The Art of Deception in Statistics

The video explores the power and misuse of statistics in various contexts, including marketing, court cases, and advertising. It highlights how Target used statistical algorithms to predict pregnancies, leading to unintended consequences. The video also discusses famous court cases where statistical errors led to wrongful convictions. It explains how statistics can be misleading in advertising and the importance of understanding correlation versus causation. The video concludes with examples of Simpson's paradox and how data representation can skew perceptions.

What was the main goal of Target's statistical algorithm?

• To increase sales of baby products

• To reduce inventory costs

• To improve customer service

• To identify pregnant customers

Tags

CCSS.7.SP.A.2

How did Target avoid making it obvious that they knew customers were pregnant?

• By not sending any coupons

• By calling the customers directly

• By mixing baby product coupons with unrelated items

• By sending only baby product coupons

In the case of Janet Collins, what was the probability of an innocent couple fitting the given descriptions?

• 1 in 1,000

• 1 in 12 million

• 1 in 10,000

• 1 in 1 million

Tags

CCSS.HSS.MD.B.6

CCSS.HSS.MD.B.7

What was the main statistical error in Sally Clark's case?

• Ignoring witness testimony

• Using incorrect data

• Assuming events were independent

• Misinterpreting genetic factors

Why was the Colgate ad claiming '80% of dentists recommend Colgate' misleading?

• Dentists could recommend multiple brands

• It was based on a small sample size

• It was not approved by the advertising authority

• It was based on outdated data

Tags

CCSS.7.SP.A.2

What is the difference between correlation and causation?

• Correlation implies causation

• Causation implies correlation

• Causation does not imply correlation

• Correlation does not imply causation

Tags

CCSS.HSS.ID.C.9

What is the third cause fallacy?

• Assuming A and B are mutually exclusive

• Assuming A causes B when both are caused by C

• Assuming B causes A when A causes B

• Assuming A and B are independent

Tags

CCSS.HSS.ID.C.9

What is Simpson's paradox?

• When correlation implies causation

• When data grouped differently tells a different story

• When independent events are assumed to be dependent

• When small sample sizes are used

Tags

CCSS.HSS.ID.C.9

What is the prosecutor's fallacy?

• Assuming correlation implies causation

• Assuming the probability of A given B is the same as B given A

• Using small sample sizes

• Ignoring witness testimony

Tags

CCSS.HSS.CP.B.6

Why are graphs with non-zero baselines misleading?

• They are based on small sample sizes

• They exaggerate differences

• They use incorrect data

• They are difficult to read