Comprehensive Study Notes on Probability, Discrete Distributions, and Normal Distributions

Foundations of Probability and Random Processes

  • Definition of a Random Process:

    • A random process represents scenarios where the outcome of any particular trial of an experiment is unknown, but the proportion (or relative frequency) with which a particular outcome is observed approaches a specific, constant value as the number of trials increases.
    • Probability is the measure of the likelihood of a random phenomenon or chance behavior occurring.
    • Probability deals with experiments that yield unpredictable, random short-term results, yet reveal structured long-term predictability.
    • The long-term proportion in which a certain outcome is observed defines the probability of that outcome.
  • The Law of Large Numbers (LLN):

    • As the number of repetitions of a probability experiment increases, the proportion with which a certain outcome is observed gets closer to the theoretical probability of that outcome.
    • The Nonexistent "Law of Averages": The Law of Large Numbers describes long-term behavior over many trials; it does not apply to short-term sequences. A common misconception (the gambler's fallacy) posits that an outcome is "due" because it has not occurred recently (e.g., expecting heads after rolling or flipping five tails in a row, or asserting a baseball batter is due for a hit after several hitless at-bats).
    • Trials in random processes are memoryless: the mechanism behind the process does not recall past results, and the likelihood of observing any outcome on the next trial remains identical to its likelihood on the first trial.
  • Probability Experiments, Sample Spaces, and Events:

    • Experiment: Any process that can be repeated and produces uncertain outcomes.
    • Sample Space (SS): The collection of all possible outcomes of a probability experiment.
    • Event: Any collection of outcomes from a probability experiment. An event may consist of a single outcome or multiple outcomes. Events are typically denoted by capital letters (such as EE).
    • Simple Event (e1e_1): An event containing exactly one outcome.
    • Example (Rock, Paper, Scissors played twice):
      • Outcomes represent ordered pairs of selections across the two rounds.
      • Sample space: S={(R,R),(R,P),(R,S),(P,R),(P,P),(P,S),(S,R),(S,P),(S,S)}S = \{(R, R), (R, P), (R, S), (P, R), (P, P), (P, S), (S, R), (S, P), (S, S)\}.
      • Event E="choosing one Rock"E = \text{"choosing one Rock"}: E={(R,P),(R,S),(P,R),(S,R)}E = \{(R, P), (R, S), (P, R), (S, R)\}.
  • The Rules of Probabilities:

    • Rule 1 (Probability Boundaries): For any event EE, the probability P(E)P(E) must be greater than or equal to 00 and less than or equal to 11:         0≤P(E)≤10 \le P(E) \le 1
    • Rule 2 (Sum of Probabilities): The sum of the probabilities of all simple outcomes in the sample space S={e1,e2,…,en}S = \{e_1, e_2, \dots, e_n\} must equal 11:         P(e1)+P(e2)+⋯+P(en)=1P(e_1) + P(e_2) + \dots + P(e_n) = 1
    • Probability Model: A comprehensive listing of all possible outcomes of a probability experiment along with their associated probabilities. To be valid, a probability model must satisfy both Rule 1 and Rule 2.
    • Certainty and Impossibility:
      • If an event cannot happen (impossible), P(E)=0P(E) = 0.
      • If an event is guaranteed to happen (certainty), P(E)=1P(E) = 1.
      • An event with probability 0.650.65 is expected to occur approximately 6565 times out of every 100100 repetitions of the experiment.
    • Unusual Event: An event that has a low probability of occurring. The conventional threshold for classifying an event as unusual is a probability strictly less than or equal to 0.050.05 (P(E)≤0.05P(E) \le 0.05).
  • Methods of Computing Probabilities:

    • The Empirical Approach (Relative Frequency): Approximates the probability of an event based on observed experimental data:         P(E)≈Frequency of ENumber of trials of experimentP(E) \approx \frac{\text{Frequency of } E}{\text{Number of trials of experiment}}
    • The Classical Approach (Theoretical): Applicable when an experiment has nn equally likely outcomes and event EE can occur in mm different ways:         P(E)=mn=N(E)N(S)P(E) = \frac{m}{n} = \frac{N(E)}{N(S)}         where N(E)N(E) is the number of outcomes in event EE and N(S)N(S) is the total number of outcomes in the sample space SS.
    • Subjective Probability: A probability value assigned on the basis of personal judgment, expertise, or individual belief rather than relative frequencies or counting equally likely outcomes (e.g., an economic forecast predicting a 20%20\% chance of recession, or racetrack betting odds set by betting volume).
  • Empirical Probability Case Study (Pass the Pigs):

    • In a game of Pass the Pigs (Milton-Bradley), a class of 5252 students tossed a pig die a total of 39393939 times, recording six landing positions:
      • Side with no dot: frequency 13441344 (P≈13443939≈0.3412P \approx \frac{1344}{3939} \approx 0.3412)
      • Side with dot: frequency 12941294 (P≈12943939≈0.3285P \approx \frac{1294}{3939} \approx 0.3285)
      • Razorback: frequency 767767 (P≈7673939≈0.1947P \approx \frac{767}{3939} \approx 0.1947)
      • Trotter: frequency 365365 (P≈3653939≈0.0927P \approx \frac{365}{3939} \approx 0.0927)
      • Snouter: frequency 137137 (P≈1373939≈0.0348P \approx \frac{137}{3939} \approx 0.0348)
      • Leaning Jowler: frequency 3232 (P≈323939≈0.0081P \approx \frac{32}{3939} \approx 0.0081)
    • Because P(Leaning Jowler)≈0.0081<0.05P(\text{Leaning Jowler}) \approx 0.0081 < 0.05, obtaining a Leaning Jowler is classified as an unusual event.
  • Classical Probability Case Study (Peanut and Fun Size M&M Bags):

    • Peanut M&M Distribution Model: Brown (0.120.12), Yellow (0.150.15), Red (0.120.12), Blue (0.230.23), Orange (0.230.23), Green (0.150.15). All probabilities satisfy 0≤P(x)≤10 \le P(x) \le 1 and sum exactly to 1.001.00, confirming validity.
    • Fun Size Bag Analysis: A bag contains 99 brown, 66 yellow, 77 red, 44 orange, 22 blue, and 22 green candies (N(S)=30N(S) = 30 total candies):         P(Yellow)=630=0.20P(\text{Yellow}) = \frac{6}{30} = 0.20P(Blue)=230≈0.067P(\text{Blue}) = \frac{2}{30} \approx 0.067         Drawing a yellow candy is three times more likely than drawing a blue candy.

Probability Rules for Combined Events and Complements

  • Disjoint (Mutually Exclusive) Events:

    • Two events are disjoint (or mutually exclusive) if they have no outcomes in common. That is, both events cannot occur simultaneously on a single trial.
    • Addition Rule for Disjoint Events: If EE and FF are disjoint events, then:         P(E or F)=P(E)+P(F)P(E \text{ or } F) = P(E) + P(F)
    • Generalization to Multiple Disjoint Events: If events E,F,G,…E, F, G, \dots are pairwise disjoint, then:         P(E or F or G or … )=P(E)+P(F)+P(G)+…P(E \text{ or } F \text{ or } G \text{ or } \dots) = P(E) + P(F) + P(G) + \dots
    • Example (Housing Units in the United States):
      • One room: 0.0100.010
      • Two rooms: 0.0320.032
      • Three rooms: 0.0930.093
      • Four rooms: 0.1760.176
      • Five rooms: 0.2190.219
      • Six rooms: 0.1890.189
      • Seven rooms: 0.1220.122
      • Eight rooms: 0.0790.079
      • Nine or more rooms: 0.0800.080
      • P(Two or Three rooms)=0.032+0.093=0.125P(\text{Two or Three rooms}) = 0.032 + 0.093 = 0.125
      • P(One, Two, or Three rooms)=0.010+0.032+0.093=0.135P(\text{One, Two, or Three rooms}) = 0.010 + 0.032 + 0.093 = 0.135
  • The General Addition Rule:

    • When two events EE and FF are not disjoint, adding their individual probabilities double-counts the outcomes in their intersection (E and FE \text{ and } F).
    • Formula: For any two events EE and FF:         P(E or F)=P(E)+P(F)−P(E and F)P(E \text{ or } F) = P(E) + P(F) - P(E \text{ and } F)
    • Example: If P(E)=0.32P(E) = 0.32, P(F)=0.61P(F) = 0.61, and P(E and F)=0.10P(E \text{ and } F) = 0.10:         P(E or F)=0.32+0.61−0.10=0.83P(E \text{ or } F) = 0.32 + 0.61 - 0.10 = 0.83
    • If EE and FF were mutually exclusive instead, P(E and F)=0P(E \text{ and } F) = 0, yielding P(E or F)=0.32+0.61=0.93P(E \text{ or } F) = 0.32 + 0.61 = 0.93.
  • Contingency Tables and Probability Calculations:

    • Contingency tables record joint frequencies for two categorical variables across a defined population or sample.
    • Example (2021 American Community Survey Living Arrangements, in millions of households; Total = 127.5 million127.5\text{ million}):
      • 1-person household: Owner-Occupied = 18.918.9, Renter-Occupied = 17.117.1 (Subtotal = 36.036.0)
      • 2-person household: Owner-Occupied = 31.231.2, Renter-Occupied = 12.412.4 (Subtotal = 43.643.6)
      • 3-person household: Owner-Occupied = 13.313.3, Renter-Occupied = 6.36.3 (Subtotal = 19.619.6)
      • 4-or-more-person household: Owner-Occupied = 20.020.0, Renter-Occupied = 8.38.3 (Subtotal = 28.328.3)
      • Totals: Total Owner-Occupied = 83.483.4, Total Renter-Occupied = 44.144.1.
    • Calculations:
      • P(Renter)=44.1127.5≈0.3459P(\text{Renter}) = \frac{44.1}{127.5} \approx 0.3459
      • P(1-person)=36.0127.5≈0.2824P(\text{1-person}) = \frac{36.0}{127.5} \approx 0.2824
      • P(1-person or 2-person)=36.0+43.6127.5=79.6127.5≈0.6243P(\text{1-person or 2-person}) = \frac{36.0 + 43.6}{127.5} = \frac{79.6}{127.5} \approx 0.6243
      • P(1-person or Renter)=P(1-person)+P(Renter)−P(1-person and Renter)=36.0127.5+44.1127.5−17.1127.5=63.0127.5≈0.4941P(\text{1-person or Renter}) = P(\text{1-person}) + P(\text{Renter}) - P(\text{1-person and Renter}) = \frac{36.0}{127.5} + \frac{44.1}{127.5} - \frac{17.1}{127.5} = \frac{63.0}{127.5} \approx 0.4941
  • The Complement Rule:

    • Complement of an Event (EcE^c): The set of all outcomes in the sample space SS that are not contained in event EE.
    • Formula: For any event EE:         P(Ec)=1−P(E)P(E^c) = 1 - P(E)
    • Examples:
      • If the American Veterinary Medical Association reports that 31.6%31.6\% (P(E)=0.316P(E) = 0.316) of American households own a dog, the probability that a randomly chosen household does not own a dog is:             P(Ec)=1−0.316=0.684P(E^c) = 1 - 0.316 = 0.684
      • Commute time in Hartford County, CT: If P(Commute ≥90 min)=pP(\text{Commute } \ge 90\text{ min}) = p, then the probability of commuting less than 9090 minutes is P(Commute <90 min)=1−P(Commute ≥90 min)P(\text{Commute } < 90\text{ min}) = 1 - P(\text{Commute } \ge 90\text{ min}).

Independence, Conditional Probability, and Multiplication Rules

  • Independent versus Dependent Events:

    • Independent Events: Two events EE and FF are independent if the occurrence of event EE does not alter the probability that event FF occurs.
    • Dependent Events: Two events are dependent if the occurrence of event EE affects the probability of event FF.
    • Disjoint versus Independent: Disjointness and independence are fundamentally distinct concepts. If two events with non-zero probability are disjoint, they cannot be independent, because knowing that event EE occurred guarantees that event FF cannot occur (P(F∣E)=0≠P(F)P(F|E) = 0 \ne P(F)).
  • The Multiplication Rule for Independent Events:

    • If EE and FF are independent events, then:         P(E and F)=P(E)⋅P(F)P(E \text{ and } F) = P(E) \cdot P(F)
    • Extension to nn Independent Events: If E1,E2,E3,…,EnE_1, E_2, E_3, \dots, E_n are independent events:         P(E1 and E2 and … and En)=P(E1)⋅P(E2)…P(En)P(E_1 \text{ and } E_2 \text{ and } \dots \text{ and } E_n) = P(E_1) \cdot P(E_2) \dots P(E_n)
    • Applications:
      • If 35%35\% (0.350.35) of residents hold a bachelor's degree or higher: for two randomly selected individuals, P(both)=(0.35)(0.35)=0.1225P(\text{both}) = (0.35)(0.35) = 0.1225. For four individuals, P(all four)=(0.35)4=0.0150P(\text{all four}) = (0.35)^4 = 0.0150. For three holding a degree and the fourth not: P=(0.35)3⋅(1−0.35)=(0.35)3(0.65)=0.02787P = (0.35)^3 \cdot (1 - 0.35) = (0.35)^3(0.65) = 0.02787.
      • Exercise Equipment Warranty: A manufacturer knows 10%10\% of equipment is defective (P(D)=0.10P(D) = 0.10) and 30%30\% of customers use it in the first year (P(U)=0.30P(U) = 0.30). Assuming independence, the proportion making a valid claim within the 1-year warranty is:             P(D and U)=(0.10)(0.30)=0.03P(D \text{ and } U) = (0.10)(0.30) = 0.03
  • Computing "At-Least" Probabilities:

    • The phrase "at least one" represents the complement of "none".
    • Formula:P(at least one success in n trials)=1−P(zero successes)=1−(1−p)nP(\text{at least one success in } n \text{ trials}) = 1 - P(\text{zero successes}) = 1 - (1 - p)^n
    • Example: Among 500500 randomly selected residents where p=0.35p = 0.35 hold a degree:         P(at least one degree)=1−(1−0.35)500=1−(0.65)500≈1.0000P(\text{at least one degree}) = 1 - (1 - 0.35)^{500} = 1 - (0.65)^{500} \approx 1.0000
  • Conditional Probability:

    • The notation P(E∣F)P(E|F) represents the probability of event EE occurring given that event FF has already occurred.
    • Formula:P(E∣F)=P(E and F)P(F)=N(E and F)N(F)P(E|F) = \frac{P(E \text{ and } F)}{P(F)} = \frac{N(E \text{ and } F)}{N(F)}
    • Pet Ownership Survey (Algebra Students):
      • Male students: Dog 4242, Cat 1010, Bird 88 (Total = 6060)
      • Female students: Dog 1919, Cat 3232, Bird 66 (Total = 5757)
      • Overall totals: Dog 6161, Cat 4242, Bird 1414; Grand total = 117117
      • P(Bird)=14117≈0.1197P(\text{Bird}) = \frac{14}{117} \approx 0.1197
      • P(Bird∣Female)=N(Bird and Female)N(Female)=657≈0.1053P(\text{Bird} | \text{Female}) = \frac{N(\text{Bird and Female})}{N(\text{Female})} = \frac{6}{57} \approx 0.1053
      • P(Female∣Bird)=N(Female and Bird)N(Bird)=614≈0.4286P(\text{Female} | \text{Bird}) = \frac{N(\text{Female and Bird})}{N(\text{Bird})} = \frac{6}{14} \approx 0.4286
    • Household Connectivity Survey (Sample of 200200 Households):
      • High-speed internet (HH): 110110
      • Land-line phone service (LL): 3838
      • Mobile phone service (MM): 128128
      • High-speed and land-line (H and LH \text{ and } L): 2727
      • Land-line and mobile (L and ML \text{ and } M): 3131
      • Of those with mobile service, 8080 have high-speed (N(H and M)=80N(H \text{ and } M) = 80)
      • P(H∣L)=N(H and L)N(L)=2738≈0.7105P(H | L) = \frac{N(H \text{ and } L)}{N(L)} = \frac{27}{38} \approx 0.7105
  • The General Multiplication Rule:

    • Derived directly from rearranging the conditional probability formula:         P(E and F)=P(E)⋅P(F∣E)=P(F)⋅P(E∣F)P(E \text{ and } F) = P(E) \cdot P(F|E) = P(F) \cdot P(E|F)
    • Equivalence Condition for Independence: Two events EE and FF are independent if and only if:         P(E∣F)=P(E)orP(F∣E)=P(F)P(E|F) = P(E) \quad \text{or} \quad P(F|E) = P(F)

Counting Techniques and Combined Probability Problems

  • The Multiplication Rule of Counting:

    • If a task consists of a sequence of choices where there are pp selections for the first choice, qq selections for the second choice, rr selections for the third choice, and so on, the total number of distinct ways to complete the task is:         p⋅q⋅r⋅…p \cdot q \cdot r \cdot \dots
    • Fixed-Price Dinner Menu Example (Merichka's Restaurant):
      • Appetizer (22 options): soup or salad
      • Entrée (44 options): baked chicken, broiled beef patty, baby beef liver, or roast beef au jus
      • Dessert (22 options): ice cream or cheesecake
      • Total unique meal selections: 2⋅4⋅2=162 \cdot 4 \cdot 2 = 16
  • Factorial Notation:

    • For any non-negative integer nn:         n!=n(n−1)(n−2)…(3)(2)(1)n! = n(n - 1)(n - 2) \dots (3)(2)(1)
    • By definition: 0!=10! = 1 and 1!=11! = 1.
    • Arranging Test Questions: If an instructor selects 1010 questions from a test bank of 3030, the number of distinct ways to arrange those 1010 chosen questions on an exam is 10!=3,628,80010! = 3,628,800.
  • Permutations of Distinct Objects:

    • A permutation is an ordered arrangement of rr objects selected from nn distinct objects without replacement (r≤nr \le n), where order is significant.
    • Formula:nPr=n!(n−r)!_n P_r = \frac{n!}{(n - r)!}
    • Trifecta Wager Example: In a 1010-horse race, the number of ways horses can finish in first, second, and third place is:         10P3=10!(10−3)!=10!7!=10⋅9⋅8=720_{10} P_3 = \frac{10!}{(10 - 3)!} = \frac{10!}{7!} = 10 \cdot 9 \cdot 8 = 720
  • Combinations of Distinct Objects:

    • A combination is an arrangement of rr objects chosen from nn distinct objects without replacement (r≤nr \le n), where order does not matter.
    • Formula:nCr=nPrr!=n!r!(n−r)!_n C_r = \frac{_n P_r}{r!} = \frac{n!}{r!(n - r)!}
    • Carrot Treat Example: The number of ways to select 44 horses from a group of 1010 to receive a carrot treat is:         10C4=10!4!(10−4)!=10⋅9⋅8⋅74⋅3⋅2⋅1=210_{10} C_4 = \frac{10!}{4!(10 - 4)!} = \frac{10 \cdot 9 \cdot 8 \cdot 7}{4 \cdot 3 \cdot 2 \cdot 1} = 210
  • Permutations with Nondistinct Items:

    • The number of distinct permutations of nn objects of which n1n_1 are of one type, n2n_2 are of a second type, …, and nkn_k are of a kkth type (where n=n1+n2+⋯+nkn = n_1 + n_2 + \dots + n_k) is given by:         n!n1!⋅n2!⋅⋯⋅nk!\frac{n!}{n_1! \cdot n_2! \cdot \dots \cdot n_k!}
    • Arranging Letters in "MISSISSIPPI":
      • Total letters: n=11n = 11
      • Letter counts: M=1M = 1, I=4I = 4, S=4S = 4, P=2P = 2
      • Distinct arrangements: 11!1!⋅4!⋅4!⋅2!=39,916,8001⋅24⋅24⋅2=34,650\frac{11!}{1! \cdot 4! \cdot 4! \cdot 2!} = \frac{39,916,800}{1 \cdot 24 \cdot 24 \cdot 2} = 34,650
  • Applications of Counting to Probability:

    • Guiseppe's Pizza: Deal offering a large pizza with 88 available toppings (sausage, pepperoni, mushroom, olive, tomato, spinach, onion, extra cheese).
      • Total possible two-topping pizzas: 8C2=8⋅72=28_8 C_2 = \frac{8 \cdot 7}{2} = 28.
      • Two-topping pizzas containing pepperoni: 1⋅7C1=71 \cdot _7 C_1 = 7.
      • Probability of getting pepperoni: P=728=14=0.25P = \frac{7}{28} = \frac{1}{4} = 0.25.
    • Friday Night Free Pizza Drawing: Out of 100100 orders, 55 are randomly chosen to be free. If an individual orders 22 pizzas:
      • Total ways to choose 22 winning pizzas from 100100: 100C2=4950_{100} C_2 = 4950.
      • Ways to receive zero free pizzas: 95C2=95⋅942=4465_{95} C_2 = \frac{95 \cdot 94}{2} = 4465.
      • Probability of receiving at least one free pizza: P(at least one free)=1−44654950=4854950≈0.0980P(\text{at least one free}) = 1 - \frac{4465}{4950} = \frac{485}{4950} \approx 0.0980.
    • Marbles in a Jar Problem: A jar contains 44 red, 55 blue, and 33 yellow marbles (total N=12N = 12). Three marbles are selected at random without replacement:
      • Total selections: 12C3=12⋅11⋅106=220_{12} C_3 = \frac{12 \cdot 11 \cdot 10}{6} = 220.
      • Selections with exactly two red and one blue: (4C2)⋅(5C1)=6⋅5=30(_4 C_2) \cdot (_5 C_1) = 6 \cdot 5 = 30.
      • Probability: P(2 red and 1 blue)=30220=322≈0.1364P(\text{2 red and 1 blue}) = \frac{30}{220} = \frac{3}{22} \approx 0.1364.
    • Child Advocacy Center Case Distribution (Total = 240240 Cases):
      • Age under 1: 88 cases (P=8240≈0.0333P = \frac{8}{240} \approx 0.0333; unusual since P≤0.05P \le 0.05)
      • Age 1 to 5: 6262 cases (P=62240≈0.2583P = \frac{62}{240} \approx 0.2583)
      • Age 6 to 10: 5454 cases (P=54240=0.2250P = \frac{54}{240} = 0.2250)
      • Age 11 to 14: 7272 cases (P=72240=0.3000P = \frac{72}{240} = 0.3000)
      • Age 15 to 17: 4444 cases (P=44240≈0.1833P = \frac{44}{240} \approx 0.1833)
      • P(11 to 17 years old)=P(11-14)+P(15-17)=72+44240=116240≈0.4833P(\text{11 to 17 years old}) = P(11\text{-}14) + P(15\text{-}17) = \frac{72 + 44}{240} = \frac{116}{240} \approx 0.4833

Discrete Random Variables and Probability Distributions

  • Distinguishing Discrete and Continuous Random Variables:
    • Random Variable (XX): A numerical measure of the outcome of a probability experiment, whose value is determined by chance.
    • Discrete Random Variable: Possesses either a finite or a countable number of values. Its values can be plotted as discrete points on a number line with gaps separating them.
    • Continuous Random Variable: Possesses infinitely many values corresponding to intervals on a real line, plotted in an uninterrupted manner without gaps.

Number line comparison of discrete and continuous random variables

*   *Classification Examples (16-ounce Bag of M&Ms):*
    *   Mass of each individual M&M: Continuous (measured weight, e.g., x>0 gx > 0\,g).
    *   Number of pieces of candy in the bag: Discrete (counted values, x∈{0,1,2,… }x \in \{0, 1, 2, \dots\}).
    *   Diameter of each piece of candy: Continuous (measured dimension, e.g., x>0 mmx > 0\,mm).
    *   Number of distinct colors in the bag: Discrete (counted values, x∈{1,2,3,4,5,6}x \in \{1, 2, 3, 4, 5, 6\}).
  • Rules for a Discrete Probability Distribution:

    • A discrete probability distribution lists all possible numerical values xx that the random variable XX can assume, along with the corresponding probability P(x)P(x).
    • Requirements:
      1. ∑P(x)=1\sum P(x) = 1
      2. 0≤P(x)≤10 \le P(x) \le 1 for all individual values of xx
  • Mean, Expected Value, and Standard Deviation of a Discrete Random Variable:

    • Mean (Expected Value, E(X)E(X)): Represents the theoretical weighted average of the outcomes in the long run:         μX=∑[x⋅P(x)]\mu_X = \sum [x \cdot P(x)]
    • Interpretation of the Mean: If an experiment is repeated nn independent times and the values of XX are recorded as x1,x2,…,xnx_1, x_2, \dots, x_n, the sample mean:         xˉ=x1+x2+⋯+xnn\bar{x} = \frac{x_1 + x_2 + \dots + x_n}{n}         approaches μX\mu_X as nn increases, causing the difference ∣xˉ−μX∣|\bar{x} - \mu_X| to converge to 00.
    • Standard Deviation: Measures the dispersion of the distribution around its mean:         σX=∑[(x−μX)2⋅P(x)]=∑[x2⋅P(x)]−μX2\sigma_X = \sqrt{\sum [(x - \mu_X)^2 \cdot P(x)]} = \sqrt{\sum [x^2 \cdot P(x)] - \mu_X^2}
  • Case Study: Chuck-a-Luck Game Analysis:

    • Game Mechanics: Three dice are tumbled in a wire cage. A player wagers 1\ on a selected integer from 11 to 66. If the chosen number appears on kk dice (k∈{1,2,3}k \in \{1, 2, 3\}), the player receives k\ in winnings and gets their 1\ stake back. If no dice match (k=0k = 0), the player loses their 1\ stake.
    • Distribution Table:
      • 00 matches (x = -1\): P(x)=0.5787P(x) = 0.5787
      • 11 match (x = +1\): P(x)=0.3472P(x) = 0.3472
      • 22 matches (x = +2\): P(x)=0.0695P(x) = 0.0695
      • 33 matches (x = +3\): P(x)=0.0046P(x) = 0.0046
    • Verification: ∑P(x)=0.5787+0.3472+0.0695+0.0046=1.0000\sum P(x) = 0.5787 + 0.3472 + 0.0695 + 0.0046 = 1.0000.
    • Mean Computation:μX=(−1)(0.5787)+(1)(0.3472)+(2)(0.0695)+(3)(0.0046)=−0.5787+0.3472+0.1390+0.0138=−0.0787≈−$0.079\mu_X = (-1)(0.5787) + (1)(0.3472) + (2)(0.0695) + (3)(0.0046) = -0.5787 + 0.3472 + 0.1390 + 0.0138 = -0.0787 \approx -\$0.079Interpretation: In the long run, a player expects to lose an average of approximately 7.97.9 cents (or 88 cents) per 1\ wager. In a simulated trial of 650650 games, the running mean converges directly toward this theoretical average (-0.08\).
    • Standard Deviation Computation:
      • ∑[x2⋅P(x)]=(−1)2(0.5787)+(1)2(0.3472)+(2)2(0.0695)+(3)2(0.0046)=0.5787+0.3472+0.2780+0.0414=1.2453\sum [x^2 \cdot P(x)] = (-1)^2(0.5787) + (1)^2(0.3472) + (2)^2(0.0695) + (3)^2(0.0046) = 0.5787 + 0.3472 + 0.2780 + 0.0414 = 1.2453
      • \sigma_X = \sqrt{1.2453 - (-0.0787)^2} = \sqrt{1.2453 - 0.00619} = \sqrt{1.2391} \approx 1.1131\
  • Case Study: Expected Profit on Real Estate Renovation:

    • Shawn and Maddie acquire a foreclosed property for 50,000\ and spend an additional 27,000\ on repairs (total fixed cost = 77,000\).
    • Selling Price Scenarios and Net Profit (xx):
      • Resell at 120,000\ (x = 43,000\), P(x)=0.15P(x) = 0.15
      • Resell at 100,000\ (x = 23,000\), P(x)=0.45P(x) = 0.45
      • Resell at 80,000\ (x = 3,000\), P(x)=0.25P(x) = 0.25
      • Resell at 60,000\ (x = -17,000\), P(x)=0.15P(x) = 0.15
    • Expected Net Profit Computation:E(X)=(43,000)(0.15)+(23,000)(0.45)+(3,000)(0.25)+(−17,000)(0.15)E(X) = (43,000)(0.15) + (23,000)(0.45) + (3,000)(0.25) + (-17,000)(0.15)         E(X) = 6,450 + 10,350 + 750 - 2,550 = 15,000\         Interpretation: The expected average net profit across many similar real estate ventures is 15,000\.

The Binomial Probability Distribution

  • Criteria for a Binomial Probability Experiment:

    1. The experiment is conducted for a fixed number of trials, denoted by nn.
    2. The trials are independent; the outcome of one trial has no influence on the outcomes of any other trials.
    3. Each trial has exactly two mutually exclusive (disjoint) outcomes: success or failure.
    4. The probability of success, denoted by pp, remains constant for each trial. The probability of failure is 1−p1 - p.
  • Binomial Notation and Probability Distribution Function:

    • nn: Total number of independent trials.
    • pp: Constant probability of success on any single trial.
    • 1−p1 - p: Probability of failure on any single trial.
    • XX: Discrete random variable denoting the total number of successes in nn trials (x∈{0,1,2,…,n}x \in \{0, 1, 2, \dots, n\}).
    • Binomial Probability Formula:P(x)=nCx⋅px⋅(1−p)n−x=n!x!(n−x)!⋅px⋅(1−p)n−xP(x) = {_n C_x} \cdot p^x \cdot (1 - p)^{n - x} = \frac{n!}{x!(n - x)!} \cdot p^x \cdot (1 - p)^{n - x}
  • Mean and Standard Deviation of a Binomial Distribution:

    • Mean (Expected Number of Successes):μX=np\mu_X = np
    • Standard Deviation:σX=np(1−p)\sigma_X = \sqrt{np(1 - p)}
  • Shape and Approximate Normality of the Binomial Distribution:

    • For a fixed value of pp, as the number of trials nn increases, the binomial probability distribution becomes progressively more symmetric and bell-shaped.
    • Rule of Thumb for Bell-Shaped Distribution:np(1−p)≥10np(1 - p) \ge 10         When this condition is satisfied, the distribution is approximately normal.
  • Worked Examples (Binomial Distribution):

    • McCarran International Airport Flight Punctuality:
      • Parameters: In April 2023, on-time proportion was p=0.70p = 0.70. A sample of n=4n = 4 flights is randomly selected.
      • Sample space values for XX: 0,1,2,3,40, 1, 2, 3, 4.
      • P(0)=4C0(0.70)0(0.30)4=1⋅1⋅0.0081=0.0081P(0) = {_4 C_0}(0.70)^0(0.30)^4 = 1 \cdot 1 \cdot 0.0081 = 0.0081
      • P(1)=4C1(0.70)1(0.30)3=4⋅(0.70)⋅(0.027)=0.0756P(1) = {_4 C_1}(0.70)^1(0.30)^3 = 4 \cdot (0.70) \cdot (0.027) = 0.0756
      • P(2)=4C2(0.70)2(0.30)2=6⋅(0.49)⋅(0.09)=0.2646P(2) = {_4 C_2}(0.70)^2(0.30)^2 = 6 \cdot (0.49) \cdot (0.09) = 0.2646
      • P(3)=4C3(0.70)3(0.30)1=4⋅(0.343)⋅(0.30)=0.4116P(3) = {_4 C_3}(0.70)^3(0.30)^1 = 4 \cdot (0.343) \cdot (0.30) = 0.4116
      • P(4)=4C4(0.70)4(0.30)0=1⋅(0.2401)⋅1=0.2401P(4) = {_4 C_4}(0.70)^4(0.30)^0 = 1 \cdot (0.2401) \cdot 1 = 0.2401
    • Public Opinion on Divorce (Harris International):
      • Parameters: Proportion of adult Americans accepting divorce is p=0.58p = 0.58. Sample size n=30n = 30.
      • Exactly 20 accept: P(X=20)=30C20(0.58)20(0.42)10≈0.0988P(X = 20) = {_{30} C_{20}} (0.58)^{20} (0.42)^{10} \approx 0.0988
      • Fewer than 10 accept: P(X<10)=P(X≤9)=∑x=0930Cx(0.58)x(0.42)30−xP(X < 10) = P(X \le 9) = \sum_{x=0}^{9} {_{30} C_x} (0.58)^x (0.42)^{30 - x}
      • At least 10 accept: P(X≥10)=1−P(X≤9)P(X \ge 10) = 1 - P(X \le 9)
      • Between 18 and 22 accept, inclusive: P(18≤X≤22)=∑x=182230Cx(0.58)x(0.42)30−xP(18 \le X \le 22) = \sum_{x=18}^{22} {_{30} C_x} (0.58)^x (0.42)^{30 - x}
    • Evaluating Unusual Observations (n=500n = 500, p=0.58p = 0.58):
      • Mean: μX=np=500(0.58)=290\mu_X = np = 500(0.58) = 290
      • Standard deviation: σX=np(1−p)=500(0.58)(0.42)=121.8≈11.0363\sigma_X = \sqrt{np(1 - p)} = \sqrt{500(0.58)(0.42)} = \sqrt{121.8} \approx 11.0363
      • Normality condition: np(1−p)=121.8≥10np(1 - p) = 121.8 \ge 10 (well satisfied).
      • Usual interval (Empirical Rule, μ±2σ\mu \pm 2\sigma): 290±2(11.04)=[267.92,312.08]290 \pm 2(11.04) = [267.92, 312.08].
      • Result: An outcome of 310310 successes falls within this interval (310≤312.08310 \le 312.08); therefore, observing 310310 people who believe divorce is acceptable is not unusual.

The Poisson Probability Distribution

  • Criteria for a Poisson Process:

    • A discrete random variable XX, representing the number of occurrences (successes) of an event over a fixed interval of time or space, follows a Poisson process if:
      1. The probability of two or more successes in any sufficiently small subinterval is 00.
      2. The probability of success is identical for any two subintervals of equal length.
      3. The number of successes in any interval is independent of the number of successes in any other interval, provided the intervals do not overlap.
  • Poisson Probability Distribution Function:

    • If XX represents the number of occurrences in an interval of fixed length tt, and occurrences arrive at an average rate of λ\lambda events per unit interval of length 11, the probability of observing exactly xx occurrences is:         P(x)=(λt)xx!e−λtfor x=0,1,2,3,…P(x) = \frac{(\lambda t)^x}{x!} e^{-\lambda t} \quad \text{for } x = 0, 1, 2, 3, \dots         where e≈2.71828e \approx 2.71828.
  • Mean and Standard Deviation of a Poisson Random Variable:

    • Mean (Expected Value):μX=λt\mu_X = \lambda t
    • Standard Deviation:σX=λt=μX\sigma_X = \sqrt{\lambda t} = \sqrt{\mu_X}
  • Worked Examples (Poisson Distribution):

    • Textbook Typos Analysis:
      • Parameters: Typos occur at a rate of λ=0.25\lambda = 0.25 per page. The interval is t=50t = 50 pages. Thus, expected occurrences: λt=(0.25)(50)=12.5\lambda t = (0.25)(50) = 12.5.
      • Probability of exactly 10 typos:P(10)=(12.5)1010!e−12.5≈0.0849P(10) = \frac{(12.5)^{10}}{10!} e^{-12.5} \approx 0.0849
      • Probability of less than 10 typos:P(X<10)=P(X≤9)=∑x=09(12.5)xx!e−12.5P(X < 10) = P(X \le 9) = \sum_{x=0}^{9} \frac{(12.5)^x}{x!} e^{-12.5}
      • Probability of at least 10 typos:P(X≥10)=1−P(X≤9)P(X \ge 10) = 1 - P(X \le 9)
    • Hospital Emergency Room Arrivals:
      • Parameters: Patients arrive at a rate of λ=0.5\lambda = 0.5 arrivals per minute between 9:00 p.m. and midnight. Interval length t=30 minutest = 30\text{ minutes}.
      • Expected arrivals per interval: μX=λt=(0.5)(30)=15 patients\mu_X = \lambda t = (0.5)(30) = 15\text{ patients}.
      • Standard deviation: σX=15≈3.8730 patients\sigma_X = \sqrt{15} \approx 3.8730\text{ patients}.
      • Probability of exactly 10 arrivals:P(10)=151010!e−15≈0.0486P(10) = \frac{15^{10}}{10!} e^{-15} \approx 0.0486
      • Unusual Threshold Evaluation (more than 20 arrivals):
        • Usual interval: μX±2σX=15±2(3.873)=[7.25,22.75]\mu_X \pm 2\sigma_X = 15 \pm 2(3.873) = [7.25, 22.75].
        • Because 2020 falls within two standard deviations (≤22.75\le 22.75), observing more than 2020 arrivals is not unusual according to the two-standard-deviation criterion, although P(X>20)P(X > 20) can be calculated precisely using technology to confirm whether P(X>20)≤0.05P(X > 20) \le 0.05.

Properties and Applications of the Normal Distribution

  • Continuous Uniform Distribution:

    • A continuous random variable has a uniform distribution over an interval [a,b][a, b] if the likelihood of landing in any subinterval depends exclusively on the subinterval's length.
    • Amazon Delivery Example: Delivery is scheduled between 10:00 a.m. and 11:00 a.m. Let XX be minutes elapsed past 10:00 a.m. (0<X<600 < X < 60). The density function is constant: f(x)=160f(x) = \frac{1}{60} for 0≤x≤600 \le x \le 60.
    • Probability Calculation: The probability that the package arrives between 10:15 a.m. and 10:30 a.m. (15≤X≤3015 \le X \le 30) corresponds to the rectangular area:         P(15≤X≤30)=Base×Height=(30−15)⋅160=1560=0.25P(15 \le X \le 30) = \text{Base} \times \text{Height} = (30 - 15) \cdot \frac{1}{60} = \frac{15}{60} = 0.25
    • To find the time within which there is a 30%30\% probability of arrival: Base⋅160=0.30  ⟹  Base=18 minutes\text{Base} \cdot \frac{1}{60} = 0.30 \implies \text{Base} = 18\text{ minutes}.
  • Properties of Probability Density Functions (pdf):

    • A probability density function f(x)f(x) for a continuous random variable must satisfy:
      1. The total area under the graph over all possible values of the random variable equals 11.
      2. The curve lies on or above the horizontal axis everywhere: f(x)≥0f(x) \ge 0 for all possible values of xx.
    • The area under the graph of a density function between two values aa and bb equals the probability that the random variable assumes a value in that interval: P(a≤X≤b)P(a \le X \le b).
    • For any continuous random variable, the probability of assuming any exact single value is zero (P(X=c)=0P(X = c) = 0); therefore, P(a≤X≤b)=P(a<X<b)P(a \le X \le b) = P(a < X < b).
  • Properties of the Normal Density Curve:

    1. The normal curve is symmetric about its population mean, μ\mu.
    2. Because mean=median=mode\text{mean} = \text{median} = \text{mode}, the curve exhibits a single peak occurring precisely at x=μx = \mu.
    3. The curve has inflection points (points where concavity changes) at x=μ−σx = \mu - \sigma and x=μ+σx = \mu + \sigma.
    4. The total area bounded under the normal curve and the horizontal axis is exactly 11.
    5. Due to symmetry, the area under the curve to the left of μ\mu is 0.50.5 (12\frac{1}{2}), and the area to the right of μ\mu is 0.50.5 (12\frac{1}{2}).
    6. As xx increases or decreases without bound (x→∞x \to \infty or x→−∞x \to -\infty), the graph approaches, but never touches or crosses, the horizontal axis (the horizontal axis serves as an asymptote).
    7. The Empirical Rule: For any normally distributed random variable:
      • Approximately 68%68\% of the area lies within 11 standard deviation of the mean: (μ−σ,μ+σ)(\mu - \sigma, \mu + \sigma).
      • Approximately 95%95\% of the area lies within 22 standard deviations of the mean: (μ−2σ,μ+2σ)(\mu - 2\sigma, \mu + 2\sigma).
      • Approximately 99.7%99.7\% of the area lies within 33 standard deviations of the mean: (μ−3σ,μ+3σ)(\mu - 3\sigma, \mu + 3\sigma).

The Empirical Rule for normal distributions showing percentage of area within 1, 2, and 3 standard deviations

  • Role of Parameters in the Normal Curve:

    • Changing the mean μ\mu shifts the normal curve along the horizontal axis without modifying its spread or shape.
    • Changing the standard deviation σ\sigma alters the spread of the distribution: larger values of σ\sigma flatten and widen the curve, whereas smaller values compress and heighten the curve around μ\mu.
  • Interpreting Area under the Normal Curve:

    • If XX is normally distributed with parameters μ\mu and σ\sigma, the area under the curve over an interval [a,b][a, b] represents:
      1. The proportion of the population that possesses the characteristic defined by the interval.
      2. The probability that a randomly chosen individual from the population possesses that characteristic.
    • Major League Baseball (MLB) Home Run Example (2022 Season):
      • Distance traveled is normally distributed with μ=398.6 feet\mu = 398.6\,feet and σ=24.8 feet\sigma = 24.8\,feet.
      • If the area under the curve to the right of x=440 feetx = 440\,feet is 0.04750.0475:
        1. The proportion of all home runs hit during the 2022 season that traveled more than 440 feet440\,feet is 0.04750.0475 (4.75%4.75\%).
        2. The probability that a randomly chosen home run traveled more than 440 feet440\,feet is 0.04750.0475.
  • Standardizing Values Using ZZ-Scores:

    • The standard normal distribution has mean μ=0\mu = 0 and standard deviation σ=1\sigma = 1. Any normal variable XX can be converted to a standard normal variable ZZ via:         Z=X−μσZ = \frac{X - \mu}{\sigma}
    • Reporting Extreme Tail Probabilities: Because the normal curve extends infinitely in both directions, the cumulative area to the left or right of any finite value cannot mathematically equal 00 or 11. When computational tools round extreme cumulative areas to 11 or 00, they should be recorded as >0.9999> 0.9999 or <0.0001< 0.0001, respectively.
  • Worked Application Examples (Normal Distribution):

    • Ball Bearing Factory Specifications:
      • Parameters: Mean diameter μ=1 cm\mu = 1\,cm, standard deviation σ=0.002 cm\sigma = 0.002\,cm. Tolerance limits: 0.995 cm0.995\,cm to 1.005 cm1.005\,cm.
      • Proportion under 0.999 cm0.999\,cm:Z=0.999−10.002=−0.0010.002=−0.50Z = \frac{0.999 - 1}{0.002} = \frac{-0.001}{0.002} = -0.50             Using standard normal tables, P(Z<−0.50)=0.3085P(Z < -0.50) = 0.3085.
      • Probability exceeding 1.0055 cm1.0055\,cm:Z=1.0055−10.002=0.00550.002=2.75Z = \frac{1.0055 - 1}{0.002} = \frac{0.0055}{0.002} = 2.75P(Z>2.75)=1−P(Z<2.75)=1−0.9970=0.0030P(Z > 2.75) = 1 - P(Z < 2.75) = 1 - 0.9970 = 0.0030.
      • Proportion meeting specifications (0.9950.995 to 1.005 cm1.005\,cm):Z1=0.995−10.002=−2.50andZ2=1.005−10.002=2.50Z_1 = \frac{0.995 - 1}{0.002} = -2.50 \quad \text{and} \quad Z_2 = \frac{1.005 - 1}{0.002} = 2.50P(−2.50<Z<2.50)=P(Z<2.50)−P(Z<−2.50)=0.9938−0.0062=0.9876P(-2.50 < Z < 2.50) = P(Z < 2.50) - P(Z < -2.50) = 0.9938 - 0.0062 = 0.9876             Thus, approximately 98.76%98.76\% of ball bearings meet specifications.
      • Middle 99%99\% of Diameters:             Middle 99%99\% leaves 0.5%0.5\% (0.0050.005) in each tail. The critical ZZ-values corresponding to areas of 0.0050.005 and 0.9950.995 are Z=±2.575Z = \pm 2.575.             X=μ±Zσ=1±(2.575)(0.002)=1±0.00515X = \mu \pm Z\sigma = 1 \pm (2.575)(0.002) = 1 \pm 0.00515             Middle 99%99\% range: [0.99485 cm,1.00515 cm][0.99485\,cm, 1.00515\,cm].
    • Percentile Calculation (Home Run Distance):
      • To find the percentile rank of a 450 foot450\,foot home run (μ=398.6\mu = 398.6, σ=24.8\sigma = 24.8):             Z=450−398.624.8=51.424.8≈2.07Z = \frac{450 - 398.6}{24.8} = \frac{51.4}{24.8} \approx 2.07             Cumulative area to the left: P(Z<2.07)=0.9808P(Z < 2.07) = 0.9808. The home run is at the 9898th percentile.
      • Finding Distance at the 9090th Percentile:             For an area of 0.90000.9000 to the left, the corresponding ZZ-score is approximately Z=1.28Z = 1.28.             X=μ+Zσ=398.6+(1.28)(24.8)=398.6+31.744=430.344 feetX = \mu + Z\sigma = 398.6 + (1.28)(24.8) = 398.6 + 31.744 = 430.344\,feet
    • Critical Value Notation (zαz_\alpha):
      • The notation zαz_\alpha denotes the specific ZZ-score whose area under the standard normal curve to the right equals α\alpha.
      • For example, z0.25z_{0.25} represents the ZZ-score with an area of 0.250.25 to its right, which corresponds to the 7575th percentile (0.750.75 area to the left). Standard normal tables provide z0.25≈0.67z_{0.25} \approx 0.67.

Assessing Normality and Normal Approximations

  • Assessing Normality via Normal Probability Plots:

    • Histograms drawn from small samples (n<30n < 30) do not reliably reflect the true shape of the underlying population distribution.
    • A normal probability plot graphs observed sample values on the horizontal axis against their expected normal scores (ZZ-scores) on the vertical axis.
    • Step-by-Step Procedure for Constructing a Normal Probability Plot:
      1. Step 1: Sort the observed data values in ascending order (x1,x2,…,xnx_1, x_2, \dots, x_n).
      2. Step 2: For each sorted value at index ii (where i=1,2,…,ni = 1, 2, \dots, n), compute the expected cumulative proportion fif_i:             fi=i−0.375n+0.25f_i = \frac{i - 0.375}{n + 0.25}             where fif_i represents the expected area under the standard normal curve to the left of the iith observation.
      3. Step 3: Determine the expected normal score (ZZ-score) corresponding to each cumulative area fif_i using standard normal probability tables.
      4. Step 4: Plot the observed data values (xx) on the horizontal axis and the corresponding expected ZZ-scores on the vertical axis.
    • Evaluation of Linearity: If the plotted points form an approximately straight linear pattern, it is reasonable to conclude that the sample was drawn from a normally distributed population.
    • Formal Decision Rule: Calculate the linear correlation coefficient rr between the observed data and expected ZZ-scores. If rr exceeds the critical value from Table VI for sample size nn, the assumption of population normality is supported.
    • Worked Example (Statistics Exam Scores, n=7n = 7):
      • Given scores: 52,67,71,75,83,87,9252, 67, 71, 75, 83, 87, 92.
      • With n=7n = 7, denominator n+0.25=7.25n + 0.25 = 7.25:
        • i=1i = 1 (x=52x = 52): f1=1−0.3757.25=0.6257.25≈0.0862  ⟹  Z≈−1.36f_1 = \frac{1 - 0.375}{7.25} = \frac{0.625}{7.25} \approx 0.0862 \implies Z \approx -1.36
        • i=2i = 2 (x=67x = 67): f2=2−0.3757.25=1.6257.25≈0.2241  ⟹  Z≈−0.76f_2 = \frac{2 - 0.375}{7.25} = \frac{1.625}{7.25} \approx 0.2241 \implies Z \approx -0.76
        • i=3i = 3 (x=71x = 71): f3=3−0.3757.25=2.6257.25≈0.3621  ⟹  Z≈−0.35f_3 = \frac{3 - 0.375}{7.25} = \frac{2.625}{7.25} \approx 0.3621 \implies Z \approx -0.35
        • i=4i = 4 (x=75x = 75): f4=4−0.3757.25=3.6257.25≈0.5000  ⟹  Z=0.00f_4 = \frac{4 - 0.375}{7.25} = \frac{3.625}{7.25} \approx 0.5000 \implies Z = 0.00
        • i=5i = 5 (x=83x = 83): f5=5−0.3757.25=4.6257.25≈0.6379  ⟹  Z≈0.35f_5 = \frac{5 - 0.375}{7.25} = \frac{4.625}{7.25} \approx 0.6379 \implies Z \approx 0.35
        • i=6i = 6 (x=87x = 87): f6=6−0.3757.25=5.6257.25≈0.7759  ⟹  Z≈0.76f_6 = \frac{6 - 0.375}{7.25} = \frac{5.625}{7.25} \approx 0.7759 \implies Z \approx 0.76
        • i=7i = 7 (x=92x = 92): f7=7−0.3757.25=6.6257.25≈0.9138  ⟹  Z≈1.36f_7 = \frac{7 - 0.375}{7.25} = \frac{6.625}{7.25} \approx 0.9138 \implies Z \approx 1.36
    • Occupational Health Study (FEV/VC Ratio): An investigation evaluating seventeen (n=17n = 17) factory workers exposed to Bacillus subtilis enzyme measured the ratio of forced expiratory volume in one second (FEV) to vital capacity (VC) to determine if lung function ratios conform to a normal distribution via probability plotting.
  • The Normal Approximation to the Binomial Probability Distribution:

    • When the number of trials nn is sufficiently large, the discrete binomial distribution can be approximated by a continuous normal distribution.
    • Rule of Thumb Criterion:np(1−p)≥10np(1 - p) \ge 10
    • Approximation Parameters:
      • Mean: μX=np\mu_X = np
      • Standard deviation: σX=np(1−p)\sigma_X = \sqrt{np(1 - p)}
    • Flight Cancellations Evaluation (LaGuardia Airport):
      • Data: Cancellation rate is 3.23%3.23\% (p=0.0323p = 0.0323) with a sample of n=100n = 100 flights.
      • Checking the condition:             np(1−p)=100(0.0323)(1−0.0323)=3.23(0.9677)=3.1257np(1 - p) = 100(0.0323)(1 - 0.0323) = 3.23(0.9677) = 3.1257
      • Because 3.1257<103.1257 < 10, the condition is violated, and the normal approximation cannot be used.
    • Valentine's Day Gifts for Pet Dogs:
      • Data: 8.5%8.5\% (p=0.085p = 0.085) of American households buy gifts for pet dogs; sample size n=200n = 200.
      • Checking the condition:             np(1−p)=200(0.085)(1−0.085)=17(0.915)=15.555≥10np(1 - p) = 200(0.085)(1 - 0.085) = 17(0.915) = 15.555 \ge 10             The condition is satisfied; the normal approximation is valid.
      • Expected Number (Mean):μX=np=200(0.085)=17 households\mu_X = np = 200(0.085) = 17\text{ households}
      • Standard Deviation:σX=15.555≈3.9440 households\sigma_X = \sqrt{15.555} \approx 3.9440\text{ households}
      • Probability of 20 or More Gifts (X≥20X \ge 20):             Using the standardizing transformation (with continuity correction or directly):             Z=20−173.944=33.944≈0.76Z = \frac{20 - 17}{3.944} = \frac{3}{3.944} \approx 0.76P(Z≥0.76)=1−P(Z<0.76)=1−0.7764=0.2236P(Z \ge 0.76) = 1 - P(Z < 0.76) = 1 - 0.7764 = 0.2236
      • Probability of More Than 25 Gifts (X>25X > 25):Z=25−173.944=83.944≈2.03Z = \frac{25 - 17}{3.944} = \frac{8}{3.944} \approx 2.03P(Z>2.03)=1−P(Z<2.03)=1−0.9788=0.0212P(Z > 2.03) = 1 - P(Z < 2.03) = 1 - 0.9788 = 0.0212