SPSS/Stats Humor

At 03:42 PM 9/27/2005, King Douglas wrote: Okay, so I ask you again, why are there no SPSS jokes? Too small and too specific an audience? SPSS programmers have no sense of humor?

Submitted by Marta García-Granero


	I hear that SPSS 14 with Python will allow you go write a program to make regular coffee, but to make espresso you will need a macro. [written by Marta, slightly reworded by King Douglas]

	[Marta submitted a very funny "Garfield" the cat cartoon that I don't want to reproduce without permission, so I'll describe its three panels]: Three birds are pecking at the ground while Garfield pretends not to notice them. Some feathers are scattered on the ground, while Garfield pretends not to notice the remaining two birds More feathers are scattered on the ground. The last remaining bird says scornfully to Garfield, "You must think birds aren't very observant," to which Garfield replies, "On average, two out of three of them aren't."

	The Spanish translation of SPSS 9 gave me the chance to laugh a bit. It was clear that some terms had been translated by someone who had a good knowledge of both languages, but without a clue of the statistical meaning of the words (I remember that in LML survival graphs, the Y axis title 'Log minus log' - log being 'logarithm' had been translated as if 'log' meant a record, a file where you store information - like in 'computer log'; the title in Spanish was 'Reg menos reg', and I needed a bit of time to understand what was going on). I must recognize that SPSS 11/12 has been better translated, but I still frown when I see the term 'mean centered' translated as 'centered mean' (RATIO procedure, BTW).

	Statisticians can't be trusted because on the average, they are mean.

Submitted by Richard Ristow:
	You don't know software until you can hate it creatively. All computing humor tends to the black. There is, of course, the famous Correct, adj.: refers to eliminating faults, errors, or defects. Used to describe a computer program containing only hidden errors.

	If it isn't broken, you haven't looked closely enough.

	Or, my own: Thou shalt not seek for bugs in thy programs, nor yet for errors in thy data; for, if thou dost, lo, thou shalt surely find them.

	Of course, 90% of a project can be done in 90% of the allotted time. The remaining 10% takes the remaining 90%.

	Or technica used colloquially. In the Unix world, if it isn't worth listening to, you "pipe it to /dev/null". (I've suggested "storing it in the write-only memory.")

	Or the famous object-oriented language "Add one to COBOL"

	"If they can get you asking the wrong questions, they don't have to worry about the answers." -Thomas Pynchon, writer (1937- ) "If there's no good answer, change the question." Richard Ristow

	Yes, I've found things to hate in SPSS. I've badgered them that there should be a format to display a variable as its value labels, or value and labels; and something like SAS's PROC CONTENTS, to write the data dictionary, including variable attributes, to an SPSS data file. They tell me, yes, it's a good idea, but you can grab it with OMS: "Metadata? SPSS users don't need no stinking metadata." "If God had meant you to read it, he'd have put it in OMS." Sometimes a glitch is so grand it becomes its own joke, like the "DO ten I" gotcha in FORTRAN. Or in JCL, in the IBM mainframe world: I found a way to make a correct statement incorrect, by specifying a certain parameter instead of letting it default, specifying it correctly. SPSS has a few of those: . FREQUENCIES !varlist/STATISTICS=MEAN. is valid; . FREQUENCIES !varlist /STATISTICS=MEAN. isn't. . AGGREGATE /OUTFILE=* /PRESORTED /* Count sub-categories / /BREAK=FAMKEY HV_DT VISIT# HVIV1 /HAS_BLNK = MAX(IS_BLANK) /HAS_A = MAX(IS_A) /HAS_B = MAX(IS_B). is valid; . AGGREGATE /OUTFILE= /PRESORTED /BREAK=FAMKEY HV_DT VISIT# HVIV1 /* Presence of sub-categories */ /HAS_BLNK = MAX(IS_BLANK) /HAS_A = MAX(IS_A) /HAS_B = MAX(IS_B). isn't. Output presentation? Some things are just too sad for humor.

Submitted by Hector Maletta:
	Statisticians can't be trusted because they are standard deviates.

	People really mean have non-standard deviations, I presume.

	You have probably heard the story of the two statisticians who go hunting. They spot a deer. One of them shoots first, and the shot goes way off to the left of the prey. The other one shoots, this time way off to the right, and then both congratulate each other and exclaim "We hit it, we hit it!!"

Submitted by Art Kendall:
	A standard deviation is not a run-of-the-mill perversion.

	Art suggests this statistics humor web site

Submitted by Raynald Levesque:
	King You ask: "why are there no SPSS jokes?" My explanation: Jokes are usually cruel. Since SPSS users love SPSS, they wouldn't make jokes about it. Regards Ray

SPSSX List humor submitted by Jan Spousta:
	HOW CAN I MAKE A TEST HYPOTHESES BY USING SPSS PROGRAM ? But have patience with authors of such questions, please. Sometimes even non-statisticians simply need to use statistics. And there are others on the list who perhaps will learn from your answers even if the author of the original question will grasp nothing because of his unability to understand the matter. And in the worst case, you can answer like Larson Daley on the forementioned question: Hit the Test button and then hit the Hypotheses button. You might also want to buy the book called "How Can I Make A Test Hypotheses By Using SPSS Program?"