ECON0006 Chap.9 Conditional Probability and Expectation
Conditional Probability and Expectation
Conditional Probability and Expectation turns uncertainty into weighted calculation. A continuous density needs a support, nonnegativity and total area one before it can define probabilities. Expectations weight outcomes by that density, while conditioning changes the information and often the relevant support. Several random variables introduce joint and marginal views.
Noisy equations may produce multiple solution branches that must be kept distinct when expected solutions are formed.
What this chapter covers
- 01
Probability support
- 02
Density normalisation
- 03
Moments and means
- 04
Joint distributions
- 05
Conditional probability
- 06
Conditional expectation
- 07
Expected solutions
Worked example · free
Normalise a linear density
- 3Set the area c/2 equal to one, giving c=2.
- 2Weight ξ by 2ξ over [0,1], giving an expectation of 2/3.
- 1Check that 2/3 lies in the support and above one half.
Key terms
- Support
- The set of values a random variable can take under the model.
- Probability density
- A nonnegative weighting function whose total area over the support is one.
- Conditional expectation
- An expected value calculated after stated information is known.
Conditional Probability and Expectation FAQ
Can a density exceed one?
Yes. A density is probability per unit and may exceed one over a narrow region. Its integrated probability over an event must remain between zero and one.
Why normalise before finding a mean?
Without total probability one, the weights do not form a valid distribution and the calculated average has no probability interpretation. Normalisation fixes the constant before any moment is evaluated.
What changes after conditioning?
The information set changes the relevant weights and possibly the support. State the event or observed variable before calculating the conditional object, then renormalise over the remaining possibilities.
How can I check an expectation before calculating it?
Inspect the support and the shape of the probability weights. The expected value should lie within the range of possible outcomes and should be pulled toward regions receiving more weight. Use that qualitative forecast after integration: a result outside the support or on the wrong side of an obviously skewed distribution signals an error in bounds, normalisation or the outcome-times-density integrand.
Exam move
Write the support, proposed density and normalisation check on three separate lines. Predict the approximate location of the mean from the shape before integrating. Practise distinguishing point values of a density from probabilities over intervals. For several variables, label joint and marginal quantities and specify what is held fixed when conditioning.
When a noisy equation has multiple roots, keep the branches separate through weighting. Review errors by missing support, negative density, area not one, unconditional mean presented as conditional and expected value outside plausible bounds. Use a fixed probability audit for every proposed distribution. Mark its support, test nonnegativity throughout that support and verify that the total mass or area equals one.
Only after those checks should a moment be calculated. Sketching the shape provides a useful independent prediction: a density weighted toward high outcomes should not produce a mean near the lower boundary. Keep constants exact until the end so normalisation errors do not disappear inside rounding. If an integral is improper, state the limiting operation and check that the probability and requested moment actually exist.
Build fluency by translating among three representations of the same event: words, set notation and a region on a line or plane. For a continuous variable, shade the interval whose area gives the probability rather than treating the density height at one point as probability. For two variables, draw the joint support and label the order of integration before calculating a marginal.
Reverse the order on a second pass when the geometry permits and compare answers. This exercise exposes missing bounds and helps distinguish an impossible region from one that merely has small probability. Conditioning should visibly change the information set.
Write the conditioning event beside the numerator, calculate its probability for the denominator and verify that the resulting weights sum or integrate to one over the remaining support. Compare the conditional answer with the unconditional one and explain which outcomes became relatively more important.
If the conditioning event has zero probability under the elementary formula, say that the displayed ratio is undefined instead of dividing mechanically. In multivariable work, name what is observed, what remains random and which value or range is being held fixed. When uncertainty enters an equation, distinguish the expected input, the solution for each input and the expected solution.
Nonlinearity means these operations need not commute. If several solution branches exist, identify the branch rule and weight each admissible outcome consistently rather than averaging roots without explanation. End a timed response with checks on units, support and bounds: probabilities stay between zero and one, a density height need not, and an expectation should be plausible for the outcomes it averages.
These checks often locate an incorrect integrand faster than repeating the entire calculation.
Working through Conditional Probability and Expectation in ECON0006? Sia is AskSia’s AI Economics tutor — ask any ECON0006 Conditional Probability and Expectation question and get a clear, step-by-step explanation grounded in how ECON0006 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.