Summary Statistics

Song copyright by Gary Cziko

Well I [G] got me some data [C] to make me more [G] wise,
But the [C] data I [G] had to [D7] summarize,
Since my [G] Ed Psych 480 I did [C] mem-o-[G]-rize ,
I [C] knew just [D7] what to [G] do.

Refrain: Oh, the [D7] mode, the mean, and the [C] med-i-[G]-an,
For a [C] large set of [G] data you [D7] choose just one,
As a typical value we [C] have these [G] three
Measures [C] of Central [D7] Ten-den-[G]-cy.

Now the [G] mode is quite easy to [C] find, you [G] see,
But it [C] suffers from [G] insta-[D7]-bil-ity,
Yet for [G] categorical data use it [C] confident-[G]- ly,
There [C] just ain't no [D7] other [G] choice.

Repeat Refrain

If you [G] like to rank data use the [C] me-di-[G]-an,
It's so [C] much [G] fun to find the [D7] middle one,
But if [G] you have two middle values [C] not [G] one
Their [C] midpoint is [D7] what [G] you need.

Repeat Refrain

The [G] a-rith-metic mean may be [C] your best [G] bet,
And [C] you can be [G] sure that I'll [D7] never for-get,
It [G] reflects all the values of the [C] data [G] set,
[C] E-ven those [D7] quite [G] ex-treme.

Lyrics copyright by Sheila O'Leary Weaver
may sing to the tune of "I Will Survive" (Gloria Gaynor)

At first I was afraid; I was petrified.
I just hadn't realized my data was skewed to one side.
And now I'd spent so many hours finding the mean and the SD,
But now I see, just how misleading they can be.

So now I'll go; walk out the door.
I'll get an F in here; I may not stay in college anymore.
I should have dropped this stupid class; I should have bought the answer key.
I should have thought for just one second, office hours were for me.

But no, I'll try, a new technique.
I'll rank the data, and the middle is the median I seek.
The middles of the halves are called the Quartiles 1 and 3,
Add the min and max - that's the five number summary.

Stat 51, I will survive
As long as I can pass this class, I know I'll be alive,
I've got all my life to live,
One semester's time to give:
I will survive, I will survive,
Hey, Hey....

Dance Interlude

To put it in a graph, I'll draw a number line.
Make spacing even or interpretation's undermined.
Draw vertical lines at each number and connect the middle three,
And a boxplot will result to show the lack of symmetry.

And now I'll find the IQR.
It's Q3 minus Q1. If it's big, the data spreads out far.
If a number's more than 1.5 IQR from the box,
It gets a star, and you can call it an outlier.

And now I'm done. My summary,
Is more appropriate for data that is lacking symmetry
The mean and the SD may be affected by the skew.
But my way's strong: It's more robust and it's not wrong....

Stat 51, I will survive
As long as I can pass this class, I know I'll be alive,
I've got all my life to live,
One semester's time to give:
I will survive, I will survive,
Hey, Hey....

More Dancing

Lyric copyright by Dennis Pearl
may sing to the tune of "Hit the Road Jack" (Percy Mayfield/Ray Charles)
(italicized words are sung by male soloist. Remaining words sung by female singers.)

Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.
What you say?
Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.

Woah Woman, oh woman, won't ya give me my mean,
You're the meanest old woman that I've ever seen.
I guess if you say so
I'll have to take x-bar and go. (That's right)

Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.
What you say?
Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.

Well baby, listen baby, don't ya treat me this-a way
It's just a summary stat to say.
Don't care what you say 'cause it's understood
You ain't got no data for my likelihood.
Well, I guess if you say so
I'll have to take x-bar and go. (That's right)

Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.
What you say?
Pick the mode Jack and don't you keep track of no mu, no mu, no mu, no mu.
Pick the mode Jack and don't you keep track of no mu.

Well!!
Don't you keep track of no mu.
Uhh what did you say?
Don't you keep track of no mu.
I did not understand it
Don't you keep track of no mu.
I came to talk it over
Don't you keep track of no mu.
I thought we had a better understanding
Don't you keep track of no mu.
Oh baby don't be so common
Don't you keep track of no mu.
Don't you want a hat on pi? x2
Don't you keep track of no mu.
Oh baby it isn't fair
Ooh yeahh

Lyric copyright ©2011, 2024 by Lawrence Mark Lesser.
May sing to the tune of "Mean" (Taylor Swift)

You.......know the average lies with lots of skew among your exes.
You.......could sum up and divide for a measure of center.
You.....know "all too well" to check the shape of the data before
You......choose to use the mean.
'Cause it can get pulled out by just one single point,
And you don't know what you don't know....

CHORUS:
Sometimes data have a real asymmetry
Or values that are really quite extreme.
In those cases, better use the median
If the middle is what you need - that's not gonna be the mean!

That won't be the mean with outliers, just forget it,
Do you see just what I mean & mean & mean & mean? (repeat chorus)