Foreshadowing the Law of Total Expectation
How do local averages inside pathways combine into one correct population average?
How this lesson unfolds.
Begin with Whose treatment cost are we averaging? The sequence then connects Three pathways create three local averages and builds toward Conditional averaging remains linear.
Whose treatment cost are we averaging?
Select one patient at random from the hospital population and let X be that patient's total treatment cost; the pathway H records whether the patient used OPD, Emergency or ICU.
- Learning goalExplain experiment definition and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionThree pathways create three local averages
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Why does each hospital pathway have its own mean?
OPD, Emergency and ICU define different conditional cost distributions, with conditional means Rs 2,000, Rs 8,000 and Rs 40,000.
- Learning goalExplain conditional distributions and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionSolve a new hospital problem locally
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What should we calculate inside each pathway?
First find the local conditional mean E[X given H] within each pathway, then pair it with the probability that a random patient follows that pathway.
- Learning goalExplain local solution strategy and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionThe hospital-wide average is Rs 7,600
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
How do the three pathways merge into one mean?
Weight each conditional mean by its pathway probability: 2000(0.60) + 8000(0.30) + 40000(0.10) = 7600.
Weight each conditional mean by its pathway probability: 2000(0.60) + 8000(0.30) + 40000(0.10) = 7600.
- Learning goalExplain hospital calculation and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionConditional expectation foreshadows the theorem
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What does foreshadowing mean in this lesson?
A local average E[X given H] is a smaller idea that anticipates the Law of Total Expectation, which recombines all local averages into the global mean.
- Learning goalExplain foreshadowing bridge and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionKnown information can simplify an expectation
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
How does conditioning help with a complicated quantity?
Conditioning freezes useful information, so an expression can be simplified locally before its conditional results are averaged over the possible conditions.
- Learning goalExplain conditioning simplification and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionProbabilities tell how often each pathway occurs
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Where does real-world frequency enter the calculation?
The pathway probabilities 0.60, 0.30 and 0.10 describe how often OPD, Emergency and ICU occur and sum to one because the pathways form a complete partition.
- Learning goalExplain pathway probabilities and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionA conditional mean is itself a random variable
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What is the deeper form of conditional expectation?
E[X given H] depends on the realized pathway H: it equals 2000 for OPD, 8000 for Emergency and 40000 for ICU; taking its expectation gives E[X].
E[X given H] depends on the realized pathway H: it equals 2000 for OPD, 8000 for Emergency and 40000 for ICU; taking its expectation gives E[X].
- Learning goalExplain conditional expectation random variable and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionThe complete Law of Total Expectation
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What formula combines every pathway correctly?
For a countable partition H_j, E[X] equals the sum over j of E[X given H_j] times P(H_j), provided the expectation exists and zero-probability terms are handled appropriately.
For a countable partition H_j, E[X] equals the sum over j of E[X given H_j] times P(H_j), provided the expectation exists and zero-probability terms are handled appropriately.
- Learning goalExplain theorem statement and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionIndicators explain why the theorem works
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Why can the global average be split by pathways?
Because a partition satisfies sum_j I(H_j)=1, we may write X as sum_j X I(H_j); linearity and E[X I(H_j)] = E[X given H_j]P(H_j) yield the theorem.
Because a partition satisfies sum_j I(H_j)=1, we may write X as sum_j X I(H_j); linearity and E[X I(H_j)] = E[X given H_j]P(H_j) yield the theorem.
- Learning goalExplain indicator proof and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionThe pathways must form a partition
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What makes conditional averaging valid?
The pathways must cover every relevant outcome and be mutually exclusive, so each patient belongs to exactly one pathway and the probabilities add to one.
- Learning goalExplain partition requirement and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionDo not average conditional means equally
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Why is (2000 + 8000 + 40000) / 3 wrong?
An equal average silently assigns probability one third to every pathway, contradicting the observed 0.60, 0.30 and 0.10 pathway frequencies.
An equal average silently assigns probability one third to every pathway, contradicting the observed 0.60, 0.30 and 0.10 pathway frequencies.
- Learning goalExplain equal weight misconception and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionConditioning changes the population in view
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
What actually happens when we condition on ICU?
Conditioning on ICU filters the population to ICU patients and then computes the average within that restricted group; it does not claim every patient used ICU.
- Learning goalExplain conditioning filter and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionThe arithmetic was already the theorem
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Where was the theorem hiding in the calculation?
The three products 2000(0.60), 8000(0.30) and 40000(0.10) are exactly the terms E[X given H_j]P(H_j) in the symbolic theorem.
- Learning goalExplain arithmetic symbol bridge and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionA local mean is not a weighted contribution
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Why distinguish conditional mean from contribution?
E[X given H_j] is a cost within pathway j, while E[X given H_j]P(H_j) is that pathway's contribution to the overall mean; a rare costly pathway can still contribute strongly.
- Learning goalExplain contribution distinction and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionConditional averaging remains linear
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.
Can we split a complicated ICU cost into parts?
If ICU cost X equals laboratory L plus treatment T plus bed B, then E[X given ICU] equals E[L given ICU] + E[T given ICU] + E[B given ICU], assuming the expectations exist.
If ICU cost X equals laboratory L plus treatment T plus bed B, then E[X given ICU] equals E[L given ICU] + E[T given ICU] + E[B given ICU], assuming the expectations exist.
- Learning goalExplain conditional linearity and connect the local conditional average to the population expectation.
- Helpful foundationsexpected value · conditional probability · weighted average · partition
- Common trapConditional means are weighted by pathway probabilities; they are not averaged equally unless the pathways are equally likely.
- Next connectionApply total expectation to a new partition
More context
The portrait diagram moves from the question to exact notation, a hospital pathway visual, and one concise conclusion.