QTM-Chapter 4

3.4 Random Variables

  • Random variables can be classified into two types:

    • Discrete Random Variables

    • Often take only integer values.

    • Example: Number of credit hours, Difference in the number of credit hours this term vs last.

    • Continuous Random Variables

    • Take real (decimal) values.

    • Example: Cost of books this term, Difference in cost of books this term vs last.

  • Definition of Random Variable:
    A random variable is a numeric quantity whose value depends on the outcome of a random event.

  • Notation:

    • Denoted with a capital letter (e.g., X).

    • Values of a random variable are expressed with a lowercase letter (e.g., x).

    • Example: P(X=x)P(X = x) where P represents the probability of X taking the value of x.

Expected Value

  • We are often interested in the average outcome of a random variable.

  • This average is referred to as the expected value (mean), and it is calculated as a weighted average of the possible outcomes.

  • Example: In a card game, winnings are as follows:

    • Win $1 if you draw a heart.

    • Win $5 if you draw an ace (including the ace of hearts).

    • Win $10 if you draw the king of spades.

    • Win $0 for any other card.

  • Calculate expected winnings based on the probability model for the game.

Variability of Random Variables

  • In addition to the expected value, we also analyze the variability of a random variable's values.

Variability in Context
  • How much would you expect the winnings in the card game to vary from game to game?

  • An example question regarding different test formats:

    • Statements about true/false (T/F) vs multiple choice test:

      • Options:

        • A. T/F test has a higher mean and more variability in responses than the multiple choice test.

        • B. T/F test has a higher mean and less variability in responses than the multiple choice test.

        • C. T/F test has a lower mean and more variability in responses than the multiple choice test.

        • D. T/F test has a lower mean and less variability in responses than the multiple choice test.

    • Consider:

      • 10 true/false questions, P(correct)=0.5P(correct) = 0.5.

      • 10 multiple choice questions, P(correct)=0.25P(correct) = 0.25.

3.5 Continuous Distributions

From Histograms to Continuous Distributions
  • Height, as a continuous numerical variable, utilizes a probability density function depicted as a smooth curve.

  • The probability that a randomly sampled U.S. adult is between 180 cm and 185 cm can be estimated as the shaded area under the curve.

  • Note: The probability of an exact value, such as exactly 180 cm, is defined as 0 because continuous probabilities are computed as areas under the curve.

Hours of Sleep on School Nights
  • Summary Statistics:

    • Mean = 6.88 hours

    • Standard Deviation (SD) = 0.92 hours

    • Data Distribution:

      • 72% of data within 1 SD of the mean: 6.88 ± 0.926.88 \text{ } \text{± 0.92}.

      • 92% of data within 1 SD of the mean: 6.88 \text{ } \text{± 2 \times 0.92}.

      • 99% of data within 1 SD of the mean: 6.88 \text{ } \text{± 3 \times 0.92}.

4.1 Normal Distribution

Describing Normal Distributions Using Parameters
  • Characteristics:

    • Bell-shaped curve, unimodal, and symmetric.

    • Many variables approximate a normal distribution, but none are perfectly normal.

    • Denoted as N(μ,σ)N(\mu, \sigma), where μ\mu is the mean and σ\sigma is the standard deviation.

Normal Distributions with Different Parameters
  • Example: SAT scores are normally distributed with mean 1500 and standard deviation 300.

  • Example: ACT scores are normally distributed with mean 21 and standard deviation 5.

Standardizing with Z Scores
  • Definition of Z Score:
    Z score is the number of standard deviations an observation is above or below the mean.

  • Z scores can be calculated for distributions of any shape; however, only for normal distributions can Z scores be used to calculate percentiles.

  • Observations more than 2 standard deviations away from the mean (∣Z∣>2\vert Z \vert > 2) are considered unusual.

  • To assess which of two applicants performed better on their standardized tests with respect to their peers, compare their scores in terms of standard deviations.

    • Pam's SAT Score Calculation:

      • Z=(1800−1500)300=1 standard deviation above the mean.Z = \frac{(1800 - 1500)}{300} = 1 \text{ standard deviation above the mean.}

    • Jim's ACT Score Calculation:

      • Z=(24−21)5=0.6 standard deviation above the mean.Z = \frac{(24 - 21)}{5} = 0.6 \text{ standard deviation above the mean.}

Calculating Percentiles and Probabilities
  • Definition of Percentile:
    The percentage of observations below a specific data point.

  • Graphically represented as the area under the probability distribution curve to the left of that data point.

Using Tables
  • 68-95-99.7 Rule (Empirical Rule):

    • For nearly normally distributed data:

      • About 68% falls within 1 SD of the mean.

      • About 95% falls within 2 SD of the mean.

      • About 99.7% falls within 3 SD of the mean.

    • While occurrences of observations falling 4, 5, or more standard deviations from the mean are rare for nearly normal data.

  • Examples based on SAT scores distribution:

    • Approximately 68% of students score between 1200 and 1800 on the SAT.

    • Approximately 95% of students score between 900 and 2100 on the SAT.

    • Approximately 99.7% of students score between 600 and 2400 on the SAT.

Quality Control Example
  • In the Heinz ketchup factory, the amounts in bottles are normally distributed with:

    • Mean = 36 oz.

    • Standard Deviation = 0.11 oz.

  • Quality Control Standards:

    • A bottle fails inspection if it contains less than 35.8 oz. or more than 36.2 oz.

  • Probability Calculation Example:

    • Let XX represent the amount of ketchup in a bottle: X   N(μ=36,σ=0.11)X \text{ } \text{~ N}(\mu = 36, \sigma = 0.11).

Finding the Exact Probability - Using the Z Table
  • Practice Question: What percent of bottles pass the quality control inspection?

    • Options:

      • (a) 1.82%

      • (b) 6.88%

      • (c) 96.56%

      • (d) 3.44%

      • (e) 93.12%

  • With X   N(μ=36,σ=0.11)X \text{ } \text{~ N}(\mu = 36, \sigma = 0.11), find the probability for bottles with less than 35.8 oz. and more than 36.2 oz. to determine passing rates in quality control.

Finding Cutoff Points
  • Human Body Temperature Example:

    • Nearly normally distributed with mean 98.2°F and standard deviation 0.73°F.

  • Determine cutoff for the lowest 3% of body temperatures using Z scores.

Practice Question (Z-Scores)
  • What is the cutoff for the highest 10% of human body temperatures?

    • Options:

      • (a) 97.3°F

      • (b) 99.4°F

      • (c) 99.1°F

      • (d) 99.6°F

  • Identify the false statement regarding Z scores in various distributions.

4.3 Binomial Distribution

Conditions for a Binomial Distribution
  • Definition: Used for dichotomous categorical variables (two possible outcomes).

  • Examples:

    • Winning the lottery (Yes/No).

    • Mortality of mosquito in the presence of pesticide (Yes/No).

    • Correctly answering a multiple-choice question (Yes/No).

    • The outcomes are defined as “success” (winning, surviving, answering correctly) and “failure.”

  • The binomial random variable XX represents the number of successes in nn trials (e.g., X=successes in the trial setX = \text{successes in the trial set}).

  • This is characterized by parameters nn and pp: thus, X   Bin(n,p)X \text{ } \text{~ Bin}(n,p).

  • Characteristics of the Binomial Distribution - Conditions:

    1. Trials are independent.

    2. The number of trials nn is fixed.

    3. Each trial has two possible outcomes: success or failure.

    4. Each trial has the same probability of success pp.

Calculating Binomial Probabilities
  • Consider an experiment selecting four individuals. What is the probability that exactly one will refuse to administer shock?

  • For computations:

    • Compile the relevant number of scenarios based on event outcomes.

    • Determine P(single scenario)P( \text{single scenario}) using:
      P=pk(1−p)n−kP = p^k(1-p)^{n-k}
      where pp is the probability of success, with kk and nn representing the respective counts of successes and trials.

  • The Choose Function:

    • Useful for calculating the number of combinations to choose successes among trials.

  • Computing Binomial Probabilities:

    • Definitions:

      • If pp is the probability of success,

      • (1−p)(1-p) is the probability of failure,

      • nn is the number of trials, and

      • kk is the number of successes,

    • The binomial distribution explains the exact probability of achieving kk successes in nn independent trials with a consistent probability pp.

Mean and Standard Deviation of a Binomial Distribution
  • The mean (expected value) of a binomial distribution is given by:

    • μ=np\mu = np

  • The standard deviation of a binomial distribution is given by:

    • σ=np(1−p)\sigma = \sqrt{np(1-p)}

Practice Questions on Binomial Distribution
  • Identify which condition is NOT applicable for the binomial distribution:

    • (a) Trials must be independent.

    • (b) Number of trials, nn, must be fixed.

    • (c) Each trial outcome must be classified as success or failure.

    • (d) Desired successes, kk, must exceed the number of trials.

    • (e) Probability of success, pp, must remain consistent across trials.

  • Example: Calculate the probability of correctly guessing 2 out of 3 questions when guessing on a multiple-choice quiz with 4 options per question, considering trial independence.