Bayes theorem and diagnostic tests study guide

Bayes' Theorem definition

A mathematical formula that describes how to update the probability of a hypothesis based on new evidence. It combines prior knowledge with observed data.

True or False: Bayes' theorem only applies to independent events.

False. Bayes' theorem can be applied to dependent events, as it considers how probabilities change with new information.

Prior probability

The initial assessment of the probability of an event before considering new evidence. Example: A store estimates a 30% chance of rain tomorrow.

Posterior probability

The updated probability of an event after considering new evidence. Example: A store revises the rain forecast to 70% after seeing new weather data.

Formula for Bayes' Theorem

$\displaystyle P(H|E) = \\frac{P(E|H) \cdot P(H)}{P(E)}$ where: - $\displaystyle P(H|E)$ = posterior - $\displaystyle P(E|H)$ = likelihood - $\displaystyle P(H)$ = prior - $\displaystyle P(E)$ = marginal likelihood

Example: Testing for a disease

If the disease's prior probability is 1%, likelihood of a positive test is 90%, and false positive rate is 5%, what’s the updated probability given a positive test result?

Likelihood in Bayes' theorem

The probability of observing the evidence given the hypothesis is true. Example: The chance of a test being positive if a person has a disease.

Marginal likelihood meaning

The total probability of evidence across all hypotheses. It acts as a normalization factor in Bayes' theorem.

Cause → Effect: Positive test result

Cause: Person has the disease. Effect: Test shows positive result. This updates the probability of having the disease.

Comparison: Prior vs. Posterior

- Prior: Initial belief before evidence. - Posterior: Revised belief after considering evidence.

Fill in the blank: P(H|E) is the __________ probability.

posterior probability.

Common mistake in Bayes' theorem

Assuming that the posterior is the same as the prior after new evidence is received. They are often different.

Bayes' theorem in diagnostics →

It's used to update the probability of a disease based on test results.

True or False: A test's sensitivity is the same as its specificity.

False. Sensitivity measures true positives; specificity measures true negatives.

Fill in the blank: If a test has a sensitivity of 90%, then ____ of positive cases are detected.

90% of positive cases are detected.

How does a positive test result affect disease probability?

It increases the probability of having the disease, influenced by pre-test probability.

Diagnostic test sensitivity vs. specificity:

Sensitivity: true positive rate; Specificity: true negative rate.

Example: Test for flu with sensitivity 85%, prior probability 10%.

Calculate posterior probability of flu after a positive test using Bayes' theorem.

What does a high false positive rate imply?

It suggests that many healthy individuals are incorrectly identified as having the disease.

Positive predictive value (PPV) →

The probability that a person actually has the disease given a positive test result.

Impact of low prevalence on PPV:

Lower prevalence leads to lower PPV because more false positives occur.

How is Bayes' theorem applied in interpreting negative results?

A negative result can decrease the probability of disease, but not eliminate it completely.

Question: If sensitivity is 95% and specificity is 90%, what's the implication?

High sensitivity means few false negatives; high specificity means few false positives.

Calculate the effect of test accuracy on disease probability:

Use Bayes' theorem to integrate pre-test probability, sensitivity, and specificity.

A store sells two types of phones.

If a customer buys a phone, there's a 70% chance it's a premium model. If 80% of premium models are returned, what's the chance a return is premium?

True or False: Bayes' theorem applies to financial investments.

True. It helps investors update beliefs about stock value based on new data.

If a test for a disease is 90% accurate, how does it affect decisions?

A positive result means there's a 90% chance of having the disease. If the disease is rare, the actual risk might be lower.

Fill in the blank: If you have a prior belief of 20% for event A, and new evidence increases it to 50%, then Bayes' theorem shows: ____ is updated.

The probability of event A is updated based on new evidence.

Two car models: Model X has a 10% defect rate, Model Y has 2%. If a car is defective, what’s the chance it's Model X?

Use Bayes' theorem: P(X|defect) = P(defect|X) * P(X) / P(defect).

Scenario: You have a 30% chance of rain. If it rains, your event attendance drops to 50%. What’s the overall attendance?

P(attendance) = P(attendance|rain) * P(rain) + P(attendance|no rain) * P(no rain).

Comparing two rental properties: Rent A is $\displaystyle 1200/month, Rent B is$1500/month. Which is better with the same amenities?

Use Bayes' theorem to weigh costs against likelihood of tenant satisfaction based on past data.

A medical test gives a false positive 5% of the time. If you test positive, what’s the probability you actually have the disease?

Bayes' theorem helps calculate the true probability considering the test's accuracy and the disease's prevalence.

A video game has a 25% chance of being a bestseller. If it receives positive reviews, what’s the updated chance?

Bayes' theorem can be used to adjust the initial 25% based on the likelihood of receiving positive reviews.

True or False: Bayes' theorem only applies to situations with binary outcomes.

False. It can be applied to multiple outcomes and probabilities as well.

Calculate the probability of winning a raffle if you buy a ticket from 100 sold.

Using Bayes' theorem, your chance of winning is 1% before the draw; adjust based on any new information about ticket sales.

Two brands of coffee: Brand A is preferred 70% of the time. If a customer has tried both, what’s the chance they choose A?

Bayes' theorem can show that preference increases if customer’s prior experience is known.