MODELVILLE
Scratch’n out likelihoods,
watchin’ inferences fail;
All of those parameters not backed up by data.
Programming models while swilling 90 schilling.
Smell those assumptions–
They’re beginnin’ to spoil.
Wasted away again in Modelville,
Searchin’ for a last bit of data.
Some people claim that there’s a model to blame,
But I know it’s my data’s fault.
Don’t know the reason,
inferences are not pleas’n
With nothing to show but this new posterior.
But it’s a real beauty,
A Bayesian cutie, how it got here
I haven’t a clue.
Wasted away again in Modelville,
Searchin’ for a last bit of data.
Some people claim that there’s a model to blame,
Now I think,– hell it could be the data at fault.
I blew out my prior,
Slipped on a P-value;
Shrunk my MLEs, had to converge on back home.
But there’s placebo in the blender,
And soon it will render
That worthless null that helps me hang on.
Wasted away again in Modelville
Searchin’ for a last bit of data.
Some people claim that there’s a model to blame,
But I know, it’s your data at fault.
Yes, and some people claim there’s a model to blame, but I know it’s your data at fault.
– Adapted by Ken Burnham (Oct. 24, 2008) from MARGARITAVILLE by Jimmy Buffett
The Testers – Hotel Small P-value
In a dark crowded stat lab, algorithms fill my head,
sound of keyboards beckoning me to bed.
Then up on the screen, I saw shimmering nulls.
My data summarized by statistics not disguised,
I went to stop for the night.
But there they were on the monitor,
P-values! – how they did glower.
I was thinking to myself,
’this could be real or just lack of power.’
Then Student lit a candle, and Fisher showed the way
There were voices down the corridor,
I thought I heard them say…
Welcome to the hotel Small P-Value.
Such a lovely place,
for ongoing disgrace.
Plenty of room at the hotel Small P-Value.
Any time of year, you’ll find significance here.
The method is logically-twisted, it got the frequentist bends.
It has a lot of small -small values, that it calls the ends.
How they dwell on the page, sweet inferences all.
Some starred to remember, not small ones to forget.
So I called the statistician, here’s my plea,
’please bring a bigger alpha.’
He said, ’we haven’t had that option here since nineteen thirty-three.’
And still those voices are calling from far away.
Wake you up in the middle of the night
just to hear them say…
Welcome to the hotel Small P-Value.
Such a lovely place,
for ongoing disgrace.
We’re livin’ it up at the hotel Small P-Value;
point oh-5 – what a nice surprise.
Likelihoods on the ceiling,
Sample spaces that are nice
And she said ’we all just prisoners here, of our own tail device.’
And in the Bayesian chambers,
They gathered for the feast.
They stab nulls with their MCMC knives,
But they just can’t kill the beast.
Last thing I remember,
I was running from a prior.
I had to find the method back
to the place I was before.
’Relax,’ said Dr. Deming,
We are predisposed to discern.
You can have P-values any time you like,
But they’ll never help you learn!
– by Kenneth P. Burnham (August 28, 2007) based on The Eagles – Hotel California