EduProd
EduProd
JEE PCM PLATFORM
πŸ“Š Dashboard
LEARN MODULE Β· SCHEMATIC MODE

Trace it. Derive it. Know it.

Every cluster below maps one schematic path β€” from first principles to the final equation. Expand a node to follow the full derivation trace.

⚑ INTEGRATED VIRTUAL LABORATORIES (4)
Open Subject Labs Catalog β†—
Bayes Tree
Stage β†—
Conditional Probability
Stage β†—
Independent Events
Stage β†—
Random Variables
Stage β†—
01

Conditional Probability P(A|B) & Fundamental Properties

3 concepts Β· Core
3 nodes

How does prior knowledge that event B has occurred restrict the sample space to update the probability of event A?

The conditional probability of event E given that event F has already occurred is P(E|F) = P(E intersection F) / P(F), with P(F) > 0. Fundamental properties include P(S|F) = 1, P((A union B)|F) = P(A|F) + P(B|F) - P((A intersection B)|F), and P(E'|F) = 1 - P(E|F).

01

Definition of Conditional Probability: P(A|B) = P(A cap B) / P(B)

For any two events A and B associated with sample space S, the conditional probability of event A given that event B has occurred is P(A|B) = P(A cap B) / P(B), provided P(B) > 0.

Key Idea
The given event B becomes the new restricted sample space. Only outcomes in the intersection A cap B remain favorable to A.
Why It Works
Dividing the joint probability P(A cap B) by the new universal total P(B) renormalizes the probabilities so that P(B|B) = 1.
Boundary Trap
Dividing by P(A) instead of P(B): always divide by the event that is GIVEN (the one on the right side of the vertical bar).
Application Note
Roll a fair 6-sided die. Let A = {getting a number > 3} = {4, 5, 6}, B = {getting an even number} = {2, 4, 6}. A cap B = {4, 6}. P(A cap B) = 2/6, P(B) = 3/6. P(A|B) = (2/6)/(3/6) = 2/3.
Derivation Trace Β· Definition of Conditional Probability: P(A|B) = P(A cap B) / P(B)
Conditional probability is simply shrinking your universe. If you are told 'the rolled die landed on an EVEN number' (event F = {2, 4, 6}), you throw away {1, 3, 5} forever! Now if someone asks 'what is the chance of rolling a 6?' (event E), it is 1 out of 3, NOT 1 out of 6. The denominator shrank from 6 to 3.
If E and F are two events associated with the same sample space of a random experiment, then the conditional probability of the event E given that F has occurred, i.e. P(E|F) is given by P(E|F) = P(E intersection F) / P(F) provided P(F) != 0. The properties of ordinary probability apply identically to conditional probabilities with respect to the reduced sample space F.
Never divide by the probability of the question event! P(E|F) means 'probability of E given F has happened'. The given event F is the condition, so P(F) is always in the denominator. Writing P(E intersection F) / P(E) is P(F|E), not P(E|F).
Open the Conditional Probability Explorer. Adjust event Venn overlaps to see how P(E|F) scales dynamically as the conditioning set F shrinks or expands.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If P(A) = 0.6, P(B) = 0.5, and P(A cap B) = 0.2, what is P(A|B)?

Answer: P(A|B) = P(A cap B) / P(B) = 0.2 / 0.5 = 2/5 = 0.4.

02

Fundamental Properties of Conditional Probability

Property 1: P(S|B) = P(B|B) = 1. Property 2: P((E cup F)|B) = P(E|B) + P(F|B) - P((E cap F)|B). Property 3: Complement rule: P(A'|B) = 1 - P(A|B).

Key Idea
All standard axioms and theorems of classical probability hold identically within the conditioned probability space B.
Why It Works
Since conditioning on B simply replaces the universal set S with B, every Boolean set identity continues to apply.
Boundary Trap
Writing P(A'|B) = 1 - P(A) instead of 1 - P(A|B): both sides must be conditioned on the same event B.
Application Note
If P(A|B) = 0.35, then the probability that A does NOT occur given B has occurred is P(A'|B) = 1 - 0.35 = 0.65.
Derivation Trace Β· Fundamental Properties of Conditional Probability
Conditional probability is simply shrinking your universe. If you are told 'the rolled die landed on an EVEN number' (event F = {2, 4, 6}), you throw away {1, 3, 5} forever! Now if someone asks 'what is the chance of rolling a 6?' (event E), it is 1 out of 3, NOT 1 out of 6. The denominator shrank from 6 to 3.
If E and F are two events associated with the same sample space of a random experiment, then the conditional probability of the event E given that F has occurred, i.e. P(E|F) is given by P(E|F) = P(E intersection F) / P(F) provided P(F) != 0. The properties of ordinary probability apply identically to conditional probabilities with respect to the reduced sample space F.
Never divide by the probability of the question event! P(E|F) means 'probability of E given F has happened'. The given event F is the condition, so P(F) is always in the denominator. Writing P(E intersection F) / P(E) is P(F|E), not P(E|F).
Open the Conditional Probability Explorer. Adjust event Venn overlaps to see how P(E|F) scales dynamically as the conditioning set F shrinks or expands.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If P(A|B) = 3/7, what is P(A'|B)?

Answer: P(A'|B) = 1 - P(A|B) = 1 - 3/7 = 4/7.

03

Sample Space Reduction Technique for Discrete Experiments

In discrete experiments with equally likely outcomes, P(A|B) can be calculated directly by counting: P(A|B) = n(A cap B) / n(B), bypassing decimal division of individual probabilities.

Key Idea
List only the outcomes of event B, and count how many of those outcomes satisfy event A.
Why It Works
Since each elementary outcome has probability 1/n(S), the factors of n(S) cancel in the ratio: [n(A cap B)/n(S)] / [n(B)/n(S)] = n(A cap B)/n(B).
Boundary Trap
Answering 1/2 for the two-child problem: the sample space is {bg, gb, gg} (3 outcomes, not 4), so the answer is 1/3.
Application Note
A family has 2 children. Given that at least one child is a girl, what is probability both are girls? S = {bb, bg, gb, gg}. Given B = {bg, gb, gg}, n(B) = 3. Favorable A = {gg}, n(A cap B) = 1. P(A|B) = 1/3.
Derivation Trace Β· Sample Space Reduction Technique for Discrete Experiments
Conditional probability is simply shrinking your universe. If you are told 'the rolled die landed on an EVEN number' (event F = {2, 4, 6}), you throw away {1, 3, 5} forever! Now if someone asks 'what is the chance of rolling a 6?' (event E), it is 1 out of 3, NOT 1 out of 6. The denominator shrank from 6 to 3.
If E and F are two events associated with the same sample space of a random experiment, then the conditional probability of the event E given that F has occurred, i.e. P(E|F) is given by P(E|F) = P(E intersection F) / P(F) provided P(F) != 0. The properties of ordinary probability apply identically to conditional probabilities with respect to the reduced sample space F.
Never divide by the probability of the question event! P(E|F) means 'probability of E given F has happened'. The given event F is the condition, so P(F) is always in the denominator. Writing P(E intersection F) / P(E) is P(F|E), not P(E|F).
Open the Conditional Probability Explorer. Adjust event Venn overlaps to see how P(E|F) scales dynamically as the conditioning set F shrinks or expands.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

A coin is tossed 3 times. Given that the first toss is Heads, what is the probability of getting 3 Heads?

Answer: Outcomes starting with H: {HHH, HHT, HTH, HTT} (4 outcomes). Favorable: {HHH} (1 outcome). P = 1/4.

02

Multiplication Theorem on Probability & Criteria for Independent Events

3 concepts Β· Core
3 nodes

How does the multiplication theorem compute joint probabilities, and what is the rigorous mathematical test for independent events?

The multiplication rule states P(E intersection F) = P(E) P(F|E) = P(F) P(E|F). Events E and F are independent if the occurrence of one does not affect the other: P(E|F) = P(E), yielding the gold standard test P(E intersection F) = P(E) P(F). Pairwise independence does not imply mutual independence for three or more events.

01

The Multiplication Rule: P(A cap B) = P(A) * P(B|A)

The joint probability that both events A and B occur sequentially is P(A cap B) = P(A) * P(B|A). For three events: P(A cap B cap C) = P(A) * P(B|A) * P(C | (A cap B)).

Key Idea
The probability of a compound chain of events is the product of conditional probabilities along the sequential path.
Why It Works
Directly algebraic from P(B|A) = P(A cap B)/P(A); clearing the denominator yields the multiplication theorem.
Boundary Trap
Using constant denominators when sampling without replacement: the total count drops by 1 on each subsequent draw.
Application Note
Draw 2 cards without replacement from a standard 52-card deck. Let A = first card is King, B = second card is King. P(A) = 4/52. Given A occurred, 3 Kings remain among 51 cards: P(B|A) = 3/51. P(both Kings) = (4/52) * (3/51) = 1/221.
Derivation Trace Β· The Multiplication Rule: P(A cap B) = P(A) * P(B|A)
Independent means 'blind to each other'. If you toss a coin and roll a die, the die does not know or care what the coin did! You simply multiply their individual probabilities: P(Head AND 6) = (1/2) * (1/6) = 1/12. But if you draw two cards without replacement, the second draw remembers the first, so they are dependent.
Let E and F be two events associated with a sample space S. By definition of conditional probability, P(E intersection F) = P(E) P(F|E) = P(F) P(E|F). Two events E and F are said to be independent if P(E intersection F) = P(E) P(F). Three events A, B and C are mutually independent if P(A intersection B) = P(A)P(B), P(B intersection C) = P(B)P(C), P(A intersection C) = P(A)P(C), and P(A intersection B intersection C) = P(A)P(B)P(C).
The single most notorious blunder in probability is confusing 'mutually exclusive' with 'independent'! Mutually exclusive means they cannot happen together (P(A intersection B) = 0). Independent means they do not influence each other (P(A intersection B) = P(A)P(B)). Two non-empty events CANNOT be both mutually exclusive and independent!
Open the Conditional Probability Explorer. Set event probabilities to satisfy P(A intersection B) = P(A) * P(B) to confirm that P(A|B) equals P(A) identically.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

An urn has 10 black and 5 white balls. Two balls are drawn without replacement. What is the probability both are black?

Answer: P(B1 cap B2) = (10/15) * (9/14) = (2/3) * (9/14) = 18/42 = 3/7.

02

Mathematical Criterion for Independent Events: P(A cap B) = P(A) * P(B)

Two events A and B are independent if and only if P(A cap B) = P(A) * P(B). If this equality holds, then P(A|B) = P(A) and P(B|A) = P(B).

Key Idea
Independence means knowing that one event occurred provides zero information about whether the other event will occur.
Why It Works
Substitute P(A|B) = P(A) into P(A cap B) = P(B) * P(A|B) to immediately obtain P(A cap B) = P(A) * P(B).
Boundary Trap
Assuming events are independent based on intuitive feelings without testing whether P(A cap B) equals P(A)*P(B).
Application Note
If P(A) = 0.3, P(B) = 0.4, and A and B are independent: P(A cap B) = 0.3 * 0.4 = 0.12. P(A cup B) = 0.3 + 0.4 - 0.12 = 0.58.
Derivation Trace Β· Mathematical Criterion for Independent Events: P(A cap B) = P(A) * P(B)
Independent means 'blind to each other'. If you toss a coin and roll a die, the die does not know or care what the coin did! You simply multiply their individual probabilities: P(Head AND 6) = (1/2) * (1/6) = 1/12. But if you draw two cards without replacement, the second draw remembers the first, so they are dependent.
Let E and F be two events associated with a sample space S. By definition of conditional probability, P(E intersection F) = P(E) P(F|E) = P(F) P(E|F). Two events E and F are said to be independent if P(E intersection F) = P(E) P(F). Three events A, B and C are mutually independent if P(A intersection B) = P(A)P(B), P(B intersection C) = P(B)P(C), P(A intersection C) = P(A)P(C), and P(A intersection B intersection C) = P(A)P(B)P(C).
The single most notorious blunder in probability is confusing 'mutually exclusive' with 'independent'! Mutually exclusive means they cannot happen together (P(A intersection B) = 0). Independent means they do not influence each other (P(A intersection B) = P(A)P(B)). Two non-empty events CANNOT be both mutually exclusive and independent!
Open the Conditional Probability Explorer. Set event probabilities to satisfy P(A intersection B) = P(A) * P(B) to confirm that P(A|B) equals P(A) identically.
πŸ”¬ VIRTUAL LAB Independent Events
Full Stage β†—
? Quick Concept Check

If P(A) = 1/2, P(B) = 1/3, and P(A cap B) = 1/6, are A and B independent?

Answer: P(A)*P(B) = (1/2)*(1/3) = 1/6 = P(A cap B). Yes, they are independent.

03

Independent Events vs Mutually Exclusive Events: Complete Distinction

Mutually exclusive events cannot occur together: A cap B = Phi, so P(A cap B) = 0. Independent events can occur together: P(A cap B) = P(A) * P(B) > 0 (for non-zero probabilities).

Key Idea
Mutually exclusive is a set-disjointness property; independence is an informational/probabilistic property.
Why It Works
If two non-zero events are mutually exclusive, the occurrence of one guarantees the other CANNOT occur (P(A|B) = 0 != P(A)), making them highly dependent!
Boundary Trap
Claiming that mutually exclusive events are independent: if P(A) > 0 and P(B) > 0, they can NEVER be independent!
Application Note
Rolling an odd number {1, 3, 5} and rolling an even number {2, 4, 6} are mutually exclusive (intersection empty), but NOT independent because P(Odd|Even) = 0.
Derivation Trace Β· Independent Events vs Mutually Exclusive Events: Complete Distinction
Independent means 'blind to each other'. If you toss a coin and roll a die, the die does not know or care what the coin did! You simply multiply their individual probabilities: P(Head AND 6) = (1/2) * (1/6) = 1/12. But if you draw two cards without replacement, the second draw remembers the first, so they are dependent.
Let E and F be two events associated with a sample space S. By definition of conditional probability, P(E intersection F) = P(E) P(F|E) = P(F) P(E|F). Two events E and F are said to be independent if P(E intersection F) = P(E) P(F). Three events A, B and C are mutually independent if P(A intersection B) = P(A)P(B), P(B intersection C) = P(B)P(C), P(A intersection C) = P(A)P(C), and P(A intersection B intersection C) = P(A)P(B)P(C).
The single most notorious blunder in probability is confusing 'mutually exclusive' with 'independent'! Mutually exclusive means they cannot happen together (P(A intersection B) = 0). Independent means they do not influence each other (P(A intersection B) = P(A)P(B)). Two non-empty events CANNOT be both mutually exclusive and independent!
Open the Conditional Probability Explorer. Set event probabilities to satisfy P(A intersection B) = P(A) * P(B) to confirm that P(A|B) equals P(A) identically.
πŸ”¬ VIRTUAL LAB Independent Events
Full Stage β†—
? Quick Concept Check

Can two events with non-zero probabilities be both mutually exclusive and independent simultaneously?

Answer: No. Mutually exclusive implies P(A cap B) = 0, while independence requires P(A cap B) = P(A)*P(B) > 0.

03

Partition of Sample Space & The Law of Total Probability

3 concepts Β· Core
3 nodes

How does decomposing a sample space into pairwise disjoint exhaustive partitions reconstruct the total probability of an observed event?

A collection of events {E1, E2, ..., En} forms a partition of S if they are pairwise disjoint (Ei intersection Ej = null for i != j), exhaustive (union Ei = S), and each has P(Ei) > 0. The Theorem on Total Probability states that for any event A in S, P(A) = sum_{i=1}^n P(Ei) P(A|Ei).

01

Mathematical Definition and Criteria for a Partition of Sample Space

A collection of events {E1, E2, ..., En} forms a partition of sample space S if: (1) Ei cap Ej = Phi for all i != j (pairwise disjoint); (2) E1 cup E2 cup ... cup En = S (exhaustive); (3) P(Ei) > 0 for all i.

Key Idea
Every single outcome in the sample space belongs to one and only one of the events Ei.
Why It Works
A partition decomposes the entire probability space into isolated, non-overlapping scenarios that sum to 1.
Boundary Trap
Omitting a category (leaving gaps in S) or allowing overlapping categories when defining partition events.
Application Note
Selecting a card: E1 = spade, E2 = heart, E3 = diamond, E4 = club. The 4 suits are disjoint, their union is the whole deck, and all have probability 1/4. They form a partition.
Derivation Trace Β· Mathematical Definition and Criteria for a Partition of Sample Space
Imagine a car manufacturing plant with 3 assembly lines: Line 1 makes 50% of cars, Line 2 makes 30%, Line 3 makes 20%. Each line has its own defective rate (say 1%, 2%, 3%). If a customer picks a car at random, what is the total chance it is defective? You simply compute: (0.50 * 0.01) + (0.30 * 0.02) + (0.20 * 0.03). You sliced the world into 3 sources and added their contributions.
A set of events E1, E2, ..., En is said to represent a partition of the sample space S if: (a) Ei intersection Ej = null, i != j, i, j = 1, 2, ..., n; (b) E1 union E2 union ... union En = S; (c) P(Ei) > 0 for all i = 1, 2, ..., n. Let {E1, E2, ..., En} be a partition of S, and let A be any event associated with S, then P(A) = sum_{j=1}^n P(Ej) P(A|Ej).
Before applying the total probability formula, make sure the candidate events E1, E2, ..., En actually form a complete partition! Their probabilities MUST sum to 1.0 (P(E1) + P(E2) + ... + P(En) = 1). If they do not add up to 1, some source has been omitted.
Open the Bayes' Tree Explorer. Adjust the branch probabilities of causes E1, E2, E3 to see how the total probability P(A) at the terminal nodes updates dynamically.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

Do events E1 = {number is prime} and E2 = {number is composite} form a partition of the numbers from 1 to 10?

Answer: No, because the number 1 is neither prime nor composite, leaving an uncovered gap in S.

02

The Law of Total Probability: P(A) = sum P(Ei) * P(A|Ei)

Let {E1, E2, ..., En} be a partition of S. Then for any event A of S: P(A) = sum_{i=1}^n P(Ei) * P(A|Ei) = P(E1)P(A|E1) + P(E2)P(A|E2) + ... + P(En)P(A|En).

Key Idea
To find the total probability of an effect A, sum the products of the probability of each cause P(Ei) and the conditional probability of the effect given that cause P(A|Ei).
Why It Works
Since A = A cap S = A cap (cup Ei) = cup (A cap Ei), and the sets are mutually disjoint, P(A) = sum P(A cap Ei) = sum P(Ei) P(A|Ei).
Boundary Trap
Forgetting to multiply the conditional rate P(A|Ei) by the base frequency of the partition P(Ei).
Application Note
Factory has two machines: M1 produces 60% of items (defective rate 2%) and M2 produces 40% of items (defective rate 1%). P(Defective) = P(M1)*P(D|M1) + P(M2)*P(D|M2) = (0.60 * 0.02) + (0.40 * 0.01) = 0.012 + 0.004 = 0.016 (1.6%).
Derivation Trace Β· The Law of Total Probability: P(A) = sum P(Ei) * P(A|Ei)
Imagine a car manufacturing plant with 3 assembly lines: Line 1 makes 50% of cars, Line 2 makes 30%, Line 3 makes 20%. Each line has its own defective rate (say 1%, 2%, 3%). If a customer picks a car at random, what is the total chance it is defective? You simply compute: (0.50 * 0.01) + (0.30 * 0.02) + (0.20 * 0.03). You sliced the world into 3 sources and added their contributions.
A set of events E1, E2, ..., En is said to represent a partition of the sample space S if: (a) Ei intersection Ej = null, i != j, i, j = 1, 2, ..., n; (b) E1 union E2 union ... union En = S; (c) P(Ei) > 0 for all i = 1, 2, ..., n. Let {E1, E2, ..., En} be a partition of S, and let A be any event associated with S, then P(A) = sum_{j=1}^n P(Ej) P(A|Ej).
Before applying the total probability formula, make sure the candidate events E1, E2, ..., En actually form a complete partition! Their probabilities MUST sum to 1.0 (P(E1) + P(E2) + ... + P(En) = 1). If they do not add up to 1, some source has been omitted.
Open the Bayes' Tree Explorer. Adjust the branch probabilities of causes E1, E2, E3 to see how the total probability P(A) at the terminal nodes updates dynamically.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If urn 1 has probability 1/3 with 50% red balls, and urn 2 has probability 2/3 with 20% red balls, what is the total probability of drawing a red ball?

Answer: P(Red) = (1/3)(1/2) + (2/3)(1/5) = 1/6 + 2/15 = 5/30 + 4/30 = 9/30 = 3/10.

03

Multi-Stage Tree Diagrams for Probability Synthesis

A probability tree diagram visually maps sequential experiments: primary branches represent partition events P(Ei), and secondary branches emanating from each Ei represent conditional probabilities P(A|Ei).

Key Idea
Multiply probabilities along a branch path to get joint probability; sum branch path products to get total probability.
Why It Works
Tree structures enforce the multiplication rule along paths and the addition rule across mutually disjoint branches.
Boundary Trap
Adding probabilities along a path instead of multiplying: always multiply along the sequence of events!
Application Note
Branch 1: Bag A (prob 1/2) -> White ball (prob 3/7). Branch 2: Bag B (prob 1/2) -> White ball (prob 5/9). Total P(White) = (1/2)*(3/7) + (1/2)*(5/9).
Derivation Trace Β· Multi-Stage Tree Diagrams for Probability Synthesis
Imagine a car manufacturing plant with 3 assembly lines: Line 1 makes 50% of cars, Line 2 makes 30%, Line 3 makes 20%. Each line has its own defective rate (say 1%, 2%, 3%). If a customer picks a car at random, what is the total chance it is defective? You simply compute: (0.50 * 0.01) + (0.30 * 0.02) + (0.20 * 0.03). You sliced the world into 3 sources and added their contributions.
A set of events E1, E2, ..., En is said to represent a partition of the sample space S if: (a) Ei intersection Ej = null, i != j, i, j = 1, 2, ..., n; (b) E1 union E2 union ... union En = S; (c) P(Ei) > 0 for all i = 1, 2, ..., n. Let {E1, E2, ..., En} be a partition of S, and let A be any event associated with S, then P(A) = sum_{j=1}^n P(Ej) P(A|Ej).
Before applying the total probability formula, make sure the candidate events E1, E2, ..., En actually form a complete partition! Their probabilities MUST sum to 1.0 (P(E1) + P(E2) + ... + P(En) = 1). If they do not add up to 1, some source has been omitted.
Open the Bayes' Tree Explorer. Adjust the branch probabilities of causes E1, E2, E3 to see how the total probability P(A) at the terminal nodes updates dynamically.
πŸ”¬ VIRTUAL LAB Bayes Tree
Full Stage β†—
? Quick Concept Check

In a tree diagram, what must the sum of probabilities of all branches departing from a single node equal?

Answer: The sum must equal exactly 1.

04

Bayes' Theorem & Inverse Probability Diagnosis

3 concepts Β· Core
3 nodes

How does Bayes' theorem invert cause-and-effect conditional probabilities to compute posterior diagnostic probabilities?

Bayes' Theorem inverts cause-and-effect reasoning: given that effect A has been observed, the posterior probability that it arose from cause Ei is P(Ei|A) = [P(Ei) P(A|Ei)] / [sum_{j=1}^n P(Ej) P(A|Ej)]. Prior probabilities P(Ei) are updated to posterior probabilities P(Ei|A) in light of observed evidence.

01

Bayes' Theorem Mathematical Statement and Formula

Let {E1, E2, ..., En} be a partition of sample space S with P(Ei) > 0. For any event A with P(A) > 0, the posterior probability is P(Ei|A) = [P(Ei) * P(A|Ei)] / [sum_{k=1}^n P(Ek) * P(A|Ek)].

Key Idea
The numerator is the specific branch probability P(Ei cap A); the denominator is the Law of Total Probability P(A).
Why It Works
Directly derived by substituting the Law of Total Probability for P(A) into the definition P(Ei|A) = P(Ei cap A) / P(A).
Boundary Trap
Forgetting to include all terms of the partition in the denominator sum.
Application Note
Bag I has 3 red, 4 black balls. Bag II has 5 red, 6 black balls. One bag is chosen at random and a ball drawn is red. Probability it was from Bag I: P(B1|R) = [ (1/2)*(3/7) ] / [ (1/2)*(3/7) + (1/2)*(5/11) ].
Derivation Trace Β· Bayes' Theorem Mathematical Statement and Formula
Forward probability asks: 'If a patient has a disease, what is the chance they test positive?' (easy, given by the lab). Bayes' Theorem answers the real-world medical question: 'A patient just tested positiveβ€”what is the actual chance they HAVE the disease?' It runs the movie backwards by comparing the true-positive path against all possible ways a positive test could happen.
Theorem (Bayes' Theorem): If E1, E2, ..., En are n non-empty events which constitute a partition of sample space S, i.e. E1, E2, ..., En are pairwise disjoint and E1 union ... union En = S and A is any event of non-zero probability, then P(Ei|A) = [P(Ei) P(A|Ei)] / [sum_{j=1}^n P(Ej) P(A|Ej)] for any i = 1, 2, ..., n.
The biggest student mistake in Bayes' Theorem questions is defining event A incorrectly! Event A is NOT the bag or the machine; event A is the SYMPTOM or EVIDENCE that is already observed to have taken place (e.g., 'the ball drawn is red' or 'the person tests positive'). The bags/machines are the partition causes E1, E2, E3.
Open the Bayes' Tree Explorer. Enter base rates and test accuracy to observe how false positives in rare diseases dramatically depress posterior confidence.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

What is the relationship between the denominator of Bayes' theorem and the Law of Total Probability?

Answer: They are identical: the denominator is the total probability P(A) = sum P(Ek)*P(A|Ek).

02

Terminology: Prior Probabilities, Likelihoods, and Posterior Probabilities

Prior probability P(Ei) is the initial probability of hypothesis Ei before evidence is collected. Likelihood P(A|Ei) is the probability of observing evidence A under hypothesis Ei. Posterior probability P(Ei|A) is the updated probability after evidence A is observed.

Key Idea
Posterior is proportional to Prior times Likelihood: P(Ei|A) proportional to P(Ei) * P(A|Ei).
Why It Works
Evidence selectively boosts hypotheses that predict that evidence with high likelihood, weighted by their initial plausibility.
Boundary Trap
Confusing Likelihood P(A|Ei) with Posterior P(Ei|A): they are conditional probabilities in opposite directions!
Application Note
A doctor knows a disease has 1% prevalence in population (Prior = 0.01). A patient tests positive (Evidence A). The doctor calculates the updated probability the patient is sick (Posterior).
Derivation Trace Β· Terminology: Prior Probabilities, Likelihoods, and Posterior Probabilities
Forward probability asks: 'If a patient has a disease, what is the chance they test positive?' (easy, given by the lab). Bayes' Theorem answers the real-world medical question: 'A patient just tested positiveβ€”what is the actual chance they HAVE the disease?' It runs the movie backwards by comparing the true-positive path against all possible ways a positive test could happen.
Theorem (Bayes' Theorem): If E1, E2, ..., En are n non-empty events which constitute a partition of sample space S, i.e. E1, E2, ..., En are pairwise disjoint and E1 union ... union En = S and A is any event of non-zero probability, then P(Ei|A) = [P(Ei) P(A|Ei)] / [sum_{j=1}^n P(Ej) P(A|Ej)] for any i = 1, 2, ..., n.
The biggest student mistake in Bayes' Theorem questions is defining event A incorrectly! Event A is NOT the bag or the machine; event A is the SYMPTOM or EVIDENCE that is already observed to have taken place (e.g., 'the ball drawn is red' or 'the person tests positive'). The bags/machines are the partition causes E1, E2, E3.
Open the Bayes' Tree Explorer. Enter base rates and test accuracy to observe how false positives in rare diseases dramatically depress posterior confidence.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If evidence A is equally likely under all hypotheses, how do posterior probabilities relate to prior probabilities?

Answer: Posterior probabilities will be exactly proportional to prior probabilities (no update occurs).

03

Diagnostic Tests, False Positives, and Industrial Quality Control

Bayes' theorem resolves the 'base rate fallacy': when a disease or manufacturing defect is very rare, even a small false positive rate can cause the vast majority of positive flags to be false alarms.

Key Idea
Always structure into clear partition: E1 = person has disease, E2 = person is healthy; A = test result is positive.
Why It Works
The false positive pool P(E2)*P(A|E2) can easily dwarf the true positive pool P(E1)*P(A|E1) because healthy people E2 are vastly more numerous.
Boundary Trap
Assuming that a 99% accurate test means a positive patient has a 99% chance of being sick: this ignores the low base rate!
Application Note
Disease prevalence 0.1%, test accuracy 99% true positive, 1% false positive. P(D|+) = [0.001 * 0.99] / [0.001*0.99 + 0.999*0.01] = 0.00099 / (0.00099 + 0.00999) = 0.00099 / 0.01098 = 9%. Only 9% of positive testers actually have the disease!
Derivation Trace Β· Diagnostic Tests, False Positives, and Industrial Quality Control
Forward probability asks: 'If a patient has a disease, what is the chance they test positive?' (easy, given by the lab). Bayes' Theorem answers the real-world medical question: 'A patient just tested positiveβ€”what is the actual chance they HAVE the disease?' It runs the movie backwards by comparing the true-positive path against all possible ways a positive test could happen.
Theorem (Bayes' Theorem): If E1, E2, ..., En are n non-empty events which constitute a partition of sample space S, i.e. E1, E2, ..., En are pairwise disjoint and E1 union ... union En = S and A is any event of non-zero probability, then P(Ei|A) = [P(Ei) P(A|Ei)] / [sum_{j=1}^n P(Ej) P(A|Ej)] for any i = 1, 2, ..., n.
The biggest student mistake in Bayes' Theorem questions is defining event A incorrectly! Event A is NOT the bag or the machine; event A is the SYMPTOM or EVIDENCE that is already observed to have taken place (e.g., 'the ball drawn is red' or 'the person tests positive'). The bags/machines are the partition causes E1, E2, E3.
Open the Bayes' Tree Explorer. Enter base rates and test accuracy to observe how false positives in rare diseases dramatically depress posterior confidence.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

In medical screening for a rare disease, why is the posterior probability often much lower than the test's sensitivity?

Answer: Because the rare base rate means the small percentage of false positives from the large healthy majority outnumbers the true positives.

05

Random Variables, Probability Distributions, Expectation (Mean) & Variance

3 concepts Β· Core
3 nodes

How are real-valued random variables mapped to probability distributions to calculate expected values (mean) and variance?

A random variable X is a real-valued function X: S -> R. Its probability distribution lists pairs (xi, pi) with pi >= 0 and sum pi = 1. The expectation (mean) is mu = E(X) = sum xi pi. The variance is Var(X) = sigma^2 = E(X^2) - [E(X)]^2 = sum xi^2 pi - mu^2, and standard deviation is sigma = +sqrt(Var(X)).

01

Random Variable Definition and Valid Probability Distribution Table

A random variable X maps sample points to real numbers. A discrete probability distribution lists values x_i alongside probabilities p_i = P(X = x_i). It is valid if and only if: (1) p_i >= 0 for all i; and (2) sum_{i=1}^n p_i = 1.

Key Idea
The probability distribution table completely characterizes the stochastic behavior of the discrete random variable.
Why It Works
Probabilities represent proportions of the total universal probability 1; negative probabilities or sums != 1 violate Kolmogorov axioms.
Boundary Trap
Accepting negative probabilities in tables when solving for an unknown parameter k: probabilities can never be negative!
Application Note
Toss 2 coins. Let X = number of heads. S = {TT, HT, TH, HH}. X can be 0, 1, 2. P(X=0) = 1/4, P(X=1) = 2/4 = 1/2, P(X=2) = 1/4. Sum = 1/4 + 1/2 + 1/4 = 1.
Derivation Trace Β· Random Variable Definition and Valid Probability Distribution Table
A random variable is simply an accountant that converts messy qualitative outcomes (like 'Heads-Tails-Heads') into numbers (like '2 heads'). Once you have a table of numbers X and their probabilities P(X), the Mean E(X) is where you would place your finger to balance the probability seesaw, and the Variance measures how spread out the weights are.
A random variable is a real valued function whose domain is the sample space of a random experiment. The probability distribution of a random variable X is the system of numbers X: x1, x2, ..., xn with P(X): p1, p2, ..., pn where pi > 0 and sum pi = 1. The mean or expectation of X is E(X) = sum xi pi. The variance of X is Var(X) = E(X - mu)^2 = E(X^2) - [E(X)]^2.
Variance can NEVER be negative! If your arithmetic yields a negative variance, you accidentally subtracted in the wrong order or made a sign error. Remember: E(X^2) is ALWAYS greater than or equal to [E(X)]^2, ensuring Var(X) >= 0.
Open the Random Variables Explorer. Modify values and probabilities to see the center-of-mass balance line E(X) shift along the real axis.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If P(X=1) = k, P(X=2) = 2k, P(X=3) = 3k, and these are all possible values, what is k?

Answer: k + 2k + 3k = 1 => 6k = 1 => k = 1/6.

02

Mathematical Expectation (Mean): mu = E(X) = sum x_i * p_i

The mean or expectation of a discrete random variable X, denoted mu or E(X), is the weighted average of its possible values: mu = E(X) = sum_{i=1}^n x_i * p_i.

Key Idea
E(X) represents the long-run theoretical average value of X over an infinite number of independent repetitions.
Why It Works
Weighting each numerical value x_i by its relative frequency of occurrence p_i computes the expected center of mass of the distribution.
Boundary Trap
Assuming E(X) must be one of the possible values of X: E(X) is an average and can be a fraction (e.g. 3.5 on a 6-sided die).
Application Note
For die roll X in {1, 2, 3, 4, 5, 6} with each p_i = 1/6: E(X) = (1+2+3+4+5+6)/6 = 21/6 = 3.5. (Note: expected value does not have to be an integer!).
Derivation Trace Β· Mathematical Expectation (Mean): mu = E(X) = sum x_i * p_i
A random variable is simply an accountant that converts messy qualitative outcomes (like 'Heads-Tails-Heads') into numbers (like '2 heads'). Once you have a table of numbers X and their probabilities P(X), the Mean E(X) is where you would place your finger to balance the probability seesaw, and the Variance measures how spread out the weights are.
A random variable is a real valued function whose domain is the sample space of a random experiment. The probability distribution of a random variable X is the system of numbers X: x1, x2, ..., xn with P(X): p1, p2, ..., pn where pi > 0 and sum pi = 1. The mean or expectation of X is E(X) = sum xi pi. The variance of X is Var(X) = E(X - mu)^2 = E(X^2) - [E(X)]^2.
Variance can NEVER be negative! If your arithmetic yields a negative variance, you accidentally subtracted in the wrong order or made a sign error. Remember: E(X^2) is ALWAYS greater than or equal to [E(X)]^2, ensuring Var(X) >= 0.
Open the Random Variables Explorer. Modify values and probabilities to see the center-of-mass balance line E(X) shift along the real axis.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

If X takes values 0 and 1 with probabilities 0.4 and 0.6 respectively, what is E(X)?

Answer: E(X) = 0*(0.4) + 1*(0.6) = 0.6.

03

Variance and Standard Deviation: Var(X) = E(X^2) - [E(X)]^2

Variance measures dispersion around the mean: Var(X) = sigma^2 = E[(X - mu)^2] = sum (x_i - mu)^2 * p_i = E(X^2) - [E(X)]^2, where E(X^2) = sum x_i^2 * p_i. Standard deviation is sigma = sqrt(Var(X)).

Key Idea
Use the computational formula Var(X) = E(X^2) - [E(X)]^2; it is far faster and avoids dealing with squared decimals of (x_i - mu).
Why It Works
Subtracting [E(X)]^2 from E(X^2) expands the quadratic difference sum (x_i - mu)^2 p_i using linearity of expectation.
Boundary Trap
Subtracting E(X^2) from [E(X)]^2 or ending up with a negative variance: variance must ALWAYS be >= 0!
Application Note
If E(X) = 2 and E(X^2) = 6: Var(X) = 6 - (2)^2 = 6 - 4 = 2. Standard deviation sigma = sqrt(2).
Derivation Trace Β· Variance and Standard Deviation: Var(X) = E(X^2) - [E(X)]^2
A random variable is simply an accountant that converts messy qualitative outcomes (like 'Heads-Tails-Heads') into numbers (like '2 heads'). Once you have a table of numbers X and their probabilities P(X), the Mean E(X) is where you would place your finger to balance the probability seesaw, and the Variance measures how spread out the weights are.
A random variable is a real valued function whose domain is the sample space of a random experiment. The probability distribution of a random variable X is the system of numbers X: x1, x2, ..., xn with P(X): p1, p2, ..., pn where pi > 0 and sum pi = 1. The mean or expectation of X is E(X) = sum xi pi. The variance of X is Var(X) = E(X - mu)^2 = E(X^2) - [E(X)]^2.
Variance can NEVER be negative! If your arithmetic yields a negative variance, you accidentally subtracted in the wrong order or made a sign error. Remember: E(X^2) is ALWAYS greater than or equal to [E(X)]^2, ensuring Var(X) >= 0.
Open the Random Variables Explorer. Modify values and probabilities to see the center-of-mass balance line E(X) shift along the real axis.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

Can the variance of a random variable ever be negative?

Answer: No. Variance is the expected value of a squared real quantity, so Var(X) >= 0 always.

06

Bernoulli Trials, Binomial Distribution B(n, p) & Parameter Relationships

3 concepts Β· Core
3 nodes

How does repeating independent Bernoulli trials generate the Binomial distribution B(n, p) with mean n*p and variance n*p*q?

Bernoulli trials are n independent, repeated trials with exactly two outcomes (success with probability p, failure with q = 1 - p). The number of successes X follows a Binomial distribution B(n, p) with probability mass function P(X = r) = nCr p^r q^(n-r) for r = 0, 1, ..., n. Its mean is mu = np and variance is sigma^2 = npq, with sigma^2 < mu always.

01

Bernoulli Trials Definition and The Four Essential Criteria

Trials of a random experiment are Bernoulli Trials if and only if: (1) the number of trials n is finite; (2) trials are independent; (3) each trial yields only two outcomes (Success S or Failure F); (4) probability of success p is constant across all trials.

Key Idea
Drawing with replacement creates Bernoulli trials; drawing without replacement violates independence and constant p.
Why It Works
Without replacement, the composition of the population changes on each draw, changing p and coupling the trials.
Boundary Trap
Treating draws without replacement as Bernoulli trials: without replacement, trials are NOT independent!
Application Note
Tossing a fair coin 10 times: n = 10 (finite), tosses are independent, outcome is Head or Tail (binary), p = 1/2 is constant. These are Bernoulli trials.
Derivation Trace Β· Bernoulli Trials Definition and The Four Essential Criteria
Think of shooting 10 basketball free throws. Each shot is either IN (success) or OUT (failure). If each shot has a 70% chance (p = 0.7) and past shots don't shake your confidence (independent), you have a Binomial distribution B(10, 0.7). The formula nCr p^r q^(10-r) counts all the different combinations of makes and misses that give exactly r baskets.
Trials of a random experiment are called Bernoulli trials if: (i) there are a finite number of trials, (ii) trials are independent, (iii) each trial has only two outcomes, (iv) probability of success remains constant. A random variable X taking values 0, 1, ..., n with P(X = r) = nCr p^r q^(n-r) is said to follow a Binomial distribution with parameters n and p, denoted by B(n, p).
Watch for the phrasing 'at least r successes' vs 'at most r successes'! 'At least 1 success' means P(X >= 1) = 1 - P(X = 0) = 1 - q^n (vastly faster than summing all r from 1 to n!). 'At most 1 success' means P(X <= 1) = P(X = 0) + P(X = 1).
Open the Binomial Distribution Explorer. Vary n from 1 to 20 and p from 0.1 to 0.9 to see the histogram transition from skewed to a symmetric bell curve.
πŸ”¬ VIRTUAL LAB Bayes Tree
Full Stage β†—
? Quick Concept Check

Are 5 draws of cards from a deck WITHOUT replacement Bernoulli trials?

Answer: No, because without replacement the probability of success changes after each draw.

02

Binomial Probability Formula: P(X = r) = nCr * p^r * q^(n - r)

In n Bernoulli trials with success probability p and failure probability q = 1 - p, the probability of exactly r successes is P(X = r) = nCr * p^r * q^(n - r), where r = 0, 1, 2, ..., n.

Key Idea
nCr counts the number of distinct ways to arrange r successes among n trials; p^r * q^(n-r) is the probability of any single such sequence.
Why It Works
By independence, any sequence with r successes and (n - r) failures has probability p^r q^(n-r). Since there are nCr mutually disjoint sequences, their probabilities sum.
Boundary Trap
Omitting the combination coefficient nCr and writing only p^r * q^(n-r), which calculates only one specific ordering!
Application Note
Toss a fair coin 5 times. Probability of exactly 3 heads: n=5, r=3, p=1/2, q=1/2. P(X=3) = 5C3 * (1/2)^3 * (1/2)^2 = 10 * (1/32) = 10/32 = 5/16.
Derivation Trace Β· Binomial Probability Formula: P(X = r) = nCr * p^r * q^(n - r)
Think of shooting 10 basketball free throws. Each shot is either IN (success) or OUT (failure). If each shot has a 70% chance (p = 0.7) and past shots don't shake your confidence (independent), you have a Binomial distribution B(10, 0.7). The formula nCr p^r q^(10-r) counts all the different combinations of makes and misses that give exactly r baskets.
Trials of a random experiment are called Bernoulli trials if: (i) there are a finite number of trials, (ii) trials are independent, (iii) each trial has only two outcomes, (iv) probability of success remains constant. A random variable X taking values 0, 1, ..., n with P(X = r) = nCr p^r q^(n-r) is said to follow a Binomial distribution with parameters n and p, denoted by B(n, p).
Watch for the phrasing 'at least r successes' vs 'at most r successes'! 'At least 1 success' means P(X >= 1) = 1 - P(X = 0) = 1 - q^n (vastly faster than summing all r from 1 to n!). 'At most 1 success' means P(X <= 1) = P(X = 0) + P(X = 1).
Open the Binomial Distribution Explorer. Vary n from 1 to 20 and p from 0.1 to 0.9 to see the histogram transition from skewed to a symmetric bell curve.
πŸ”¬ VIRTUAL LAB Conditional Probability
Full Stage β†—
? Quick Concept Check

In 4 tosses of a fair coin, what is the probability of getting no heads (r = 0)?

Answer: P(X = 0) = 4C0 * (1/2)^0 * (1/2)^4 = 1 * 1 * (1/16) = 1/16.

03

Mean (np), Variance (npq), and Cumulative Queries ('At Least', 'At Most')

For X ~ B(n, p): Mean = n*p, Variance = n*p*q (where q = 1 - p). Cumulative queries: 'At least r successes' = P(X >= r) = sum_{k=r}^n P(X = k); 'At most r successes' = P(X <= r) = sum_{k=0}^r P(X = k). Note: P(X >= 1) = 1 - P(X = 0).

Key Idea
Use the complement rule P(at least 1 success) = 1 - P(X = 0) = 1 - q^n to solve 'at least one' problems in seconds.
Why It Works
Since variance is np(1 - p) and 0 < p < 1, the factor (1 - p) < 1 guarantees that Variance < Mean for all non-degenerate binomial distributions.
Boundary Trap
Summing terms manually for P(X >= 1) instead of using the instant complement 1 - P(X = 0).
Application Note
If X ~ B(10, 0.5): Mean = 10 * 0.5 = 5. Variance = 10 * 0.5 * 0.5 = 2.5. P(at least one success) = 1 - (0.5)^10 = 1 - 1/1024 = 1023/1024.
Derivation Trace Β· Mean (np), Variance (npq), and Cumulative Queries ('At Least', 'At Most')
Think of shooting 10 basketball free throws. Each shot is either IN (success) or OUT (failure). If each shot has a 70% chance (p = 0.7) and past shots don't shake your confidence (independent), you have a Binomial distribution B(10, 0.7). The formula nCr p^r q^(10-r) counts all the different combinations of makes and misses that give exactly r baskets.
Trials of a random experiment are called Bernoulli trials if: (i) there are a finite number of trials, (ii) trials are independent, (iii) each trial has only two outcomes, (iv) probability of success remains constant. A random variable X taking values 0, 1, ..., n with P(X = r) = nCr p^r q^(n-r) is said to follow a Binomial distribution with parameters n and p, denoted by B(n, p).
Watch for the phrasing 'at least r successes' vs 'at most r successes'! 'At least 1 success' means P(X >= 1) = 1 - P(X = 0) = 1 - q^n (vastly faster than summing all r from 1 to n!). 'At most 1 success' means P(X <= 1) = P(X = 0) + P(X = 1).
Open the Binomial Distribution Explorer. Vary n from 1 to 20 and p from 0.1 to 0.9 to see the histogram transition from skewed to a symmetric bell curve.
πŸ”¬ VIRTUAL LAB Bayes Tree
Full Stage β†—
? Quick Concept Check

Can a binomial distribution have Mean = 6 and Variance = 9?

Answer: No. In any Binomial distribution, Variance = npq = Mean * q. Since q <= 1, Variance can never exceed the Mean.