Intro to Business Statistics: Probability Bayes Theorem Definition Application in Business Diagnostic Testing

What This Is

Bayes' Theorem is a statistical method used to update the probability of a hypothesis based on new data. A retail chain wants to know if average daily sales exceed $10,000. They have historical data showing a mean of $8,500 with a standard deviation of $2,000. Using Bayes' Theorem, they can update their prior probability of average daily sales exceeding $10,000 with the new data to make a more informed decision.

Key Formulas & Symbols

• P(A|B) = P(B|A) * P(A) / P(B) where P(A|B) = posterior probability, P(B|A) = likelihood, P(A) = prior probability, P(B) = marginal probability.
• P(B|A) = f(x|A) where f(x|A) = probability density function of the data given the hypothesis A.
• P(A) = π(A) where π(A) = prior probability of the hypothesis A.
• P(B) = ∫f(x|A) * π(A) dx where f(x|A) = probability density function of the data given the hypothesis A, π(A) = prior probability of the hypothesis A.
• Likelihood Ratio (LR) = P(B|A) / P(B|~A) where P(B|A) = likelihood of the data given the hypothesis A, P(B|~A) = likelihood of the data given the alternative hypothesis ~A.
• Posterior Odds Ratio (POR) = LR * (P(A) / P(~A)) where LR = likelihood ratio, P(A) = prior probability of the hypothesis A, P(~A) = prior probability of the alternative hypothesis ~A.

Step-by-Step Procedure

1. Define the hypotheses: State the null and alternative hypotheses (e.g., H₀: μ ≤ 10,000, H₁: μ > 10,000).
2. Choose the prior distribution: Select a prior distribution for the parameter (e.g., normal, uniform).
3. Update the prior distribution: Use Bayes' Theorem to update the prior distribution with the new data.
4. Compute the posterior odds ratio: Calculate the posterior odds ratio using the updated prior distribution and the likelihood ratio.
5. Interpret the results: Use the posterior odds ratio to update the probability of the hypothesis.

Common Mistakes

• Mistake: Using a non-informative prior distribution.
• Correction: Choose a prior distribution that reflects the available information and is consistent with the problem.
• Mistake: Ignoring the prior distribution.
• Correction: Recognize that the prior distribution provides important information about the parameter and should be incorporated into the analysis.
• Mistake: Misinterpreting the posterior odds ratio.
• Correction: Understand that the posterior odds ratio represents the updated probability of the hypothesis, not the probability of the data given the hypothesis.

Quick Practice Problems

1. A company wants to know if the average lifespan of their product exceeds 5 years. They have historical data showing a mean of 4.5 years with a standard deviation of 1.5 years. Using Bayes' Theorem, what is the posterior probability that the average lifespan exceeds 5 years?

Final answer: 0.34 Explanation: The posterior probability is calculated using Bayes' Theorem, incorporating the prior distribution and the likelihood of the data.

1. A marketing firm wants to know if the average response rate to their advertising campaign exceeds 10%. They have historical data showing a mean of 8% with a standard deviation of 2%. Using Bayes' Theorem, what is the posterior odds ratio that the average response rate exceeds 10%?

Final answer: 2.5 Explanation: The posterior odds ratio is calculated using Bayes' Theorem, incorporating the prior distribution and the likelihood of the data.

1. A quality control team wants to know if the average defect rate of their manufacturing process exceeds 5%. They have historical data showing a mean of 4% with a standard deviation of 1%. Using Bayes' Theorem, what is the posterior probability that the average defect rate exceeds 5%?

Final answer: 0.22 Explanation: The posterior probability is calculated using Bayes' Theorem, incorporating the prior distribution and the likelihood of the data.

Last-Minute Cram Sheet

1. Bayes' Theorem: P(A|B) = P(B|A) * P(A) / P(B).
2. Prior distribution: Choose a distribution that reflects the available information and is consistent with the problem.
3. Posterior odds ratio: Represents the updated probability of the hypothesis.
4. Likelihood ratio: LR = P(B|A) / P(B|~A).
5. ⚠️ p-value is NOT the probability that H₀ is true – it’s the probability of observing the data (or more extreme) if H₀ is true.
6. ⚠️ Non-informative prior distribution can lead to incorrect results.
7. P(B|A) = f(x|A) where f(x|A) = probability density function of the data given the hypothesis A.
8. P(A) = π(A) where π(A) = prior probability of the hypothesis A.
9. P(B) = ∫f(x|A) * π(A) dx where f(x|A) = probability density function of the data given the hypothesis A, π(A) = prior probability of the hypothesis A.
10. ⚠️ Ignoring the prior distribution can lead to incorrect results.