“A Contribution to Statistics” by Wislawa Szymborska

1 Comment

“A Contribution to Statistics” by Wislawa Szymborska

Poet and Nobel Laureate Wisława Szymborska — that's her, right above — died just a few years ago at 88 years old. I have included this poem in my data analysis classes for a few years because: (a) I love poetry; (b) it has statistics; and (c) as a social psychologist, I believe it summarizes human nature wonderfully.

A Contribution to Statistics

Out of a hundred people…

those who always know better:


doubting every step:

nearly all the rest

glad to lend a hand

if it doesn’t take too long:

as high as forty-nine

always good

because they can't be otherwise:

four, well maybe five

able to admire without envy:


suffering illusions

induced by fleeting youth:

sixty, give or take a few

not to be taken lightly:

forty and four

living in constant fear

of someone or something:


capable of happiness:

twenty-something tops

harmless singly, savage in crowds:

half at least


when forced by circumstances:

better not to know

even ballpark figures

wise after the fact:

just a couple more

than wise before it

taking only things from life:


(I wish I were wrong)

hunched in pain

no flashlight in the dark:


sooner or later


thirty-five, which is a lot


and understanding:


worthy of compassion:



a hundred out of a hundred.

thus far this figure still remains unchanged.

1 Comment

Analysis as translation


Analysis as translation

There is a nearly universal tendency to think of data analysis as nothing more of plugging numbers into the “correct” formulas and coming up with the “correct” results. And, yes, while it is very easy to do something boneheaded in the course of analysis (like forgetting the higher numbers on a variable indicate stronger disagreement, as occasionally but regrettably happens), it is generally true that there are many ways to analyze a given data set and many valid conclusions that can be reached.

To make this point a little clearer, it can be helpful to think about data analysis as a form of translation. For example, here is one famous poem by the great Roman lyric poet Horace (AKA Quintus Horatius Flaccus; 65-8 BCE) in its original Latin:

“Pyrrha Ode”

Quis multa gracilis te puer in rosa

perfusus liquidis urget odoribus

grato, Pyrrha, sub antro?

cui flavam religas comam,

simplex munditiis? heu quotiens fidem

mutatosque deos flebit et aspera

nigris aequora ventis

emirabitur insolens,

qui nunc te fruitur credulus aurea,

qui semper vacuam, semper amabilem

sperat, nescius aurae

fallacis! miseri, quibus

intemptata nites! me tabula sacer

votiva paries indicat uvida

suspendisse potenti

vestimenta maris deo.

I don’t read Latin and I suspect that none of you do either, so here are two English translations (from many, many others). The first is by John Milton (1608-1674), published in 1673.

“Ode 1.5” translated by John Milton

What slender Youth bedew’d with liquid odours

Courts thee on Roses in some pleasant Cave,

Pyrrha for whom bindst thou

In wreaths thy golden Hair,

Plain in thy neatness; O how oft shall he

On Faith and changèd Gods complain: and Seas

Rough with black winds and storms

Unwonted shall admire:

Who now enjoyes thee credulous, all Gold,

Who alwayes vacant alwayes amiable

Hopes thee; of flattering gales

Unmindfull. Hapless they

To whom thou untry’d seem’st fair. Me in my vow’d

Picture the sacred wall declares t’ have hung

My dank and dropping weeds

To the stern God of Sea.

The second translation (or, properly speaking, a “paraphrase”) is by Anthony Hecht (1923-2004), published in 1980.

“An Old Malediction” by Anthony Hecht

What well-heeled knuckle-head, straight from the unisex

Hairstylist and bathed in “Russian Leather,”

Dallies with you these late summer days, Pyrrha,

In your expensive sublet? For whom do you

Slip into something simple by, say, Gucci?

The more fool he who has mapped out for himself

The saline latitudes of incontinent grief.

Dazzled though he be, poor dope, by the golden looks

Your locks fetched up out of a bottle of Clairol,

He will know that the wind changes, the smooth sailing

Is done for, when the breakers wallop him broadside,

When he’s rudderless, dismasted, thoroughly swamped

In that mindless rip-tide that got the best of me

Once, when I ventured in your deeps, Piranha.

Ahh, I love it. Now, as translations, it’s clear that they’re not identical. It is, however, the same sad story told in two (three, if you count Horace’s Latin version) very different times and social contexts. They both address (among other things) the regret of love gone (very) wrong, although the first seems more resigned and the second more bitter. They give different takes on the same situation but emphasizing different aspects of the emotional experience. Interestingly, the same thing can happen with data analysis: two different analyses may use exactly the same dataset but can give different insights into the data (not opposite conclusions, mind you; just different angles) and help make the people behind the data more understandable.