Chapter 5: Discrete Probability Distributions

Learning Objectives (LOs)

  • LO 5.1: Describe a discrete random variable and its probability distribution.

  • LO 5.2: Calculate and interpret summary measures for a discrete random variable.

  • LO 5.3: Calculate and interpret probabilities for a binomial random variable.

  • LO 5.4: Calculate and interpret probabilities for a Poisson random variable.

  • LO 5.5: Calculate and interpret probabilities for a hypergeometric random variable.

Introductory Case: Available Staff for Probable Customers

  • Context: Anne Jones, Starbucks manager, is concerned about competition influencing business.

    • Typical Starbucks customers visit 15 to 18 times a month.

    • Anne believes loyal customers visit an average of 18 times in a 30-day month.

  • Decisions to Be Made:

    1. Calculate expected visits by a typical Starbucks customer.

    2. Calculate the probability of a typical customer visiting the chain a specified number of times.

5.1 Random Variables and Discrete Probability Distributions

  • A random variable is a function that assigns numerical values to outcomes of an experiment concisely summarizing uncertainty.

  • Distinct Types:

    • Discrete Random Variable: Assumes a countable number of distinct values, denoted by the letter X.

    • Example: Number of employees in a firm.

    • Continuous Random Variable: Characterized by uncountable values within an interval.

    • Example: Return on a mutual fund.

  • Probability Distribution:

    • Each discrete random variable holds an associated probability distribution known as a probability mass function.

    • Properties of the probability distribution:

    • Each probability value (for any discrete outcome x) lies between 0 and 1.

    • The sum of all probabilities equals 1.

  • A discrete random variable can be represented with a cumulative probability distribution.

Representation of Probability Distribution

  • Example: Rolling a die (Discrete Uniform Distribution)

    • Characteristics:

    • Finite number of outcomes.

    • Each outcome has an equal likelihood.

    • Symmetrical.

  • Tabular Representation:

    • For rolling die:

    • Outcomes (x): 1, 2, 3, 4, 5, 6

    • Probability P(X = x): 1/6

    • Cumulative Probability:

    • Outcomes (x): 1, 2, 3, 4, 5, 6

    • Cumulative Probabilities P(X ≤ x): 1/6, 2/6, 3/6, 4/6, 5/6, 6/6

  • Graphical and Algebraic Representation: Further detail on graphical and probability functions is elaborated in the text but not discretely transcribed here.

5.2 Expected Value, Variance, and Standard Deviation


  • Expected Value (Mean):

    • Denoted by $ar{X}$:

    • It represents a weighted average of all possible values of the random variable X indicating the central tendency.


  • Variance ($ ext{Var}(X)$) and Standard Deviation ($ ext{SD}(X)$):

    • These measures indicate clustering or scattering of values in distribution about the mean.

    • Definitions:

    • Variance is denoted as $ ext{Var}(X)$, while standard deviation is $ ext{SD}(X)$.


  • Example (Car Dealership Compensation):



    • Values (Bonus in $1,000s) and Probabilities:

      Bonus

      Probability


      10

      0.15


      6

      0.25


      3

      0.40


      0

      0.20

      • a. Calculate the expected value of bonuses.

      • b. Calculate variance and standard deviation.

      • c. Compute total bonuses expected if there are 25 employees.

      • Detailed calculations for expected value, variance, and total bonuses based on outcomes delivered in original context.

      5.3 The Binomial Distribution

      • The Binomial Distribution applies when conditions of a Bernoulli process are satisfied, defined as a series of independent trials with:

        • Two potential outcomes (success and failure).

        • Constant probabilities of success (p) and failure (1−p) across trials.

      • A binomial random variable (X) captures the number of successful outcomes within n trials.

        • Examples: Customer loan defaults, consumer reactions, drug efficacy, graduate school applications.

      • Probability Function for binomial random variable X for obtaining x successes within n trials is expressed as:

      P(X = x) = {n ext{ choose } x} p^x (1-p)^{n-x}

      • Key components:

        1. Combinatorial term to determine the number of possible sequences with x successes.

        2. Probabilistic term for any particular sequence.

      • Summary Measures:

        • Mean: $ ext{E}(X) = n imes p$

        • Variance: $ ext{Var}(X) = n imes p imes (1 - p)$

        • Standard Deviation: $ ext{SD}(X) = ext{sqrt}(n imes p imes (1 - p))$

      • Example Calculation:

        • Determine probabilities regarding customer reactions with a historical success rate of 30% across a sample of 5 customers.

        • Probability of specific successes within trials computed via binomial probability functions.

      5.4 The Poisson Distribution

      • A Poisson Process is characterized by:

        • The counting of successes in a given time/space interval.

        • Independence across non-overlapping intervals.

        • Constant probability of success throughout all intervals.

      • Poisson Random Variable (X) can be represented mathematically as:

      P(X = x) = rac{e^{- ext{λ}} ext{λ}^x}{x!}
      where λ is the mean number of successes in the interval.

      • Example: Craft breweries opening daily, treating as a Poisson process.

        • Calculate expected breweries weekly and specific probabilities (like 12 openings in a week) using lambda as a proportion of time periods.

      5.5 The Hypergeometric Distribution

      • The Hypergeometric Distribution differs from the binomial in terms of dependency in trials when sampling without replacement.

        • Hypergeometric scenarios, changes in probability from draw to draw due to a fixed population size.

      • Random variable (X) denoting successes during sampling conveys distinct formulas based on drawn success counts, failures available, and sample size:

      P(X = x) = rac{{{S ext{ choose } x} imes {(N-S) ext{ choose } (n-x)}}}{{{N ext{ choose } n}}}

      • Example: Probability of selecting damaged mangos from a batch of a known count with specified sampling limits, yielding probabilities, expected values, variances, and standard deviations as required.

      Conclusion

      • Each discrete probability distribution analyzed varies from binomial, Poisson, to hypergeometric, embodying unique properties and applications, often through structured calculations or software tools such as Excel for effective computations.