Bayes theorem: notes and practice questions
- Binomial Distribution: Models the number of successes in a fixed number of independent trials with constant success probability (, ). Key stats: Mean = , Variance = . Use Binomial PDF for exact probabilities and CDF for cumulative ones.
- Normal Distribution: A continuous distribution, denoted . Characterized by its bell curve. Standardize using to get . Use Normal CDF to find probabilities (areas) and Inverse Normal to find values given probabilities. Empirical rule: ~68% within , ~95% within , ~99.7% within .
How it is examined
Three events is the ceiling, so a question with a four-branch partition is out of syllabus. The medical-testing context (false positives) is the archetype. A tree diagram is usually the faster route and is fully acceptable, since the guidance never demands the formula. 5 to 7 marks.
Bayes' theorem is given for two and for three events.
Use of Bayes' theorem for a maximum of three events.
Linking questions
- Other contexts: probability methods in medical studies to assess risk factors for certain diseases.
Practice questions
3 questions · 1 medium · 2 hardQuestion 1
MediumPaper 1 · no calculator7 marksA factory produces microchips on two production lines, Line 1 and Line 2. Line 1 produces of the microchips and Line 2 produces the remaining . The probability that a microchip from Line 1 is defective is , and the probability that a microchip from Line 2 is defective is .
A microchip is selected at random from the factory's total output.
(a) Find the probability that it is defective.
(b) Given that a randomly selected microchip is defective, find the probability that it was produced by Line 1.
(c) find the probability that it was produced by Line 2.
Consider the two possible ways a microchip can be defective: it could be from Line 1 and defective, OR it could be from Line 2 and defective. Calculate the probability of each case and add them together. A tree diagram might be helpful.
This is a conditional probability question. Use the formula . You have already calculated the denominator in part (a).
Remember that if a defective chip is selected, it must have come from either Line 1 or Line 2. The probabilities of these two events must sum to 1.
Question 2
HardPaper 2 · calculator5 marksA factory produces two types of light bulbs: standard and long-life. The lifespan of the bulbs, in hours, can be modelled as normal distributions with the following parameters.
| Bulb type | Mean | Standard deviation |
|---|---|---|
| Standard | 800 h | 50 h |
| Long-life | 1000 h | 100 h |
(a) Find the percentage of standard bulbs that have a lifespan of less than 700 hours.
(b) The factory produces a large number of bulbs, of which 60% are standard bulbs. Both types of bulbs are produced and randomly mixed together for packaging. A quality control process identifies and removes all bulbs with a lifespan of less than 700 hours. An inspector randomly selects a bulb from this removed group. Find the probability that it is a standard bulb.
Use the normal distribution CDF with mean 800 and standard deviation 50 to find the probability of a lifespan less than 700 hours, then express it as a percentage.
This is a conditional probability problem. Let S be the event that a bulb is standard, and L be the event that its lifespan is less than 700 hours. You need to find P(S|L). Use the formula for conditional probability or Bayes' theorem. You will need to calculate the probability of a long-life bulb having a lifespan less than 700 hours first.
Question 3
HardPaper 2 · calculator16 marks(a) An electronics factory produces two types of resistors: Type A and Type B.
The resistance, (in Ohms), of Type A resistors is normally distributed with a mean of 120 Ohms and a standard deviation of 5 Ohms.
Find the probability that a randomly selected Type A resistor has a resistance less than 115 Ohms.
(b) In a random selection of 10 Type A resistors, find the probability that exactly 3 have a resistance less than 115 Ohms.
(c) The resistance, (in Ohms), of Type B resistors is normally distributed with a mean of 135 Ohms and a standard deviation of 7 Ohms.
Each day, 70% of the resistors produced are Type A, and 30% are Type B.
On a particular day, a resistor is randomly selected from all those produced at the factory.
Let represent 'Type A resistor' and represent 'Type B resistor'.
(i) Find the probability that the randomly selected resistor has a resistance less than 115 Ohms.
(ii) Given that a randomly selected resistor has a resistance less than 115 Ohms, find the probability that it is a Type A resistor.
(d) The machine that makes the Type A resistors is adjusted so that the mean resistance of the Type A resistors remains the same, but their standard deviation changes to Ohms. The machine that makes the Type B resistors is not adjusted. The probability that the resistance of a randomly selected resistor from these machines is now less than 115 Ohms is 0.18.
Find the value of .
Use the normal cumulative distribution function (CDF) on your GDC. Remember to input the lower bound, upper bound, mean, and standard deviation.
This is a binomial probability problem. Identify the number of trials, the number of successes, and the probability of success from part (a).
You need to consider both types of resistors. Calculate the probability for Type B resistors first, then use the law of total probability, taking into account the proportions of each type.
This is a conditional probability problem, often solved using Bayes' theorem. You need the probability of a Type A resistor having low resistance and the overall probability of a low resistance resistor.
Set up an equation for the new total probability, similar to part (c.i). You'll need to solve for the new probability , then use the inverse normal function to find the z-score, and finally calculate the new standard deviation .
No question on this page matches those filters. Try another difficulty or paper.
Every Bayes theorem question, marked for you
Every answer is marked mark by mark, IB-style, and the AI tutor helps when you are stuck.
Where marks are lost
- Rounding an intermediate value and then using it.
- Answering to the wrong accuracy. Two significant figures, or six, where the rule says exactly or three.
- Writing the answer and nothing else.