A STATISTICAL CONJECTURE: Some Thoughts about Causality Errors
If Student X is an egghead
and Student Y is an arsehole
what parameters must come together
for optimal learning to unfold?
It seems safe to assume most instructors are luggards
and classrooms are essentially cemeteries
How many degrees of freedom are needed
to see the inherent absurdity?
If we pump out enough data
and isolate the variables to a sufficient degree
anything can begin to look like anything
but isn't that the purpose of statistics:
to prop up desired beliefs?
Bill: Ah, flimsy conjectures! So much human activity is based upon frail premises.
Gus: Yes, we live in a house of cards: a wobbly construction of myths, half-truths, and lies.
Nadya: Hasn't it always been so?
Liao: (nodding) Perhaps. As our intelligence grows, so does our capacity for illusion.
=====================================================================================
from _Cyberpoems: Exploring the Human-Machine Interface_ by T Newfields
SUMMARY: Some thoughts about the misuse of statistics in complex systems
KEYWORDS: statistical intent, analytical errors, meta-statistics
Author: T Newfields [Nitta Hirou / Huáng Yuèwǔ] (b. 1955 - ?)
Begun: 2001 in Nagoya, Japan / Finished: 2019 in Yokohama, Japan
Creative Commons License: Attribution. {{CC-BY-4.0}} Granted
< LAST http://www.tnewfields.info/CyberPoems/quantum.htm
TOC http://www.tnewfields.info/CyberPoems/index.html
NEXT > http://www.tnewfields.info/CyberPoems/elegie.htm