- Explains common misuses of statistics and misrepresentation of probability. Use common sense as the first line of defense. This impacts advertising, criminal trials (what is the likelihood that the defendant is guilty based on the blood found at the scene vs whats the possibility that any other individual is guilty based on the blood found at the scene).

- Post hoc ergo propter hoc- After this, therefore because of this. Because this event followed another event, the subsequent event must have been caused by the first event. We link events that might be temporarily related but aren't actually cause and effect.

- Likelihood of two unrelated events both happening= probability of event 1 x probability of event 2→ lower than the probability of each event independently. Likelihood of flipping one head on a coin followed by flipping another head.

- Likelihood of two related events- for example, the likelihood of freezing weather tonight and tomorrow night→ higher given the occurrence of freezing weather the first night.

- Vaccines lead to autism? People look at the increased percentage of autism diagnoses. Autism was more frequently diagnosed because it became more understood. Autism is commonly diagnosed between 18-24 months and the MMR vaccine is given around 12 months.

- Was 9/11 an inside job? Why did the towers collapse vertically? Easy to overwhelm with questions and theories designed to cast doubt on the events. But structural engineers never found anything suspicious about it.

- Breast cancer. Pretest probability- occurrence in the population. 1/8 women develop BrCA. Mammograms can over-call diagnoses (false positive).

- Positive test + confirmed diagnosis= true positive

- Negative test + confirmed absence of diagnosis= true negative

- Positive test + confirmed absence of diagnosis= false positive (low FP= high specificity)

- Negative test + confirmed diagnosis= false negative (low FN= high sensitivity)