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:
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
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
vestimenta maris deo.
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.
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.