Winter 2013 STAT 302 Chapter 3 Conditional Probability and Independence In-class activity The odds that a randomly chosen individual has a circulation problem is 0.25. Individuals who have a circulation problem are three times as likely to be smokers as those who do not have a circulation problem. Given that an individual is a smoker, what is the probability that he/she has a circulation problem? Eugenia Yu, UBC Department of Statistics 1 Winter 2013 STAT 302 1. What is the probability that a randomly chosen individual has a circulation problem? A. B. C. D. E. 0.10 0.20 0.25 0.33 0.50 Eugenia Yu, UBC Department of Statistics 2 Winter 2013 STAT 302 2. Suppose we define the following events: C = an individual has a circulation problem T = an individual is a smoker Which of the following is correct? A. B. C. D. E. P (T ) = 3P (T c) P (C|T ) = 3P (C|T c) P (T |C) = 3P (T |C c) P (T ∩ C) = 3P (T c ∩ C) P (T ∩ C) = 3P (T ∩ C c) Eugenia Yu, UBC Department of Statistics 3 Winter 2013 STAT 302 3. Given that an individual is a smoker, what is the probability that he/she has a circulation problem? A. B. C. D. E. 3/7 = 0.4285 3/4 = 0.75 1/13 = 0.0769 2/3 = 0.6667 5/9 = 0.5556 Eugenia Yu, UBC Department of Statistics 4 Winter 2013 STAT 302 Let’s look at the solution ..... Eugenia Yu, UBC Department of Statistics 5 Winter 2013 STAT 302 We define the following events: C = an individual has a circulation problem T = an individual is a smoker P (C) = 0.25 1+0.25 = 0.20 P (T |C) = 3P (T |C c) P (C|T ) = = = = P (T |C)P (C) P (T |C)P (C) + P (T |C c)P (C c) P (T |C)P (C) P (T |C)P (C) + (1/3)P (T |C)[1 − P (C)] P (C) P (C) + (1/3)[1 − P (C)] 0.20 0.20 + (1/3)[1 − 0.20] = 3/7 Eugenia Yu, UBC Department of Statistics 6 Winter 2013 STAT 302 Conditional probability Given that an event F has occurred [P (F ) > 0], the probability that another event E occurs is called the conditional probability of E given F . Notation and formula: P (E ∩ F ) P (E|F ) = P (F ) Eugenia Yu, UBC Department of Statistics 7 Winter 2013 STAT 302 Note: 1. P (E|F ) 6= P (F |E) 2. P (E c ∩ F ) = P (F ) − P (E ∩ F ) 3. P (E c|F ) = 1 − P (E|F ) Proof : P (E c|F ) = = P (E c ∩ F ) P (F ) P (F ) − P (E ∩ F ) P (F ) P (E ∩ F ) = 1− P (F ) = 1 − P (E|F ) Eugenia Yu, UBC Department of Statistics 8 Winter 2013 STAT 302 Properties of P (E|F ) 1. 0 ≤ P (E|F ) ≤ 1 Proof: P (E|F ) = P (E∩F ) P (F ) Here P (E ∩ F ) ≥ 0 and P (F ) > 0, so P (E|F ) ≥ 0 Also, {E ∩ F } ⊂ F ⇒ P (E ∩ F ) ≤ P (F ) , so P (E|F ) ≤ 1 2. P (F |F ) = 1; P (S|F ) = 1 Proof : P (F |F ) = P (F ∩ F ) P (F ) = =1 P (F ) P (F ) P (S|F ) = P (S ∩ F ) P (F ) = =1 P (F ) P (F ) Eugenia Yu, UBC Department of Statistics 9 Winter 2013 STAT 302 3. For a sequence of mutually exclusive events E1, E2, ...(Ei ∩ Ej = φ for all i 6= j), ∞ ∞ [ X P Ei|F = P (Ei|F ) i=1 Proof : P ∞ [ Ei|F = i=1 P S∞ i=1 Ei S∞ i=1 (Ei ∩ F ) P (F ) P P∞ = ∩F P (F ) i=1 = i=1 P (Ei by the Distributive law ∩ F) P (F ) since {Ei ∩ F }0s are mutually exclusive ∞ X = P (Ei|F ) i=1 Eugenia Yu, UBC Department of Statistics 10 Winter 2013 STAT 302 Bayes’ formula Total probability: P (E) = Pn i=1 P (E|Fi )P (Fi ) The Bayes’ formula: P (Fi|E) = P (Fi ∩ E) P (E|Fi)P (Fi) = Pn P (E) j=1 P (E|Fj )P (Fj ) Eugenia Yu, UBC Department of Statistics 11 Winter 2013 STAT 302 Odds of an event • The odds of an event E is defined as the ratio of the probability of an event E to the probability of its complement P (E) P (E) = P (E c) 1 − P (E) • The odds tells how much more likely E occurs compared to its complement. – if odds > 1, E is more likely than E c – if odds < 1, E is less likely than E c – if odds = 1, E and E c are equally likely Eugenia Yu, UBC Department of Statistics 12 Winter 2013 STAT 302 • Suppose the odds of E is α, then P (E) = αP (E c) P (E) + P (E c) = 1 αP (E c) + P (E c) = 1 P (E c) = P (E) = Eugenia Yu, UBC Department of Statistics 1 1+α α 1+α 13