But, using only 3 eggs does reduce the probabilty of getting salmonella 75% IF the 1 out of 4 stat is true, which it probably isn’t. Of course, then using 2 egg would further reduce the chances 75%, etc… And throwing three eggs out of the initial dozen, but still using four eggs in the original does not have the same effect.

Not quite. Assuming that 1 out of 4 eggs carries salmonella, the probability of a four egg omelette having salmonella is 68%. The probability of a 3 egg omelette having salmonella is 58%, and two egg omelette having salmonella is 44%. Using three instead of four reduces the odds by 15%, and using two instead of 3 reduces the odds by anther 24%. Reduce that to 1, and you reduce the odds by anther 57%.

Cook the eggs to a proper temperature (until the yolk, where the salmonella resides, is fully solidified), and you reduce the odds to nothing because you kill the salmonella.

---

<div class="post-metadata">

**Author:** ![InternetLegend](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/internetlegend/32/13154_2.png) [@InternetLegend](https://boards.straightdope.com/u/InternetLegend)\
**Post date:** [October 22, 2002, 6:22am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/54 "2002-10-22T06:22:35Z")

</div>

The Chairman of the Board of our Chamber of Commerce wrote a letter to the morning paper that decried the “low expectations” our public schools have of their students. The focus of his outrage was the fact that the state considers schools to meet standards if their students’ median test scores fall between the 40th and 60th percentiles for the nation and it considers a school to be “exemplary” if its median test scores are in the 70th percentile.

This brilliant community leader said:

> [@](#):
>
> "When did 40 percent become an acceptable passing score in anything and when did 70 percent come to mean that achievement was exemplary…?
> 
> …Exemplary must mean 90 percent not 70 percent…"

Now, I can understand how someone who had glanced over a newspaper and seen those numbers could mistake a percentile for a raw score. I cannot understand how the chairman of a large city’s Chamber of Commerce, a body which recently made it its business to endorse a slate of candidates for the school board, outspending their opponents many times over in a race that had historically been low-key, inexpensive, and non-political, could make a mistake of this magnitude. I plan to write my own letter to the paper and the Chamber. I don’t, however, expect the chairman to understand my reasoning.

Albuquerque, where all the children are above average…

---

<div class="post-metadata">

**Author:** ![don\_t\_ask](https://avatars.discourse-cdn.com/v4/letter/d/e68b1a/32.png) [@don\_t\_ask](https://boards.straightdope.com/u/don_t_ask)\
**Post date:** [October 22, 2002, 7:01am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/55 "2002-10-22T07:01:29Z")

</div>

My all time favourite was the clever misrepresentation in the old Pepsi ads.

“In recent blind taste tests more Coke drinkers preferred the taste of Pepsi!”

So if I get Coke drinkers and split them into groups of 3 what would I need to be telling the truth? I would only need 2 drinkers in a group to pick the Pepsi and repeat until it happens again. Even if I polled 1,000 groups before that happened the statement above is true.

---

<div class="post-metadata">

**Author:** ![Early\_Out](https://avatars.discourse-cdn.com/v4/letter/e/6f9a4e/32.png) [@Early\_Out](https://boards.straightdope.com/u/Early_Out)\
**Post date:** [October 22, 2002, 9:51am UTC](https://boards.straightdope.com/t/statistical-howlers/132578/56 "2002-10-22T09:51:11Z")

</div>

The chances of someone having a bomb on a plane are one in a million. The chances of _two_ people on the same plane each having a bomb are one in a trillion.

Moral: When you fly, always take a bomb with you.

---

<div class="post-metadata">

**Author:** ![Coil](https://avatars.discourse-cdn.com/v4/letter/c/bb73d2/32.png) [@Coil](https://boards.straightdope.com/u/Coil)\
**Post date:** [October 22, 2002, 3:25pm UTC](https://boards.straightdope.com/t/statistical-howlers/132578/57 "2002-10-22T15:25:30Z")

</div>

> [@](#):
>
> \*Originally posted by Bob Scene \*  
> \*\*Remember how pissed off the boss in Dilbert got when he found out that 40 % of sick leave was used on Mondays and Fridays? \*\*

I just got an evil idea. I think I am going to e-mail a letter to every member of parliament and ask them what they think should be done about the problem of 40% of all sick leave being taken on these two days.

It´ll be interesting to see how many out of the 349 will catch on.

[Previous page](https://boards.straightdope.com/t/statistical-howlers/132578.md?page=2)
