Probability (continued) Supplementary Material for CFB3333/PHY3333 - - PowerPoint PPT Presentation

▶
probability continued
SMART_READER_LITE
LIVE PREVIEW

Probability (continued) Supplementary Material for CFB3333/PHY3333 - - PowerPoint PPT Presentation

Mar 06, 2023 197 likes •306 views

http://www.physics.smu.edu/pseudo Probability (continued) Supplementary Material for CFB3333/PHY3333 Professors John Cotton and Stephen Sekula March 21, 2012 Based on the following information on the web:

slide-1
SLIDE 1

http://www.physics.smu.edu/pseudo

Probability (continued)

Supplementary Material for CFB3333/PHY3333 Professors John Cotton and Stephen Sekula March 21, 2012 Based on the following information on the web:

http://www.physics.smu.edu/pseudo/Probability/

slide-2
SLIDE 2

http://www.physics.smu.edu/pseudo

The Birthday Paradox

  • We act surprised when we find out that

somebody we meet at a party has the same birthday as us.

  • Let's try it in class.
  • We'll start people calling out their birthday (month

and day only, please!) one at a time

  • Yell “Me too!” if you hear somebody else with your

birthday

slide-3
SLIDE 3

http://www.physics.smu.edu/pseudo

The Telephone Game

  • What happens when we base our assumptions
  • n something from “word of mouth”?
  • Let's play the Telephone Game!
slide-4
SLIDE 4

http://www.physics.smu.edu/pseudo

The Telephone Game

  • Information has a low probability of surviving

word-of-mouth transmission

  • Information degradation is a huge problem!
  • You can't always base assessments on what you

hear.

slide-5
SLIDE 5

http://www.physics.smu.edu/pseudo

The Telephone Game Story

  • Here's what the passage SHOULD have said.

“A rock band was playing in a submarine The noise awakened a whale Which ordered a pizza for lunch Anchovies go well with squid.”

slide-6
SLIDE 6

http://www.physics.smu.edu/pseudo

The Clustering Illusion

slide-7
SLIDE 7

http://www.physics.smu.edu/pseudo

Simpson's Paradox

  • Two or more studies may each reach a common

conclusion

  • But . . . when combined they reach the opposing

conclusion.

slide-8
SLIDE 8

http://www.physics.smu.edu/pseudo

The non-transitive paradox

  • If A is better than B, and B is better than C, how

is A related to C?

  • C is not necessarily better than A!
  • Consider rock, paper, scissors...
slide-9
SLIDE 9

http://www.physics.smu.edu/pseudo

Extrapolation and Coincidence

  • Extrapolation: predicting what happens next in

the data based on what has happened before.

  • Coincidence: when two unrelated events occur

in relation to one another.

  • we often assign correlation or causality and neglect

that they may just be a coincidence