Register to reply

Statistical Constant

by Thor
Tags: constant, statistical
Share this thread:
Thor
#1
Jun16-05, 06:31 PM
P: 60
McNally's Statistical Constant: No set of statistics can be more than 6.25% accurate

This may be represented by the mathematical equation -
A(S
S) = A(D) x E(M) x B(J -->N) x I(I <--N)

Where
A(S
S) is the ACCURACY of any (SET of STATISTICS)

A(D) is the ACCURACY of the (DATA) from which they were developed

NOTE: DATA is either accurate or it is not, so this factor is always = 50%]
E(M) is the ERROR of the (METHODOLOGY) used to select the data

NOTE: METHODOLOGY is either in error or it is not, so this factor is also always = 50%]
B(J -->N) is the BIAS of the (JERK who conjures the NUMBER)

NOTE: The JERK is either BIASED or he is not, so this factor is also always = 50%]
I(I <--N) is the INTERPRETATION of the (IDIOT who uses the NUMBER)

NOTE: The IDIOT is either fooled by the NUMBER or he is not, so this factor is also always = 50%]

Given the above
A(S
S ) = 50% x 50% x 50% x 50% = 6.25%

CONCLUSION: The next time someone spouts statistics to you be sure to question their A(S
S )
Phys.Org News Partner Science news on Phys.org
Experts defend operational earthquake forecasting, counter critiques
EU urged to convert TV frequencies to mobile broadband
Sierra Nevada freshwater runoff could drop 26 percent by 2100

Register to reply

Related Discussions
Constant velocity and constant acce Introductory Physics Homework 2
So, say the Fine Structure Constant is not a Constant... Quantum Physics 9
Is Boltzman's Constant related to the Fine Structure Constant? Quantum Physics 6
Connection of Newtonian Constant of Gravitation G and Plank Constant h General Physics 11