Compute probabilities for chatbot response quality

Context

A chatbot response is considered good only when it is both:

• Helpful, and
• Honest. You are given:
• $P(\text{Helpful}) = 0.8$
• $P(\text{Honest}) = 0.9$ Unless you state otherwise, treat helpfulness and honesty as independent for a single response, and treat responses on different turns as i.i.d.

Questions

1. For one response, what is $P(\text{Good})$?
2. What is the probability of seeing two good responses in a row?
3. If the first three responses are good, what is the probability that the 4th response is good?

Overview: This question checks basic probability understanding: independence, joint event probabilities, and i.i.d. assumptions. It is often used to test probabilistic reasoning about repeated trials in the Statistics & Math domain, and it is mainly a conceptual problem with a simple practical application.

Read the full Data Scientist interview experience this question came from.

Community answers

Answer by akshayprasath

Good = helpful and honest. If independent, then $P(\text{Helpful} \cap \text{Honest}) = P(\text{Helpful}) \cdot P(\text{Honest}) = \frac{9}{10} \times \frac{8}{10} = 0.72 = p$. For 2 responses in a row: $p \times p$. $p$