Independent vs mutually exclusive events: probability traps
Independent events do not change each other's probability. Mutually exclusive events cannot occur together in the same trial. These are different conditions: two mutually exclusive events with positive probabilities cannot be independent. For 'both', use an intersection; for 'either', allow for overlap. For conditional probability, restrict the sample space to the event you already know occurred before counting favourable outcomes.
In an IMAT probability problem, “independent” is a mathematical condition, not a synonym for “different”. Decide what the events mean before choosing a formula.
Two questions, two different tests
| Property | Question to ask | Mathematical test |
|---|---|---|
| Independence | Does knowing B occurred change the probability of A? | P(A ∩ B) = P(A) × P(B) |
| Mutual exclusivity | Can A and B happen in the same trial? | P(A ∩ B) = 0 |
The symbol ∩ means “both”; ∪ means “at least one”. In the usual inclusive meaning of “A or B”, outcomes where both happen are included.
If A and B are mutually exclusive and both have positive probability, observing B makes A impossible. Its probability therefore changes, so the events are not independent. OpenStax compares these two conditions.
One die, two contrasting examples
All examples here are original practice scenarios. Assume a fair six-sided die.
Let A mean an even number: {2, 4, 6}. Let B mean a number greater than 4: {5, 6}.
- P(A) = 3/6 = 1/2.
- P(B) = 2/6 = 1/3.
- Both happen only on 6, so P(A ∩ B) = 1/6.
Because 1/6 = (1/2) × (1/3), these events are independent. They are not mutually exclusive: rolling 6 satisfies both.
Now let C mean an odd number: {1, 3, 5}. A and C cannot occur together on one roll, so they are mutually exclusive. They are not independent: knowing that the roll is odd rules out an even result.
“Or” needs an overlap check
The general addition rule is:
P(A ∪ B) = P(A) + P(B) − P(A ∩ B)
For the even-or-greater-than-4 example, this gives 1/2 + 1/3 − 1/6 = 2/3. Listing {2, 4, 5, 6} confirms four favourable outcomes out of six.
Simply adding 1/2 and 1/3 counts the outcome 6 twice. Addition without subtracting overlap works for mutually exclusive events because their intersection is zero.
“Given” changes the denominator
Conditional probability is:
P(A | B) = P(A ∩ B) / P(B), provided P(B) > 0.
Suppose a separate practice question says the die result is greater than 3. The remaining possibilities are {4, 5, 6}. Two are even, so the probability of an even result given that information is 2/3, not 3/6.
Write the reduced sample space before calculating. “Probability of even given greater than 3” and “probability of greater than 3 given even” ask different questions; swapping them is not a general rule. The basic probability rules explain the conditional multiplication rule as well.
Replacement can change the answer
A bag contains three blue and two red counters. Draw two at random.
Without replacement: after drawing blue first, only two blue counters remain among four. The probability of two blues is (3/5) × (2/4) = 3/10.
With replacement and mixing between draws: the original proportions return. The probability is (3/5) × (3/5) = 9/25.
The word “and” tells you to find an intersection, but it does not justify multiplying the original probabilities. The general rule is P(A ∩ B) = P(A) × P(B | A); independence lets you replace the conditional term with P(B).
Self-check: at least one
Using the same bag, draw two counters without replacement. What is the probability of at least one red?
Use the complement: the only way to get no red is to get two blues. Therefore the answer is 1 − 3/10 = 7/10. “At least one red” includes both one red and two reds.
When reviewing a wrong answer, label the cause: overlooked overlap, wrong sample space or unjustified independence. Then try a fresh example. For a biological application, continue with genetic crosses and probability.
For categorical statements such as “if”, “only if” and “must”, use necessary and sufficient conditions. A logical implication and a conditional probability require different reasoning.
Sources
Or keep going in the iPhone app.
Found a mistake? Write to hello@imatlearn.com