Probability: Praxis Core Academic Skills For Educators: Mathematics 5733
For Praxis Core Mathematics 5733, probability questions can ask for an event’s likelihood, a complement, a combined event or a conditional result.
- Define the sample space first, then decide whether events overlap, are independent, or change after a selection.
- Keep the denominator tied to the experiment’s conditions.
On this page12 sections
- Start with the experiment and sample space
- Use the complement when it is simpler
- Combine events that can overlap
- Tell independence apart from mutual exclusivity
- Conditional probability changes the denominator
- Without replacement versus with replacement
- Use a tree diagram or organized list
- Expected value is a weighted average
- Common probability errors to catch
- A repeatable solution routine
- Build the complement of an “at least one” event
- Check an answer with a second representation
Start with the experiment and sample space
Probability measures how likely an event is within a defined experiment. For equally likely outcomes, calculate the number of outcomes that satisfy the event divided by the total number of outcomes in the sample space. Before counting, state what one trial means and what outcomes are possible. A die roll, a card draw and two draws from a bag have different sample spaces and assumptions.
For a fair six-sided die, the sample space is {1,2,3,4,5,6}. The probability of an even result is 3/6=1/2 because three outcomes are even. If a question describes a nonstandard die or unequal likelihoods, the outcomes are not automatically equally likely. Use the probabilities or experimental information supplied in the prompt.
A fraction is often the most useful exact form. Convert it to a decimal or percentage only when requested or when comparing values. Check that the probability is between zero and one. Zero describes an impossible event under the model; one describes an event that must occur. A result outside that interval usually signals a counting or denominator error.
Use the complement when it is simpler
The complement of event A is the event that A does not happen. Its probability is P(not A)=1−P(A). This is useful when the question asks for at least one success, because counting all ways to get one or more successes can be longer than finding the probability of zero successes and subtracting from one.
Suppose a student randomly chooses one of eight cards numbered 1 through 8. The probability of drawing a number greater than 6 is 2/8=1/4. The complementary event is drawing 6 or less, with probability 6/8=3/4. The two probabilities sum to one because the two events cover every possible card and cannot happen together on the same draw.
A useful check is to name the complement precisely. “Not even” on a standard die means odd, not “less than even” or “not greater than four.” The wording must match the event. In multi-step problems, the complement of “at least one” is “none,” while the complement of “more than one” is “zero or one.” Translate the boundary carefully.
Combine events that can overlap
For two events A and B, the addition rule is P(A or B)=P(A)+P(B)−P(A and B). Subtract the overlap because outcomes in both events were counted twice when the individual probabilities were added. If A and B are mutually exclusive, their overlap is zero and the simpler sum applies.
On one roll of a fair die, let A be an even result and B be a result greater than 4. A is {2,4,6}; B is {5,6}; the overlap is {6}. Thus P(A or B)=3/6+2/6−1/6=4/6=2/3. Adding the first two fractions without subtracting the shared 6 would incorrectly produce 5/6.
Mutually exclusive events cannot occur in the same trial. An odd and even result on one die roll are mutually exclusive. An even result and a number greater than 4 are not, because 6 satisfies both. When a question says “A or B,” determine whether “or” includes both events. In ordinary probability, the union includes outcomes in either event and also outcomes in both.
Tell independence apart from mutual exclusivity
Events A and B are independent when knowing that one occurred does not change the probability of the other. For independent events, P(A and B)=P(A)P(B). Mutual exclusivity means the events cannot occur together, so P(A and B)=0. These are different properties. Two events with positive probabilities cannot be both mutually exclusive and independent.
For the die example, A is even and B is greater than 4. P(A)=1/2, P(B)=1/3 and P(A and B)=1/6. Since (1/2)(1/3)=1/6, these two events are independent. By contrast, even and odd are mutually exclusive, and each has probability 1/2. Their intersection is zero, which is not equal to (1/2)(1/2), so they are not independent.
Do not infer independence just because two events sound unrelated or have equal marginal probabilities. Independence is a mathematical relationship that can be checked by comparing conditional and unconditional probabilities, or by testing the product rule. A question’s experiment may state independence directly, or its structure may support the assumption, such as repeated rolls of a fair die.
Conditional probability changes the denominator
Conditional probability asks for the probability of A given that B has occurred. When P(B) is positive, P(A given B)=P(A and B)/P(B). The condition restricts attention to outcomes in B, so the relevant sample space is smaller. The denominator is not the original total if the condition rules out some outcomes.
Using a die, suppose you know the result is greater than 3. The remaining outcomes are {4,5,6}. Within that restricted set, the probability of an even result is 2/3 because 4 and 6 are even. In the original six-outcome sample space the probability of even was 1/2. The new condition changed the denominator and the probability.
A two-way table can organize conditional probability. If 30 of 50 respondents selected option X, and 18 of those 30 are in group A, then the proportion in group A among X respondents is 18/30=3/5. If the question instead asks what fraction of all respondents are both in group A and selected X, the denominator is 50. State the group after the word “among” before dividing.
Without replacement versus with replacement
A draw without replacement changes the composition of the sample space after the first selection. A draw with replacement returns the selected object before the next draw, so the original proportions are restored. These conditions affect both the numerator and denominator on later draws and often determine whether successive events are independent.
Consider a bag containing five orange and three purple counters. The probability of drawing two orange counters without replacement is (5/8)(4/7)=20/56=5/14. After the first orange counter is removed, four orange counters remain among seven total. The probability with replacement is (5/8)(5/8)=25/64 because the first counter returns before the second draw.
The difference is not a trick in arithmetic. It comes from the experiment’s procedure. Without replacement, the second probability depends on the first outcome. With replacement, the second draw has the same orange probability as the first and is independent when the draws are made under the stated random model. Label the draws and update the counts before multiplying.
Use a tree diagram or organized list
A tree diagram is useful when a process has stages. Each branch is labeled by the probability of that outcome given the path so far. To find the probability of a particular complete path, multiply along its branches. To find an event that can occur through several distinct paths, add the probabilities of those paths.
For the counter bag, the path orange then purple without replacement has probability (5/8)(3/7)=15/56. The path purple then orange has probability (3/8)(5/7)=15/56. The probability of one orange and one purple in either order is 30/56=15/28. The outcomes are disjoint paths, so adding them is appropriate.
An organized list can replace a tree for a small sample space. For two coin tosses, the equally likely outcomes are HH, HT, TH and TT. Exactly one head occurs in HT and TH, so the probability is 2/4=1/2. Listing outcomes helps avoid treating HT and TH as one outcome when order matters.
Expected value is a weighted average
When an outcome has an associated numerical value, expected value is the sum of each value multiplied by its probability. It describes a long-run average over repeated trials under the model, not a promise about the result of one trial. Make sure the values and probabilities correspond to the same outcomes and that the probabilities sum to one.
Suppose a simple game pays 4 points with probability 1/4 and 0 points otherwise. The expected points per play are 4(1/4)+0(3/4)=1. That does not mean every play yields exactly one point. A single play gives either zero or four; one is the average that would emerge over many plays if the probabilities and payoffs remain stable.
For a decision problem, expected value is only one consideration. A question may ask for the numerical expectation, but real choices can also depend on risk, constraints and consequences. Do not claim that a higher expected value guarantees a better outcome in every individual trial.
Common probability errors to catch
Using the wrong denominator is common. In a conditional question, restrict the sample space to the condition. In a percent or proportion question, check whether the prompt asks for a share of the whole group or a share of a subgroup. Write the denominator in words before carrying out division.
Adding probabilities of overlapping events without subtracting the intersection double-counts outcomes. Multiplying probabilities without confirming independence or conditional structure can also be wrong. In sequential experiments, ask whether the first result changes the next trial. With replacement and without replacement are not interchangeable phrases.
Finally, distinguish an observed frequency from a theoretical probability. If a survey reports that 17 of 40 selected participants chose an option, 17/40 is the observed sample proportion. It may estimate a broader probability, but it is not automatically the exact probability for every person. Follow the problem’s model and avoid extending a sample result beyond its evidence.
A repeatable solution routine
Write the experiment, define the event and identify the sample space. Decide whether outcomes are equally likely or whether probabilities are supplied. Determine whether the question uses a complement, union, intersection or condition. For a sequence, record whether replacement occurs and whether the stages are independent. Calculate, simplify and check that the result is within zero and one.
For a combined event, sketch a Venn diagram or list the overlapping outcomes. For a conditional event, cross out outcomes that do not satisfy the condition before counting. For a sequence, use a tree with conditional branch labels. These representations make the assumptions visible and reduce reliance on memorized wording alone.
The ETS 5733 companion is the authority for exam scope and examples. OpenStax’s introductory statistics text provides accessible explanations of independence, mutually exclusive events and the distinction between them. The original numerical cases in this article are teaching examples created to demonstrate the rules; they are not reproduced ETS questions.
Build the complement of an “at least one” event
Suppose a fair coin is tossed three times and the question asks for at least one head. There are 2^3=8 equally likely ordered outcomes, but listing the seven outcomes with a head takes time. The complement is three tails, with probability (1/2)^3=1/8. Therefore the probability of at least one head is 1−1/8=7/8. The word “at least” often signals a useful complement.
Check an answer with a second representation
For the two-counter example, the probability of two orange draws can also be found by counting ordered pairs. There are 8×7=56 equally likely ordered selections without replacement. Five orange counters can occupy the first position and four the second, giving 20 favorable ordered pairs. The probability is 20/56=5/14. Agreement with the sequential multiplication provides an independent check.
A Venn diagram is useful for event overlap, while a tree diagram is useful for stages. Do not use a tree’s branch probabilities as if all paths have equal probability unless the model says so. For a tree, multiply within one path and add distinct paths that form the target event. Write labels directly on branches so the conditions remain clear.
Common questions
How do I know if two events are independent?
Check whether P(A and B) equals P(A)P(B), or whether conditioning on one event leaves the probability of the other unchanged. Do not infer independence from the wording alone.
What does without replacement do to probability?
It changes the available outcomes after a selection. Recalculate the next draw using the remaining total and counts; the second event may depend on the first.
Are mutually exclusive events independent?
Not when both have positive probability. Mutually exclusive events have an intersection probability of zero, while the product of their positive probabilities is greater than zero.