Econ 131 – Quantitative Economics (a.k.a. Introductory Econometrics) Review of probability concepts Instructor: Anthony G. Sabarillo University of the Philippines School of Economics Not for distribution. For lecture purposes only. 1 Outline • Basic probability concepts • Theorems of probability • Calculating probabilities and counting techniques • Conditional probability • Theorem of total probabilities, Bayes’ theorem, and the multiplication rule • Statistical independence Not for distribution. For lecture purposes only. 2 Basic probability concepts • Probability theory • the mathematics of randomness • concerned with quantifying the exact or estimated chance that a random event will occur • Experiment • a process whose outcome is not known in advance with certainty • outcomes are “mutually exclusive potential results” of an experiment (Stock and Watson 2020, p. 56) • can be repeated, theoretically, an infinite number of times • has a well-defined set of possible outcomes • e.g., throwing a die • Is measuring a person’s blood pressure an experiment? Not for distribution. For lecture purposes only. 3 Basic probability concepts • Sample space • set of all possible outcomes of the experiment • What is the sample space of the experiment of • ….throwing a six-faced die? • …tossing a coin? • Event • a subset of the sample space • e.g., an odd outcome from throwing a die Not for distribution. For lecture purposes only. 4 Basic probability concepts • Let π΄π΄ and π΅π΅ be any two events that are subsets of the same sample space ππ (i.e., π΄π΄ ⊆ ππ and π΅π΅ ⊆ ππ). • (digression: What does the ff. mean: π΄π΄ ⊂ ππ?) • Define the following events or sets: • π΄π΄ ∩ π΅π΅ • π΄π΄ ∪ π΅π΅ • π΄π΄πΆπΆ • What does it mean when π΄π΄ ∩ π΅π΅ = ∅? • De Morgan’s Law π΄π΄ ∪ π΅π΅ πΆπΆ = π΄π΄πΆπΆ ∩ π΅π΅πΆπΆ ππππ π΄π΄ ∩ π΅π΅ πΆπΆ = π΄π΄πΆπΆ ∪ π΅π΅πΆπΆ Not for distribution. For lecture purposes only. 5 Basic probability concepts • Definition of probability. • “the proportion of the time that the outcome occurs in the long run” (Stock and Watson 2020, p. 56) • Formally: Let π΄π΄ be an event in a sample space ππ. The probability of the event π΄π΄, denoted by ππ(π΄π΄), is measured as the proportion of times that event π΄π΄ will occur in repeated trials of an experiment. • Approximating probability using relative frequency: • Suppose that in a total of ππ possible equally likely outcomes of an experiment, ππ of them are favorable to the occurrence of the event π΄π΄; • then ratio ππ/ππ is the relative frequency of π΄π΄. • For a large ππ, the relative frequency will provide a very good approximation of the probability of π΄π΄. Not for distribution. For lecture purposes only. 6 Basic probability concepts • Probability ≡ a nonnegative number assigned to every event • Axioms of probability • ππ π΄π΄ ≥ 0 ∀ event π΄π΄ • ππ ππ = 1, ππ ≡ sample space Not for distribution. For lecture purposes only. 7 Basic probability concepts • Classical probability • If a sample space has a finite # of simple events with equal probabilities: number of outcomes in π΄π΄ ππ π΄π΄ = number of outcomes in ππ • Empirical Probability • We can estimate the probability of an event A by repeating the same experiment over and over (say n times) and observing the number of times the event occurs. • Other basic probability concepts Not for distribution. For lecture purposes only. 8 Calculating probabilities • When the sample space contains a finite number of simple events with equal probabilities, • and when a composite event is composed of mutually exclusive simple events: • ππ(composite event)=∑ ππ(mutually exclusive simple events) • Recall: If a sample space has a finite # of simple events with equal probabilities, • ππ π΄π΄ = number of outcomes in π΄π΄ number of outcomes in ππ • Exercise: • Consider the experiment of tossing a coin twice. What is the probability of getting at least one tail? Not for distribution. For lecture purposes only. 9 Counting methods • Multiplication rule • Permutation • Combination Not for distribution. For lecture purposes only. 10 The Multiplication Rule • if a task is done in two stages and if the first stage can be done in ππ ways and for each way that the first stage can be done, the second stage can be done in ππ ways, then the two together can be done in ππ × ππ ways. • Example 1: Suppose that a coin is tossed then a six-sided dice is rolled. Since there are two possible outcomes in the first activity and six possible outcomes in the second activity, the number of possible outcomes is ________. • Example 2: Suppose you’re given the ff. menu: • Meal: Chicken, pork, beef • Drink: Soda, coffee, tea • How many meal-drink combos can you order? Not for distribution. For lecture purposes only. 11 Permutation • The number of possible arrangements in a set when the order of the arrangements matters. • The number of permutations of ππ elements taken ππ at a time is given by the ff. formula: ππππππ = ππ! ππ − ππ ! • In other words: This is the number of ways we can take ππ objects from ππ choices (where the order matters). • Example: In how many ways can you arrange 3 letters taken from the word “ALMOST”? Not for distribution. For lecture purposes only. 12 Combination • number of ways to form a collection of unordered elements. • The number of ways to form combinations of size ππ from a set of ππ distinct objects, without repetition, is: ππ! ππ ππ πΆπΆππ = = ππ ππ − ππ ! ππ! • In other words: The combination of ππ taken ππ is the number of ways we can take ππ objects from ππ choices, where the order does not matter. • Example: (Larsen and Marx) There are nine students, five of whom are men and four of whom are women, who interviewed for four summer internships sponsored by a city newspaper. In how many ways can the newspaper choose a set of four interns? Not for distribution. For lecture purposes only. 13 Conditional Probability • Given two events π΄π΄ and π΅π΅, the probability of π΄π΄ given that π΅π΅ has occurred is the conditional probability of π΄π΄ given π΅π΅. • Notation: ππ(π΄π΄|π΅π΅) • More formally: • For events π΄π΄ & π΅π΅ where ππ(π΅π΅) > 0, • ππ π΄π΄ π΅π΅ ππ(π΄π΄∩π΅π΅) = ππ(π΅π΅) Not for distribution. For lecture purposes only. 14 Conditional Probability • Example: • Experiment: tossing a fair coin twice • Find: the probability of two heads occurring given that at least one head has occurred Not for distribution. For lecture purposes only. 15 Statistical independence Def. Independence of events Events π΄π΄ and π΅π΅ are independent (or pairwise independent or statistically independent) if and only if any one of the following conditions is satisfied: 1. ππ(π΄π΄ ∩ π΅π΅) = ππ(π΄π΄)ππ(π΅π΅) 2. ππ π΄π΄ π΅π΅ = ππ(π΄π΄) if ππ(π΅π΅) > 0 3. ππ(π΅π΅|π΄π΄) = ππ(π΅π΅) if ππ(π΄π΄) > 0 Not for distribution. For lecture purposes only. 16 Statistical independence • Example: Suppose you toss two dice. Let π΄π΄ denote the event of a 1 on the second die, π΅π΅ the event of getting a 1 on the first die. Are π΄π΄ and π΅π΅ independent? Not for distribution. For lecture purposes only. 17 Statistical independence • Def. Events π΄π΄, π΅π΅, and πΆπΆ are mutually independent if and only if • 1) ππ π΄π΄ ∩ π΅π΅ = ππ π΄π΄ ππ(π΅π΅) and • 2) ππ π΄π΄ ∩ πΆπΆ = ππ π΄π΄ ππ(πΆπΆ) and • 3) ππ π΅π΅ ∩ πΆπΆ = ππ π΅π΅ ππ(πΆπΆ) and • 4) ππ π΄π΄ ∩ π΅π΅ ∩ πΆπΆ = ππ π΄π΄ ππ(π΅π΅)ππ(πΆπΆ). Not for distribution. For lecture purposes only. 18 Statistical independence • Example • Experiment: toss a fair coin twice • π»π»π»π», π»π»π»π», ππππ, ππππ • Let π΄π΄ ≡ π»π» on the first toss • Let π΅π΅ ≡ π»π» on the second toss • Let πΆπΆ ≡ π»π»π»π» or ππππ • Determine if π΄π΄, π΅π΅, and πΆπΆ are mutually independent. Not for distribution. For lecture purposes only. 19 Statistical independence Def. Independence of several events Events π΄π΄1 , π΄π΄2 , … , π΄π΄ππ are mutually independent if and only if • any proper subset of events are mutually independent, and • ππ π΄π΄1 ∩ π΄π΄2 ∩ β― ∩ π΄π΄ππ = ππ π΄π΄1 ππ π΄π΄2 … ππ π΄π΄ππ Not for distribution. For lecture purposes only. 20 Appendix Not for distribution. For lecture purposes only. 21 Basic probability concepts (back) • 1. ππ ∅ = 0 • 2. If π΄π΄1 , π΄π΄2 , … , π΄π΄ππ are mutually exclusive events, then ππ π΄π΄1 ∪ π΄π΄2 ∪ β― ∪ π΄π΄ππ = ππ π΄π΄1 + ππ π΄π΄2 + β― + ππ(π΄π΄ππ ) • 3. ππ π΄π΄ = 1 − ππ π΄π΄πΆπΆ • 4. For any two events π΄π΄ and π΅π΅, • ππ π΄π΄ = ππ π΄π΄ ∩ π΅π΅ + ππ π΄π΄ ∩ π΅π΅πΆπΆ • ππ π΄π΄ − π΅π΅ = ππ π΄π΄ − ππ π΅π΅ • 5. ππ π΄π΄ ∪ π΅π΅ = ππ π΄π΄ + ππ π΅π΅ − ππ(π΄π΄ ∩ π΅π΅) • 6. For any two events π΄π΄ and π΅π΅, • if π΄π΄ ⊂ π΅π΅ then ππ π΄π΄ < ππ(π΅π΅) Not for distribution. For lecture purposes only. 22