Comprehensive Lecture and Study Notes on Descriptive Statistics, Experimental Design, Probability Theory, and Random Variables

Descriptive Statistics and Data Summaries

  • Foundational Statistical Terminology and Visual Summaries:

    • Mean: The arithmetic average of a dataset, calculated as the sum of all observations divided by the total number of observations. While sensitive to extreme values, it serves as the standard central benchmark in symmetric distributions.
    • Median: The physical midpoint of an ordered dataset, dividing the lower 50%50\% of observations from the upper 50%50\%. It represents a resistant measure of central tendency unaffected by extreme outliers.
    • Mode: The most frequently occurring value or values in a dataset. On a frequency histogram, it corresponds to the highest peak.
    • Categorical Data Displays: Categorical distributions are summarized visually using bar charts and pie charts.
    • Sample and Statistic: A sample is an observed subset of a population used in empirical studies; a statistic is a numerical summary metric calculated directly from sample data.
    • Range: The total spread between the extreme values in a dataset, defined as Range=Maximum−Minimum\text{Range} = \text{Maximum} - \text{Minimum}.
  • Stem-and-Leaf Displays:

    • A stem-and-leaf plot displays quantitative data by splitting each observation into a "stem" (leading digits) and a "leaf" (final trailing digit), maintaining the raw data values while displaying distributional shape.
    • Worked Stem-and-Leaf Analysis:
    • Given the stem-and-leaf table:

Stem and leaf display

* Stem 44 with leaves 5,6,85, 6, 8 represents the data values: 45,46,4845, 46, 48.
* Stem 55 with leaves 2,2,32, 2, 3 represents: 52,52,5352, 52, 53.
* Stem 66 with leaf 66 represents: 6666.
* Stem 77 with leaves 1,7,8,91, 7, 8, 9 represents: 71,77,78,7971, 77, 78, 79.
* **Sample Size:** n=11n = 11.
* **Mean Calculation:**

      ∑xi=45+46+48+52+52+53+66+71+77+78+79=667\sum x_i = 45 + 46 + 48 + 52 + 52 + 53 + 66 + 71 + 77 + 78 + 79 = 667xˉ=66711≈60.64\bar{x} = \frac{667}{11} \approx 60.64 * Median Calculation: The 6th ordered value is 5353. * Mode Calculation: The value appearing with the highest frequency is 5252 (frequency of 22).

  • Measures of Center for Grouped Frequency Distributions:

    • When data are summarized into frequency tables, the sample mean is calculated by weighting each distinct value (or midpoint) by its respective frequency:     xˉ=∑(fi⋅xi)∑fi\bar{x} = \frac{\sum (f_i \cdot x_i)}{\sum f_i}
    • Grouped Data Definitions:
    • Class Interval: Defined numerical ranges into which raw data values are grouped.
    • Class Boundaries: The true boundaries that eliminate gaps between contiguous class intervals (e.g., for interval 1−31 - 3, boundaries are 0.50.5 to 3.53.5).
    • Class Mark (xix_i): The midpoint of a class interval, calculated as the average of the upper and lower interval limits: Class Mark=Lower Limit+Upper Limit2\text{Class Mark} = \frac{\text{Lower Limit} + \text{Upper Limit}}{2}.
    • Grouped Mean Worked Computation:
    • Class interval 1−31 - 3: Class mark x1=2x_1 = 2, frequency f1=4f_1 = 4, product f1x1=8f_1 x_1 = 8.
    • Class interval 4−64 - 6: Class mark x2=5x_2 = 5, frequency f2=9f_2 = 9, product f2x2=45f_2 x_2 = 45.
    • Class interval 7−97 - 9: Class mark x3=8x_3 = 8, frequency f3=11f_3 = 11, product f3x3=88f_3 x_3 = 88.
    • Class interval 10−1210 - 12: Class mark x4=11x_4 = 11, frequency f4=8f_4 = 8, product f4x4=88f_4 x_4 = 88.
    • Class interval 13−1513 - 15: Class mark x5=14x_5 = 14, frequency f5=2f_5 = 2, product f5x5=28f_5 x_5 = 28.
    • Sum of frequencies: ∑fi=4+9+11+8+2=34\sum f_i = 4 + 9 + 11 + 8 + 2 = 34.
    • Sum of products: ∑(fi⋅xi)=8+45+88+88+28=257\sum (f_i \cdot x_i) = 8 + 45 + 88 + 88 + 28 = 257.
    • Sample mean:       xˉ=25734≈7.56\bar{x} = \frac{257}{34} \approx 7.56
  • Measures of Spread: Population and Sample Variability:

    • Variance: The average squared deviation of observations from their mean.
    • Population Variance (σ2\sigma^2): Computed across an entire population of size NN:       σ2=∑(xi−μ)2N\sigma^2 = \frac{\sum (x_i - \mu)^2}{N}
    • Sample Variance (s2s^2): Computed from a sample of size nn:       s2=∑(xi−xˉ)2n−1s^2 = \frac{\sum (x_i - \bar{x})^2}{n - 1}
    • Standard Deviation: The positive square root of the variance, expressed in original data units:
    • Population Standard Deviation: σ=σ2=∑(xi−μ)2N\sigma = \sqrt{\sigma^2} = \sqrt{\frac{\sum (x_i - \mu)^2}{N}}
    • Sample Standard Deviation: s=s2=∑(xi−xˉ)2n−1s = \sqrt{s^2} = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n - 1}}
    • Frequency Distribution Standard Deviation Formula:s=∑fi⋅(xi−xˉ)2(∑fi)−1s = \sqrt{\frac{\sum f_i \cdot (x_i - \bar{x})^2}{(\sum f_i) - 1}}
    • Theoretical Basis for Bessel's Correction (n−1n - 1 in Denominator):
    • The population variance σ2\sigma^2 is a fixed population parameter.
    • Deviations calculated relative to the sample mean xˉ\bar{x} are smaller than deviations calculated relative to the true population mean μ\mu.
    • Dividing by nn underestimates the population variance, producing a biased downward estimator.
    • Dividing by n−1n - 1 (the degrees of freedom) provides an unbiased estimator of σ2\sigma^2.
    • Sample Standard Deviation Computation from a Frequency Table:
    • Given sample mean xˉ=2.9\bar{x} = 2.9 and sample size ∑fi=20\sum f_i = 20:
      • x1=1x_1 = 1: x1−xˉ=−1.9x_1 - \bar{x} = -1.9, (x1−xˉ)2=3.61(x_1 - \bar{x})^2 = 3.61, f1=3f_1 = 3, f1(x1−xˉ)2=10.83f_1(x_1 - \bar{x})^2 = 10.83
      • x2=2x_2 = 2: x2−xˉ=−0.9x_2 - \bar{x} = -0.9, (x2−xˉ)2=0.81(x_2 - \bar{x})^2 = 0.81, f2=4f_2 = 4, f2(x2−xˉ)2=3.24f_2(x_2 - \bar{x})^2 = 3.24
      • x3=3x_3 = 3: x3−xˉ=0.1x_3 - \bar{x} = 0.1, (x3−xˉ)2=0.01(x_3 - \bar{x})^2 = 0.01, f3=8f_3 = 8, f3(x3−xˉ)2=0.08f_3(x_3 - \bar{x})^2 = 0.08
      • x4=4x_4 = 4: x4−xˉ=1.1x_4 - \bar{x} = 1.1, (x4−xˉ)2=1.21(x_4 - \bar{x})^2 = 1.21, f4=2f_4 = 2, f4(x4−xˉ)2=2.42f_4(x_4 - \bar{x})^2 = 2.42
      • x5=5x_5 = 5: x5−xˉ=2.1x_5 - \bar{x} = 2.1, (x5−xˉ)2=4.41(x_5 - \bar{x})^2 = 4.41, f5=3f_5 = 3, f5(x5−xˉ)2=13.23f_5(x_5 - \bar{x})^2 = 13.23
    • Sum of squared deviations: ∑fi(xi−xˉ)2=10.83+3.24+0.08+2.42+13.23=29.80\sum f_i (x_i - \bar{x})^2 = 10.83 + 3.24 + 0.08 + 2.42 + 13.23 = 29.80.
    • Sample variance:       s2=29.8020−1=29.8019≈1.5684s^2 = \frac{29.80}{20 - 1} = \frac{29.80}{19} \approx 1.5684
    • Sample standard deviation:       s=1.5684≈1.252s = \sqrt{1.5684} \approx 1.252
  • Relative Standing: Standardized zz-Scores:

    • A zz-score indicates the number of standard deviations a given raw data value xx lies above or below the mean:     z=x−μσz = \frac{x - \mu}{\sigma}
    • Academic Test Performance Comparison:
    • Test A: Raw score x=59x = 59, class average μ=57\mu = 57, standard deviation σ=0.8\sigma = 0.8.       zA=59−570.8=20.8=2.5z_A = \frac{59 - 57}{0.8} = \frac{2}{0.8} = 2.5
    • Test B: Raw score x=64x = 64, class average μ=62\mu = 62, standard deviation σ=1.1\sigma = 1.1.       zB=64−621.1=21.1≈1.82z_B = \frac{64 - 62}{1.1} = \frac{2}{1.1} \approx 1.82
    • Performance on Test A is comparatively better because the student scored 2.52.5 standard deviations above the class mean, compared to 1.821.82 standard deviations above the mean on Test B.
    • Typing Speed Suitability Comparison:
    • In typing evaluations, a shorter completion time represents superior performance; therefore, a more negative zz-score is optimal.
    • Pharmacy: Personal time =3.3 min= 3.3\,min, average =3.1 min= 3.1\,min, standard deviation =0.3 min= 0.3\,min.       z=3.3−3.10.3=0.20.3≈+0.67z = \frac{3.3 - 3.1}{0.3} = \frac{0.2}{0.3} \approx +0.67
    • Bank: Personal time =6.4 min= 6.4\,min, average =7.0 min= 7.0\,min, standard deviation =0.6 min= 0.6\,min.       z=6.4−7.00.6=−0.60.6=−1.0z = \frac{6.4 - 7.0}{0.6} = \frac{-0.6}{0.6} = -1.0
    • Law firm: Personal time =8.0 min= 8.0\,min, average =8.4 min= 8.4\,min, standard deviation =0.45 min= 0.45\,min.       z=8.0−8.40.45=−0.40.45≈−0.89z = \frac{8.0 - 8.4}{0.45} = \frac{-0.4}{0.45} \approx -0.89
    • The applicant is most competitive for the position at the Bank, being 1.01.0 full standard deviation faster than the workplace average.
  • Quartiles and Cumulative Frequency Distributions:

    • Quartile Location Formulas:
    • For discrete datasets of size nn:       Location of Q1=14(n+1)\text{Location of } Q_1 = \frac{1}{4}(n + 1)Location of Q2=12(n+1)\text{Location of } Q_2 = \frac{1}{2}(n + 1)Location of Q3=34(n+1)\text{Location of } Q_3 = \frac{3}{4}(n + 1)
    • For continuous datasets of size nn:       Location of Q1=14n\text{Location of } Q_1 = \frac{1}{4}nLocation of Q2=12n\text{Location of } Q_2 = \frac{1}{2}nLocation of Q3=34n\text{Location of } Q_3 = \frac{3}{4}n
    • Cumulative Frequency Ogive Analysis:

Cumulative frequency ogive of flight delays

* Total observations in flight sample: n=50n = 50.
* Median (Q2Q_2): Corresponds to cumulative frequency of 2525. Reading across to the curve yields a median delay of approximately 62 min62\,min.
* First Quartile (Q1Q_1): Corresponds to cumulative frequency of 12.512.5, which aligns with approximately 51 min51\,min.
* Third Quartile (Q3Q_3): Corresponds to cumulative frequency of 37.537.5, aligning with approximately 74 min74\,min.
* Interquartile Range (IQR\text{IQR}):

      IQR=Q3−Q1≈74−51=23 min\text{IQR} = Q_3 - Q_1 \approx 74 - 51 = 23\,min * Flights between 4040 and 50 min50\,min late: At 40 min40\,min, cumulative frequency is ≈5\approx 5; at 50 min50\,min, cumulative frequency is ≈11\approx 11. Count is 11−5=611 - 5 = 6 flights. * Flights more than 80 min80\,min late: Cumulative frequency at 80 min80\,min is ≈42\approx 42. Count is 50−42=850 - 42 = 8 flights.

  • The Five-Number Summary, Boxplots, and Outlier Identification:
    • Five-Number Summary Components: Minimum, First Quartile (Q1Q_1), Median (MM or Q2Q_2), Third Quartile (Q3Q_3), and Maximum.
    • Interquartile Range (IQR\text{IQR}):IQR=Q3−Q1\text{IQR} = Q_3 - Q_1
    • Tukey's Outlier Fences:
    • Lower Outlier Threshold: Q1−1.5×IQRQ_1 - 1.5 \times \text{IQR}
    • Upper Outlier Threshold: Q3+1.5×IQRQ_3 + 1.5 \times \text{IQR}
    • Any data observation falling strictly outside the boundary interval (Q1−1.5×IQR,  Q3+1.5×IQR)(Q_1 - 1.5 \times \text{IQR},\; Q_3 + 1.5 \times \text{IQR}) is categorized as a suspected outlier.

Anatomy of a box plot

  • Boxplot Anatomy:

    • Central box spans from Q1Q_1 to Q3Q_3, representing the middle 50%50\% of the distribution (IQR\text{IQR}).

    • Median marked by a line dividing the box.

    • Whiskers extend outward from the box edges to the most extreme data values situated within the 1.5×IQR1.5 \times \text{IQR} fences.

    • Outliers are plotted as individual points beyond the whiskers.

    • Outlier Principle: Outliers do not cause skewness; when classifying the overall shape of a dataset, isolated outliers should not be enclosed in a manufactured continuous tail.

    • Distributional Shapes and Matching:

Distribution shapes: symmetric, left-skewed, right-skewed

  • Symmetric Distribution: Balanced around the center. The mean, median, and mode coincide at the central axis of symmetry.
  • Right-Skewed (Positively Skewed) Distribution: Elongated tail extending toward higher positive values. Extreme values inflate the mean:     Mode<Median<Mean\text{Mode} < \text{Median} < \text{Mean}
  • Left-Skewed (Negatively Skewed) Distribution: Elongated tail extending toward lower values. Extreme low values depress the mean:     Mean<Median<Mode\text{Mean} < \text{Median} < \text{Mode}
  • Histogram and Boxplot Matching:

Matching histograms to box plots

* **Histogram (a):** Symmetric, bell-shaped distribution centered at 6060. Matches **Boxplot (2)**, which features a centered median at 6060 and symmetric whiskers with two-tailed outliers.
* **Histogram (b):** Uniform, rectangular distribution spanning from 00 to 100100. Matches **Boxplot (3)**, which has a median at 5050, wide, evenly distributed quartiles, and whiskers spanning the range [0,100][0, 100].
* **Histogram (c):** Heavily right-skewed distribution concentrated between 00 and 22 with a long right tail extending to 66. Matches **Boxplot (1)**, which exhibits a low median around 1.31.3, compressed lower quartile, and multiple extreme outliers reaching past 66.

Principles of Experimental Design

  • Observational Studies versus Controlled Experiments:

    • Observational Study: Researchers observe or record outcomes without manipulating exposure or assigning treatments. In a retrospective study, existing historical records are reviewed; in a prospective study, subjects are identified in advance and observed prospectively as real-time events unfold.
    • Controlled Experiment: Investigators deliberately impose specific experimental conditions (treatments) upon experimental units to evaluate causal relationships between explanatory factors and response variables.
  • Core Experimental Terminology:

    • Experimental Units: The fundamental entities or individuals on which the experiment is performed (referred to as human subjects when participants are people).
    • Factors: Explanatory variables manipulated by the experimenter.
    • Treatments: Specific experimental conditions applied to units. If an experiment investigates a single factor, treatments correspond to the levels of that factor; in multi-factor designs, treatments are combinations of factor levels.
    • Response Variable: The measured outcome reflecting the effect of the treatments.
  • The Four Principles of Experimental Design:

    • 1. Control: Restricting extraneous sources of variability by comparing treatment groups against control baselines, standardizing environments, and utilizing blinding to isolate the treatment effect.
    • 2. Randomize: Using chance mechanisms to assign experimental units to treatments, balancing unobserved confounding variables across groups.
    • 3. Replicate: Applying treatments to multiple experimental units to ensure outcomes are statistically reliable and not due to chance.
    • 4. Block: Grouping similar experimental units together into homogeneous sets (blocks) to control for known sources of variation before allocating treatments within each block.

Randomized experimental design schematic

  • Agricultural Design Scenario ("Lend Me Your Ears"):
    • Scenario: A farming trial tests whether a new fertilizer increases corn yields on a 30-acre30\text{-acre} plot without fully abandoning organic methods.
    • Experimental Units: The 30 individual plots30\text{ individual plots} (e.g., divided into 1-acre1\text{-acre} subplots) across the acreage.
    • Factor: Fertilizer type and composition.
    • Treatments: Varying concentrations of the new fertilizer blended with standard organic fertilizer compared to a pure organic control group.
    • Response Variable: Crop yield measured in bushels or number of ears of corn harvested per acre.
    • Study Classification: A controlled experiment because researchers deliberately manipulate fertilizer applications across plots rather than passively observing existing yields.
    • Replication Requirement: Replication is necessary across multiple distinct plots per treatment to measure random experimental error.
    • Random Allocation Method: Using a random number generator to randomly assign the 3030 plots into treatment conditions.
    • Blinding Feasibility: Double-blinding is not possible for the plants, but can be maintained for the field assessors evaluating final ear counts and yields.
    • Controllable Sources of Variability: Soil composition, baseline moisture, sunlight exposure, irrigation schedules, and seed strain can be controlled via blocking (grouping similar soil patches together before randomization).

Probability Foundations, Sample Spaces, and Set Operations

  • Basic Probability Concepts:

    • Experiment: Any act or observational process yielding an outcome that cannot be predicted with certainty.
    • Sample Point: The most basic elementary outcome of an experiment.
    • Sample Space (SS): The collection of all possible sample points for an experiment.
    • Event: Any specific collection or subset of sample points from the sample space.
    • Probability of an Event (AA): The sum of the probabilities assigned to the individual sample points comprising event AA:     P(A)=∑Ei∈AP(Ei)P(A) = \sum_{E_i \in A} P(E_i)
    • Classical Probability Definition: For equally likely sample points:     P(A)=Number of favorable outcomes in event ATotal number of possible outcomes in sample space SP(A) = \frac{\text{Number of favorable outcomes in event } A}{\text{Total number of possible outcomes in sample space } S}
    • Axioms of Probability:0≤P(A)≤10 \le P(A) \le 1P(S)=1P(S) = 1
  • Step-by-Step Procedure for Calculating Event Probabilities:

    • Step 1: Define the experiment by describing the physical observation process and recorded variables.
    • Step 2: List all distinct sample points in the sample space SS.
    • Step 3: Assign valid probabilities to each sample point such that 0≤P(Ei)≤10 \le P(E_i) \le 1 and ∑P(Ei)=1\sum P(E_i) = 1.
    • Step 4: Identify the collection of sample points belonging to the event of interest AA.
    • Step 5: Sum the individual probabilities of those sample points to determine P(A)P(A).
  • Standard Six-Sided Die Examples:

    • Sample space: S={1,2,3,4,5,6}S = \{1, 2, 3, 4, 5, 6\}, with each outcome having probability P(Ei)=16P(E_i) = \frac{1}{6}.
    • Probability of rolling a 11:     P({1})=16P(\{1\}) = \frac{1}{6}
    • Probability of rolling a 11 or a 22:     P({1,2})=16+16=26=13P(\{1, 2\}) = \frac{1}{6} + \frac{1}{6} = \frac{2}{6} = \frac{1}{3}
    • Probability of rolling a number in the set (1,2,3,4,5,6)(1, 2, 3, 4, 5, 6):     P(S)=66=1.0=100%P(S) = \frac{6}{6} = 1.0 = 100\%
    • Probability of rolling two 1s on two consecutive independent rolls:     P(Double 1)=16×16=136P(\text{Double } 1) = \frac{1}{6} \times \frac{1}{6} = \frac{1}{36}
  • Set Operations on Events:

Venn diagrams illustrating union and intersection

  • Union (A∪BA \cup B): The event that occurs if either event AA or event BB (or both) occurs on a single performance of the experiment. Symbolized as A∪BA \cup B, comprising all sample points belonging to AA, BB, or both.

  • Intersection (A∩BA \cap B): The event that occurs if both event AA and event BB occur simultaneously on a single trial. Denoted as A∩BA \cap B, comprising all sample points belonging to both sets.

  • Complement (AcA^c or Aˉ\bar{A}): The event that event AA does not occur, comprising all sample points in SS that do not belong to AA.

  • Rule of Complements: The sum of the probabilities of an event and its complement equals 11:     P(A)+P(Ac)=1  ⟹  P(Ac)=1−P(A)P(A) + P(A^c) = 1 \implies P(A^c) = 1 - P(A)

  • General Addition Rule: For any two events AA and BB (disjoint or non-disjoint):     P(A∪B)=P(A)+P(B)−P(A∩B)P(A \cup B) = P(A) + P(B) - P(A \cap B)

  • Exhaustive Events: Two events AA and BB are exhaustive if their union equals the entire sample space, such that P(A∪B)=1P(A \cup B) = 1.

  • Mutually Exclusive (Disjoint) Events: Two events AA and BB cannot occur simultaneously; their intersection contains no sample points:     P(A∩B)=0  ⟹  P(A∪B)=P(A)+P(B)P(A \cap B) = 0 \implies P(A \cup B) = P(A) + P(B)

  • Independent Events: Two events AA and BB are independent if the occurrence of one does not change the probability of the other:     P(A∣B)=P(A)andP(B∣A)=P(B)P(A|B) = P(A) \quad \text{and} \quad P(B|A) = P(B)P(A∩B)=P(A)⋅P(B)P(A \cap B) = P(A) \cdot P(B)

    • Discrete Sample Spaces and Household Probability Exercises:
  • Discrete Probability Space with Undetermined Probabilities:

    • Sample space S={a,b,c,d,e,f}S = \{a, b, c, d, e, f\}, with probabilities P(a)=0.12P(a) = 0.12, P(b)=0.16P(b) = 0.16, P(d)=0.14P(d) = 0.14, P(e)=0.20P(e) = 0.20, and P(c)=P(f)P(c) = P(f).
    • Sum of known outcomes:       0.12+0.16+0.14+0.20=0.620.12 + 0.16 + 0.14 + 0.20 = 0.62
    • Total remaining probability:       P(c)+P(f)=1.00−0.62=0.38  ⟹  P(c)=P(f)=0.382=0.19P(c) + P(f) = 1.00 - 0.62 = 0.38 \implies P(c) = P(f) = \frac{0.38}{2} = 0.19
    • Events are defined as Y={a,c,e}Y = \{a, c, e\} and Z={b,c,d,f}Z = \{b, c, d, f\}.       P(Y)=P(a)+P(c)+P(e)=0.12+0.19+0.20=0.51P(Y) = P(a) + P(c) + P(e) = 0.12 + 0.19 + 0.20 = 0.51P(Z)=P(b)+P(c)+P(d)+P(f)=0.16+0.19+0.14+0.19=0.68P(Z) = P(b) + P(c) + P(d) + P(f) = 0.16 + 0.19 + 0.14 + 0.19 = 0.68
  • Five-Element Sample Space Model:

    • Sample space S={a,b,c,d,e}S = \{a, b, c, d, e\} with values P(a)=0.18P(a) = 0.18, P(b)=0.16P(b) = 0.16, P(d)=0.24P(d) = 0.24, P(e)=0.21P(e) = 0.21.
    • Probability of cc:       P(c)=1.00−(0.18+0.16+0.24+0.21)=1.00−0.79=0.21P(c) = 1.00 - (0.18 + 0.16 + 0.24 + 0.21) = 1.00 - 0.79 = 0.21
    • For events Y={a,c,e}Y = \{a, c, e\} and Z={b,c,d}Z = \{b, c, d\}:       P(Y)=0.18+0.21+0.21=0.60P(Y) = 0.18 + 0.21 + 0.21 = 0.60P(Z)=0.16+0.21+0.24=0.61P(Z) = 0.16 + 0.21 + 0.24 = 0.61
  • Household Pet Ownership Analysis (30 Households):

    • Total households =30= 30.
    • 55 households own no pets; therefore, 30−5=2530 - 5 = 25 households own pets.
    • 1313 households own a cat; 77 households own both a cat and a dog.
    • Households owning only a cat: 13−7=613 - 7 = 6.
    • Households owning only a dog: 25−13=1225 - 13 = 12.
    • Probability that a randomly selected household owns only a pet dog:       P(Only Dog)=1230=0.40P(\text{Only Dog}) = \frac{12}{30} = 0.40
  • Household Pet Ownership Analysis (20 Households):

    • Total households =20= 20.

    • 33 households own no pets; 20−3=1720 - 3 = 17 households own pets.

    • 1212 households own a cat; 77 households own both a cat and a dog.

    • Households owning only a dog: 17−12=517 - 12 = 5.

    • Probability of owning only a dog:       P(Only Dog)=520=0.25P(\text{Only Dog}) = \frac{5}{20} = 0.25

    • Sampling Without Replacement and Permutations:

  • Marble Drawing Experiment:

    • A box contains 1010 white (WW) and 1010 black (BB) marbles (2020 total). Two marbles are drawn in succession without replacement.
    • Sample space of ordered outcomes:       S={WW,WB,BW,BB}S = \{WW, WB, BW, BB\}
    • Probabilities of individual paths:       P(WW)=1020×919=90380=938P(WW) = \frac{10}{20} \times \frac{9}{19} = \frac{90}{380} = \frac{9}{38}P(WB)=1020×1019=100380=1038P(WB) = \frac{10}{20} \times \frac{10}{19} = \frac{100}{380} = \frac{10}{38}P(BW)=1020×1019=100380=1038P(BW) = \frac{10}{20} \times \frac{10}{19} = \frac{100}{380} = \frac{10}{38}P(BB)=1020×919=90380=938P(BB) = \frac{10}{20} \times \frac{9}{19} = \frac{90}{380} = \frac{9}{38}
  • Three-Child Family Sample Space by Birth Order:

    • Let BB represent a boy and GG represent a girl. The complete sample space is:       S={BBB,BBG,BGB,BGG,GBB,GBG,GGB,GGG}S = \{BBB, BBG, BGB, BGG, GBB, GBG, GGB, GGG\}
    • Assuming independent, equally likely births with P(B)=P(G)=0.5P(B) = P(G) = 0.5 (88 outcomes, each with probability 18\frac{1}{8}):
      • a. At least one child is a girl: Comprises all outcomes except BBBBBB (77 outcomes). Probability =78= \frac{7}{8}.
      • b. At most one child is a girl: Comprises outcomes with zero or one girl: (BBB,BBG,BGB,GBB)(BBB, BBG, BGB, GBB) (44 outcomes). Probability =48=12= \frac{4}{8} = \frac{1}{2}.
      • c. All children are girls: (GGG)(GGG) (11 outcome). Probability =18= \frac{1}{8}.
      • d. Exactly two children are girls: (BGG,GBG,GGB)(BGG, GBG, GGB) (33 outcomes). Probability =38= \frac{3}{8}.
      • e. The first-born is a girl: (GBB,GBG,GGB,GGG)(GBB, GBG, GGB, GGG) (44 outcomes). Probability =48=12= \frac{4}{8} = \frac{1}{2}.
  • String Permutations Over Digit Alphabet:

    • Constructing a 6-digit sequence from the set (0,1,2,3)(0, 1, 2, 3). Total possible strings: 46=40964^6 = 4096.

    • Case 1: String begins and ends with 00 (format: 0−−−−00 - - - - 0). Number of arrangements: 1×4×4×4×4×1=44=2561 \times 4 \times 4 \times 4 \times 4 \times 1 = 4^4 = 256.       P(0−−−−0)=2564096=116=0.0625P(0----0) = \frac{256}{4096} = \frac{1}{16} = 0.0625

    • Case 2: String begins with 2323 and ends with 00 (format: 23−−−02 3 - - - 0). Number of arrangements: 1×1×4×4×4×1=43=641 \times 1 \times 4 \times 4 \times 4 \times 1 = 4^3 = 64.       P(23−−−0)=644096=164=0.015625P(23---0) = \frac{64}{4096} = \frac{1}{64} = 0.015625

    • Tree Diagrams and Joint Dice Rolls:

Tree diagram of two coin tosses

  • Two Fair Dice Rolling Space (3636 Elementary Outcomes):

Sample space grid for rolling two dice

* (i) Sum equals eight: Outcomes are (2,6),(3,5),(4,4),(5,3),(6,2)(2,6), (3,5), (4,4), (5,3), (6,2). Total =5= 5.

      P(Sum=8)=536P(\text{Sum} = 8) = \frac{5}{36} * (ii) Sum equals three: Outcomes are (1,2),(2,1)(1,2), (2,1). Total =2= 2.       P(Sum=3)=236=118P(\text{Sum} = 3) = \frac{2}{36} = \frac{1}{18} * (iii) Absolute difference equals zero (doubles): Outcomes are (1,1),(2,2),(3,3),(4,4),(5,5),(6,6)(1,1), (2,2), (3,3), (4,4), (5,5), (6,6). Total =6= 6.       P(∣X1−X2∣=0)=636=16P(|X_1 - X_2| = 0) = \frac{6}{36} = \frac{1}{6} * (iv) Sum is a prime number: Possible primes are 2,3,5,7,112, 3, 5, 7, 11. * Sum = 2: (1,1)(1,1) (11 outcome) * Sum = 3: (1,2),(2,1)(1,2), (2,1) (22 outcomes) * Sum = 5: (1,4),(2,3),(3,2),(4,1)(1,4), (2,3), (3,2), (4,1) (44 outcomes) * Sum = 7: (1,6),(2,5),(3,4),(4,3),(5,2),(6,1)(1,6), (2,5), (3,4), (4,3), (5,2), (6,1) (66 outcomes) * Sum = 11: (5,6),(6,5)(5,6), (6,5) (22 outcomes) * Total favorable outcomes: 1+2+4+6+2=151 + 2 + 4 + 6 + 2 = 15.       P(Sum is Prime)=1536=512P(\text{Sum is Prime}) = \frac{15}{36} = \frac{5}{12}

  • Dual Spinner Game Experiments:

Spinners A and B

  • Setup 1 (Unequal/Equal Sectors):
    • Spinner A: Sector 11 covers half the circle (P(1)=12P(1) = \frac{1}{2}); sectors 22 and 55 cover quarter circles (P(2)=14P(2) = \frac{1}{4}, P(5)=14P(5) = \frac{1}{4}).
    • Spinner B: Four equal quarter quadrants labeled 1,2,3,41, 2, 3, 4 (each with probability 14\frac{1}{4}).
    • Total joint outcomes: 3×4=123 \times 4 = 12 ordered pairs.
    • Sum equals 55: Favorable pairs are (1,4)(1, 4), (2,3)(2, 3).       P(1,4)=12×14=18=216P(1, 4) = \frac{1}{2} \times \frac{1}{4} = \frac{1}{8} = \frac{2}{16}P(2,3)=14×14=116P(2, 3) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(Sum=5)=216+116=316P(\text{Sum} = 5) = \frac{2}{16} + \frac{1}{16} = \frac{3}{16}
    • Sum equals 66: Favorable pairs are (2,4)(2, 4), (5,1)(5, 1).       P(2,4)=14×14=116P(2, 4) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(5,1)=14×14=116P(5, 1) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(Sum=6)=116+116=216=18P(\text{Sum} = 6) = \frac{1}{16} + \frac{1}{16} = \frac{2}{16} = \frac{1}{8}
    • Absolute difference equals 11: Favorable pairs are (1,2)(1, 2), (2,1)(2, 1), (2,3)(2, 3), (5,4)(5, 4).       P(1,2)=12×14=216P(1, 2) = \frac{1}{2} \times \frac{1}{4} = \frac{2}{16}P(2,1)=14×14=116P(2, 1) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(2,3)=14×14=116P(2, 3) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(5,4)=14×14=116P(5, 4) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(∣A−B∣=1)=2+1+1+116=516P(|A - B| = 1) = \frac{2 + 1 + 1 + 1}{16} = \frac{5}{16}
    • Absolute difference equals 22: Favorable pairs are (1,3)(1, 3), (2,4)(2, 4), (5,3)(5, 3).       P(1,3)=12×14=216P(1, 3) = \frac{1}{2} \times \frac{1}{4} = \frac{2}{16}P(2,4)=14×14=116P(2, 4) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(5,3)=14×14=116P(5, 3) = \frac{1}{4} \times \frac{1}{4} = \frac{1}{16}P(∣A−B∣=2)=2+1+116=416=14P(|A - B| = 2) = \frac{2 + 1 + 1}{16} = \frac{4}{16} = \frac{1}{4}
  • Setup 2 (Two Four-Quadrant Spinners):
    • Spinner 1: Equal quadrants 1,2,4,51, 2, 4, 5 (P=14P = \frac{1}{4} each).
    • Spinner 2: Equal quadrants 0,1,3,50, 1, 3, 5 (P=14P = \frac{1}{4} each).
    • Total equally likely outcomes: 4×4=164 \times 4 = 16, each having probability 116\frac{1}{16}.
    • (i) Identical scores: (1,1),(5,5)(1, 1), (5, 5) (22 outcomes). Probability =216=18= \frac{2}{16} = \frac{1}{8}.
    • (ii) Both scores are prime: Primes on Spinner 1 are (2,5)(2, 5); primes on Spinner 2 are (3,5)(3, 5). Combinations: (2,3),(2,5),(5,3),(5,5)(2, 3), (2, 5), (5, 3), (5, 5) (44 outcomes). Probability =416=14= \frac{4}{16} = \frac{1}{4}.
    • (iii) Sum is odd: Requires one even and one odd number.
      • Even on 1 (2,42, 4) and Odd on 2 (1,3,51, 3, 5): 2×3=62 \times 3 = 6 outcomes.
      • Odd on 1 (1,51, 5) and Even on 2 (00): 2×1=22 \times 1 = 2 outcomes.
      • Total favorable: 6+2=86 + 2 = 8 outcomes. Probability =816=12= \frac{8}{16} = \frac{1}{2}.
    • (iv) Sum is divisible by 55: Pairs summing to 55 or 1010:
      • Sum = 5: (2,3),(4,1),(5,0)(2, 3), (4, 1), (5, 0) (33 outcomes).
      • Sum = 10: (5,5)(5, 5) (11 outcome).
      • Total favorable: 44 outcomes. Probability =416=14= \frac{4}{16} = \frac{1}{4}.
    • (v) Product is nonzero: Requires Spinner 2 to avoid landing on 00. Spinner 2 lands on 1,3,51, 3, 5 in 33 out of 44 outcomes. Probability =4×316=1216=34= \frac{4 \times 3}{16} = \frac{12}{16} = \frac{3}{4}.

Contingency Tables and Conditional Probability

  • The Conditional Probability Theorem and Multiplicative Rule:

    • Conditional Probability Formula: For two events AA and BB, assuming P(B)≠0P(B) \neq 0:     P(A∣B)=P(A∩B)P(B)P(A|B) = \frac{P(A \cap B)}{P(B)}
    • Multiplicative Rule of Probability:P(A∩B)=P(B)⋅P(A∣B)=P(A)⋅P(B∣A)P(A \cap B) = P(B) \cdot P(A|B) = P(A) \cdot P(B|A)
  • Speeding Violations and Distracted Driving Study:

    • Given the empirical study table (N=755N = 755 drivers):
Speeding violation in last yearNo violation in last yearTotal
Uses cell phone while driving2525280280305305
Does not use cell phone4545405405450450
Total7070685685755755
  • Probability Computations:

    • a. Driver is a cell phone user:       P(Cell Phone User)=305755≈0.4040P(\text{Cell Phone User}) = \frac{305}{755} \approx 0.4040

    • b. Driver had no speeding violation in the last year:       P(No Violation)=685755≈0.9073P(\text{No Violation}) = \frac{685}{755} \approx 0.9073

    • c. Driver had no violation AND was a cell phone user:       P(No Violation∩Cell Phone User)=280755≈0.3709P(\text{No Violation} \cap \text{Cell Phone User}) = \frac{280}{755} \approx 0.3709

    • d. Driver is a cell phone user OR had no violation:       P(Cell Phone User∪No Violation)=305755+685755−280755=710755≈0.9404P(\text{Cell Phone User} \cup \text{No Violation}) = \frac{305}{755} + \frac{685}{755} - \frac{280}{755} = \frac{710}{755} \approx 0.9404

    • e. Driver is a cell phone user GIVEN they had a violation:       P(Cell Phone User∣Violation)=2570=514≈0.3571P(\text{Cell Phone User} | \text{Violation}) = \frac{25}{70} = \frac{5}{14} \approx 0.3571

    • U.S. Senate Election Affiliation Study:

  • Cross-tabulation of Senate seats up for re-election:

Up for reelectionDemocratic PartyRepublican PartyOtherTotal
Year A20201313003333
Year B10102424003434
Total30303737006767
  • Probability Questions:

    • Senator has an "Other" affiliation:       P(Other)=067=0P(\text{Other}) = \frac{0}{67} = 0

    • Senator is up for re-election in Year B:       P(Year B)=3467≈0.5075P(\text{Year B}) = \frac{34}{67} \approx 0.5075

    • Senator is a Democrat AND up for re-election in Year B:       P(Democrat∩Year B)=1067≈0.1493P(\text{Democrat} \cap \text{Year B}) = \frac{10}{67} \approx 0.1493

    • Senator is a Republican OR is up for re-election in Year A:       P(Republican∪Year A)=3767+3367−1367=5767≈0.8507P(\text{Republican} \cup \text{Year A}) = \frac{37}{67} + \frac{33}{67} - \frac{13}{67} = \frac{57}{67} \approx 0.8507

    • Senator is a Democrat GIVEN they are up for re-election in Year B:       P(Democrat∣Year B)=1034=517≈0.2941P(\text{Democrat} | \text{Year B}) = \frac{10}{34} = \frac{5}{17} \approx 0.2941

    • Senator is up for re-election in Year A GIVEN they are a Republican:       P(Year A∣Republican)=1337≈0.3514P(\text{Year A} | \text{Republican}) = \frac{13}{37} \approx 0.3514

    • Independence / Mutually Exclusive Test ("Republican" and "Year B"):

      • Joint probability: P(Republican∩Year B)=2467≠0P(\text{Republican} \cap \text{Year B}) = \frac{24}{67} \neq 0, so they are not mutually exclusive.
      • Product of marginals:         P(Republican)×P(Year B)=3767×3467≈0.5522×0.5075≈0.2802≠2467≈0.3582P(\text{Republican}) \times P(\text{Year B}) = \frac{37}{67} \times \frac{34}{67} \approx 0.5522 \times 0.5075 \approx 0.2802 \neq \frac{24}{67} \approx 0.3582
      • The events are neither mutually exclusive nor independent.
    • Independence / Mutually Exclusive Test ("Other" and "Year B"):

      • Joint probability: P(Other∩Year B)=067=0P(\text{Other} \cap \text{Year B}) = \frac{0}{67} = 0, so they are mutually exclusive.
      • Since P(Other)=0P(\text{Other}) = 0, P(Other∩Year B)=0=P(Other)×P(Year B)P(\text{Other} \cap \text{Year B}) = 0 = P(\text{Other}) \times P(\text{Year B}).
      • The events are both mutually exclusive and independent.
    • Single-Topping Pizza Ordering Contingency Table:

  • Reconstructed Lunch Table (N=120N = 120 orders):

SizeVeggieCheeseMeat loversAnchoviesTotal
Small1313441616223535
Medium21215588333737
Large1414272766114848
Total48483636303066120120
  • Step-by-Step Derivation of Missing Cells:

    • Small Meat lovers: 35−(13+4+2)=1635 - (13 + 4 + 2) = 16.
    • Medium Veggie: 48−(13+14)=2148 - (13 + 14) = 21.
    • Medium Total: 21+5+8+3=3721 + 5 + 8 + 3 = 37.
    • Large Total: 120−(35+37)=48120 - (35 + 37) = 48.
    • Large Cheese: 48−(14+6+1)=2748 - (14 + 6 + 1) = 27.
    • Column Totals: Cheese =4+5+27=36= 4 + 5 + 27 = 36; Meat lovers =16+8+6=30= 16 + 8 + 6 = 30; Anchovies =2+3+1=6= 2 + 3 + 1 = 6.
  • Probability Calculations:

    • (i) A cheese pizza is ordered:       P(Cheese)=36120=310=0.30P(\text{Cheese}) = \frac{36}{120} = \frac{3}{10} = 0.30

    • (ii) A large pizza is ordered given it is meat lovers:       P(Large∣Meat lovers)=630=15=0.20P(\text{Large} | \text{Meat lovers}) = \frac{6}{30} = \frac{1}{5} = 0.20

    • (iii) Meat lovers pizza is ordered given it is large:       P(Meat lovers∣Large)=648=18=0.125P(\text{Meat lovers} | \text{Large}) = \frac{6}{48} = \frac{1}{8} = 0.125

    • (iv) Veggie pizza is ordered assuming it is not large:

      • Non-large orders: Small ++ Medium =35+37=72= 35 + 37 = 72.
      • Veggie non-large orders: 13+21=3413 + 21 = 34.       P(Veggie∣Not Large)=3472=1736≈0.4722P(\text{Veggie} | \text{Not Large}) = \frac{34}{72} = \frac{17}{36} \approx 0.4722
    • Provincial Election Support Contingency Table:

  • Reconstructed Political Survey Table (N=630N = 630 supporters):

Political PartyFavours early electionsFavours scheduled electionsTotal
Liberal party9090110110200200
NDP party60607575135135
Green party181852527070
Tory party112112113113225225
Total280280350350630630
  • Derivation of Table Values:

    • Liberal Scheduled: 200−90=110200 - 90 = 110.
    • Green Early: 70−52=1870 - 52 = 18.
    • NDP Early: 280−(90+18+112)=280−220=60280 - (90 + 18 + 112) = 280 - 220 = 60.
    • NDP Scheduled: 135−60=75135 - 60 = 75.
    • Total Scheduled: 630−280=350630 - 280 = 350.
  • Conditional Probabilities:

    • A Tory party supporter favours early elections:       P(Early∣Tory)=112225≈0.4978P(\text{Early} | \text{Tory}) = \frac{112}{225} \approx 0.4978

    • A supporter does not favour early elections (prefers scheduled) given they belong to the NDP:       P(Scheduled∣NDP)=75135=59≈0.5556P(\text{Scheduled} | \text{NDP}) = \frac{75}{135} = \frac{5}{9} \approx 0.5556

    • A supporter belongs to the Liberal party given they prefer scheduled elections:       P(Liberal∣Scheduled)=110350=1135≈0.3143P(\text{Liberal} | \text{Scheduled}) = \frac{110}{350} = \frac{11}{35} \approx 0.3143

    • A supporter favours early elections given they belong to the Green party:       P(Early∣Green)=1870=935≈0.2571P(\text{Early} | \text{Green}) = \frac{18}{70} = \frac{9}{35} \approx 0.2571

    • Specialized Probability Scenarios:

  • Restaurant Pizza and Dessert Analysis:

    • Let AA be the event a patron orders pizza, and BB be the event a patron orders a banana split.
    • Given: P(A)=0.4P(A) = 0.4, P(B∣A)=0.36P(B|A) = 0.36, and P(B∣Ac)=0.30P(B|A^c) = 0.30, with P(Ac)=1−0.4=0.6P(A^c) = 1 - 0.4 = 0.6.
    • (i) Probability a patron orders both a pizza and a banana split:       P(A∩B)=P(A)⋅P(B∣A)=0.4×0.36=0.144P(A \cap B) = P(A) \cdot P(B|A) = 0.4 \times 0.36 = 0.144
    • (ii) Probability a patron orders a banana split (by the Law of Total Probability):       P(B)=P(A∩B)+P(Ac∩B)=0.144+P(Ac)⋅P(B∣Ac)P(B) = P(A \cap B) + P(A^c \cap B) = 0.144 + P(A^c) \cdot P(B|A^c)P(B)=0.144+(0.6×0.30)=0.144+0.180=0.324P(B) = 0.144 + (0.6 \times 0.30) = 0.144 + 0.180 = 0.324
  • Distance Learning and Part-Time Student Enrollment:

    • Let DD be the event a student takes a distance learning class, and EE be the event a student is part-time.
    • Given: P(D)=0.10P(D) = 0.10, P(E)=0.40P(E) = 0.40, and P(D∣E)=0.20P(D|E) = 0.20.
    • a. Find P(D∩E)P(D \cap E): By the multiplicative rule:       P(D∩E)=P(E)⋅P(D∣E)=0.40×0.20=0.08P(D \cap E) = P(E) \cdot P(D|E) = 0.40 \times 0.20 = 0.08
    • b. Find P(E∣D)P(E|D): By Bayes' rule:       P(E∣D)=P(D∩E)P(D)=0.080.10=0.80P(E|D) = \frac{P(D \cap E)}{P(D)} = \frac{0.08}{0.10} = 0.80
    • c. Find P(D∪E)P(D \cup E): By the addition rule:       P(D∪E)=P(D)+P(E)−P(D∩E)=0.10+0.40−0.08=0.42P(D \cup E) = P(D) + P(E) - P(D \cap E) = 0.10 + 0.40 - 0.08 = 0.42
    • d. Test for independence:       P(D)⋅P(E)=0.10×0.40=0.04P(D) \cdot P(E) = 0.10 \times 0.40 = 0.04       Since P(D∩E)=0.08≠0.04P(D \cap E) = 0.08 \neq 0.04, events DD and EE are dependent.
    • e. Test for mutual exclusivity:       Since P(D∩E)=0.08≠0P(D \cap E) = 0.08 \neq 0, events DD and EE are not mutually exclusive.
  • Abstract Conditional Scenarios:

    • Scenario 1: Given P(C)=0.4P(C) = 0.4, P(D)=0.5P(D) = 0.5, P(C∣D)=0.6P(C|D) = 0.6.
      • P(C∩D)=P(D)⋅P(C∣D)=0.5×0.6=0.30P(C \cap D) = P(D) \cdot P(C|D) = 0.5 \times 0.6 = 0.30
      • Mutually exclusive? No, because P(C∩D)=0.30≠0P(C \cap D) = 0.30 \neq 0.
      • Independent? No, because P(C∣D)=0.60≠P(C)=0.40P(C|D) = 0.60 \neq P(C) = 0.40.
      • P(C∪D)=P(C)+P(D)−P(C∩D)=0.4+0.5−0.3=0.60P(C \cup D) = P(C) + P(D) - P(C \cap D) = 0.4 + 0.5 - 0.3 = 0.60
      • P(D∣C)=P(C∩D)P(C)=0.300.40=0.75P(D|C) = \frac{P(C \cap D)}{P(C)} = \frac{0.30}{0.40} = 0.75
    • Scenario 2: Events YY and ZZ are independent with P(Y∪Z)=0.71P(Y \cup Z) = 0.71 and P(Y)=0.42P(Y) = 0.42.
      • Rewritten addition rule: P(Y∪Z)=P(Y)+P(Z)−P(Y)⋅P(Z)P(Y \cup Z) = P(Y) + P(Z) - P(Y) \cdot P(Z).
      • Solving for P(Z)P(Z): 0.71=0.42+P(Z)−0.42⋅P(Z)0.71 = 0.42 + P(Z) - 0.42 \cdot P(Z)0.29=0.58⋅P(Z)  ⟹  P(Z)=0.290.58=0.500.29 = 0.58 \cdot P(Z) \implies P(Z) = \frac{0.29}{0.58} = 0.50
    • Scenario 3: Events GG and HH are mutually exclusive with P(G)=0.5P(G) = 0.5 and P(H)=0.3P(H) = 0.3.
      • The statement P(H∣G)=0.4P(H|G) = 0.4 must be false because mutually exclusive events cannot occur together (P(H∩G)=0P(H \cap G) = 0), which requires P(H∣G)=0P(G)=0P(H|G) = \frac{0}{P(G)} = 0.
      • P(H∪G)=P(H)+P(G)=0.3+0.5=0.80P(H \cup G) = P(H) + P(G) = 0.3 + 0.5 = 0.80.
      • GG and HH are dependent because knowledge that GG has occurred alters the probability that HH occurs (reducing it to 00).

Discrete Random Variables and Probability Distributions

  • Random Variables and Discrete Distribution Properties:

    • Random Variable (XX): A variable that assumes numerical values determined by the outcomes of a random experiment.
    • Discrete Random Variable (DRV): A random variable that assumes a countable number of distinct values (finite or countably infinite).
    • Continuous Random Variable (CRV): A random variable that assumes any value within one or more continuous intervals along a line.
    • Probability Distribution Function (PDF): For a DRV, a table, graph, or mathematical formula that assigns a probability P(X=x)P(X = x) to each possible value xx.
    • Properties of Discrete PDFs:0≤P(X=x)≤1for every value x0 \le P(X = x) \le 1 \quad \text{for every value } x∑P(X=x)=1\sum P(X = x) = 1
    • Expected Value (E(X)E(X)): The long-run theoretical mean of a discrete random variable:     E(X)=μ=∑xi⋅P(X=xi)E(X) = \mu = \sum x_i \cdot P(X = x_i)
    • Variance (Var(X)Var(X)): The measure of dispersion of a discrete random variable around its mean:     Var(X)=σ2=E(X2)−[E(X)]2=∑xi2⋅P(X=xi)−[E(X)]2Var(X) = \sigma^2 = E(X^2) - [E(X)]^2 = \sum x_i^2 \cdot P(X = x_i) - [E(X)]^2
  • Dice Distributions (Sum and Absolute Difference):

    • Sum Distribution (X=Die 1+Die 2X = \text{Die 1} + \text{Die 2}):
    • XX takes on integer values from 22 to 1212.
    • P(X=2)=136P(X = 2) = \frac{1}{36}
    • P(X=3)=236P(X = 3) = \frac{2}{36}
    • P(X=4)=336P(X = 4) = \frac{3}{36}
    • P(X=5)=436P(X = 5) = \frac{4}{36}
    • P(X=6)=536P(X = 6) = \frac{5}{36}
    • P(X=7)=636P(X = 7) = \frac{6}{36}
    • P(X=8)=536P(X = 8) = \frac{5}{36}
    • P(X=9)=436P(X = 9) = \frac{4}{36}
    • P(X=10)=336P(X = 10) = \frac{3}{36}
    • P(X=11)=236P(X = 11) = \frac{2}{36}
    • P(X=12)=136P(X = 12) = \frac{1}{36}
    • XX is a DRV because its values are countable and all probabilities are non-negative and sum to 11.
    • Absolute Difference Distribution (Y=∣Die 1−Die 2∣Y = |\text{Die 1} - \text{Die 2}|):
    • YY takes on values from 00 to 55.
    • Difference 00 (doubles): P(Y=0)=636P(Y = 0) = \frac{6}{36}
    • Difference 11: P(Y=1)=1036P(Y = 1) = \frac{10}{36}
    • Difference 22: P(Y=2)=836P(Y = 2) = \frac{8}{36}
    • Difference 33: P(Y=3)=636P(Y = 3) = \frac{6}{36}
    • Difference 44: P(Y=4)=436P(Y = 4) = \frac{4}{36}
    • Difference 55: P(Y=5)=236P(Y = 5) = \frac{2}{36}
    • Sum of probabilities: 6+10+8+6+4+236=3636=1\frac{6 + 10 + 8 + 6 + 4 + 2}{36} = \frac{36}{36} = 1. Therefore, YY is a DRV.
  • Unknown Parameters in Discrete Probability Tables:

    • Parameter mm in Variable WW:
    • The probability distribution for WW is:
ww0011449913132020
P(W=w)P(W = w)mm311\frac{3}{11}222\frac{2}{22}422\frac{4}{22}211\frac{2}{11}m2\frac{m}{2}
* Common denominator 2222 conversion:

      m+m2+622+222+422+422=1m + \frac{m}{2} + \frac{6}{22} + \frac{2}{22} + \frac{4}{22} + \frac{4}{22} = 132m+1622=1  ⟹  32m=1−811=311\frac{3}{2}m + \frac{16}{22} = 1 \implies \frac{3}{2}m = 1 - \frac{8}{11} = \frac{3}{11}m=311×23=211m = \frac{3}{11} \times \frac{2}{3} = \frac{2}{11}

  • Parameter kk in Variable XX:
    • The probability distribution for XX is:

| xx | 00 | 11 | 22 | −1-1 | |---|---|---|---|---|---| | P(X=x)P(X = x) | 2k2k | 5k5k | 4k4k | 9k9k |

* Sum of probabilities: 2k+5k+4k+9k=20k=1  ⟹  k=120=0.052k + 5k + 4k + 9k = 20k = 1 \implies k = \frac{1}{20} = 0.05.
* Expected value:

      E(X)=[0⋅(2k)]+[1⋅(5k)]+[2⋅(4k)]+[−1⋅(9k)]=(0+5+8−9)k=4kE(X) = [0 \cdot (2k)] + [1 \cdot (5k)] + [2 \cdot (4k)] + [-1 \cdot (9k)] = (0 + 5 + 8 - 9)k = 4kE(X)=4×0.05=0.20E(X) = 4 \times 0.05 = 0.20

  • Coin Tosses, Keys, and Exercise Models:

    • Three Fair Coins (Number of Heads XX):
    • Sample points: TTT,THT,HTT,TTH,HHT,THH,HTH,HHHTTT, THT, HTT, TTH, HHT, THH, HTH, HHH.
    • P(X=0)=18P(X = 0) = \frac{1}{8}
    • P(X=1)=38P(X = 1) = \frac{3}{8}
    • P(X=2)=38P(X = 2) = \frac{3}{8}
    • P(X=3)=18P(X = 3) = \frac{1}{8}
    • Mean:       E(X)=0(18)+1(38)+2(38)+3(18)=0+3+6+38=128=1.5E(X) = 0\left(\frac{1}{8}\right) + 1\left(\frac{3}{8}\right) + 2\left(\frac{3}{8}\right) + 3\left(\frac{1}{8}\right) = \frac{0 + 3 + 6 + 3}{8} = \frac{12}{8} = 1.5
    • Three Indistinguishable Keys:
    • A keychain has 33 indistinguishable keys, exactly one of which opens the door. Keys are tested randomly without replacement.
    • Probability unlocking on 1st try: 13\frac{1}{3}.
    • Probability unlocking on 2nd try: 23×12=13\frac{2}{3} \times \frac{1}{2} = \frac{1}{3}.
    • Probability unlocking on 3rd try: 23×12×1=13\frac{2}{3} \times \frac{1}{2} \times 1 = \frac{1}{3}.
    • Expected number of trials:       E(X)=1(13)+2(13)+3(13)=1+2+33=63=2.0 keysE(X) = 1\left(\frac{1}{3}\right) + 2\left(\frac{1}{3}\right) + 3\left(\frac{1}{3}\right) = \frac{1 + 2 + 3}{3} = \frac{6}{3} = 2.0\text{ keys}
    • Athlete Daily Exercise and Stretches:
    • An athlete exercises for 30 min30\,min, choosing to jog (JJ) or swim (SS). P(J)=0.37P(J) = 0.37, so P(S)=1−0.37=0.63P(S) = 1 - 0.37 = 0.63.
    • Stretches given jogging: P(Stretch∣J)=0.29  ⟹  P(No Stretch∣J)=0.71P(\text{Stretch}|J) = 0.29 \implies P(\text{No Stretch}|J) = 0.71.
    • Stretches given swimming: P(Stretch∣S)=0.13  ⟹  P(No Stretch∣S)=0.87P(\text{Stretch}|S) = 0.13 \implies P(\text{No Stretch}|S) = 0.87.
    • Probability the athlete does not perform post-workout stretches:       P(No Stretch)=P(J)⋅P(No Stretch∣J)+P(S)⋅P(No Stretch∣S)P(\text{No Stretch}) = P(J) \cdot P(\text{No Stretch}|J) + P(S) \cdot P(\text{No Stretch}|S)P(No Stretch)=(0.37×0.71)+(0.63×0.87)=0.2627+0.5481=0.8108P(\text{No Stretch}) = (0.37 \times 0.71) + (0.63 \times 0.87) = 0.2627 + 0.5481 = 0.8108
  • Applied Expected Value Decision Analysis:

    • Theater Raffle Fundraiser (Exercise 70):
    • A theater group sells 100100 raffle tickets at $5\$5 each. A buyer purchases 44 tickets (total investment $20\$20). Prize: Broadway show passes valued at $150\$150.
    • Variable definition: Let XX denote the buyer's net financial gain.
    • Values of XX: Win: $150−$20=$130\$150 - \$20 = \$130; Lose: −$20-\$20.
    • Probabilities: Winning probability =4100=0.04= \frac{4}{100} = 0.04; losing probability =96100=0.96= \frac{96}{100} = 0.96.
    • PDF table:
OutcomexxP(X=x)P(X = x)
Win$130\$1300.040.04
Lose−$20-\$200.960.96
* Expected average net return per raffle:

      E(X)=(130×0.04)+(−20×0.96)=5.20−19.20=−$14.00E(X) = (130 \times 0.04) + (-20 \times 0.96) = 5.20 - 19.20 = -\$14.00

  • Card and Coin Gambling Game (Exercise 71):
    • A card is selected from a standard 52-card deck and a fair coin is flipped.
    • Deck composition: 1212 face cards (Jack, Queen, King across four suits) and 4040 non-face cards.
    • Payoff structure:
      • Face card AND Heads: Win $6\$6. P=1252×12=652P = \frac{12}{52} \times \frac{1}{2} = \frac{6}{52}
      • Face card AND Tails: Win $2\$2. P=1252×12=652P = \frac{12}{52} \times \frac{1}{2} = \frac{6}{52}
      • Non-face card: Lose $2\$2 regardless of coin outcome. P=4052P = \frac{40}{52}
    • Expected value calculation:       E(X)=(6×652)+(2×652)+(−2×4052)=36+12−8052=−3252=−813≈−$0.62E(X) = \left(6 \times \frac{6}{52}\right) + \left(2 \times \frac{6}{52}\right) + \left(-2 \times \frac{40}{52}\right) = \frac{36 + 12 - 80}{52} = -\frac{32}{52} = -\frac{8}{13} \approx -\$0.62
    • Decision: The player expects to lose approximately $0.62\$0.62 per round in the long run; therefore, one should not play this game to win money.
  • Envelope Coupon Deal (Exercise 79):
    • An entry fee of $10\$10 is charged to draw one envelope from a pool of 100100.
    • Coupon distribution:
      • 1010 coupons for a gift worth $6\$6 (net: 6−10=−$46 - 10 = -\$4)
      • 8080 coupons for a gift worth $8\$8 (net: 8−10=−$28 - 10 = -\$2)
      • 66 coupons for a gift worth $12\$12 (net: 12−10=+$212 - 10 = +\$2)
      • 44 coupons for a gift worth $40\$40 (net: 40−10=+$3040 - 10 = +\$30)
    • Expected gross coupon value:       E(Gross)=(6×0.10)+(8×0.80)+(12×0.06)+(40×0.04)=0.60+6.40+0.72+1.60=$9.32E(\text{Gross}) = (6 \times 0.10) + (8 \times 0.80) + (12 \times 0.06) + (40 \times 0.04) = 0.60 + 6.40 + 0.72 + 1.60 = \$9.32
    • Expected net return:       E(Net)=9.32−10.00=−$0.68E(\text{Net}) = 9.32 - 10.00 = -\$0.68
    • Decision: Over the long run, players expect to come out behind in money (Option b).
  • GameStop Video Game Rentals (Exercise 78):
    • Distribution of game rentals per customer (5-game5\text{-game} cap):
xx001122334455
P(X=x)P(X = x)0.030.030.500.500.240.24?0.070.070.040.04
* Definition of random variable: X=number of video games rented per customer at GameStopX = \text{number of video games rented per customer at GameStop}.
* Finding P(X=3)P(X = 3): Sum of known probabilities =0.03+0.50+0.24+0.07+0.04=0.88= 0.03 + 0.50 + 0.24 + 0.07 + 0.04 = 0.88.

      P(X=3)=1.00−0.88=0.12P(X = 3) = 1.00 - 0.88 = 0.12 * Probability a customer rents at least four video games:       P(X≥4)=P(X=4)+P(X=5)=0.07+0.04=0.11P(X \ge 4) = P(X = 4) + P(X = 5) = 0.07 + 0.04 = 0.11

  • Two-Stage Box Card Drawing Experiment:
    • Box A contains four cards: (2,2,3,3)(2, 2, 3, 3). Box B contains three cards: (2,3,4)(2, 3, 4).
    • Rules: Draw two cards without replacement from Box A. If both cards display the same number, draw one additional card from Box B. Score XX is the sum of all cards drawn.
    • Primary draws from Box A ((42)=6\binom{4}{2} = 6 combinations):
      • Two 2s: (22)=1\binom{2}{2} = 1 way; probability =16= \frac{1}{6}. Sum from Box A =4= 4.
      • Two 3s: (22)=1\binom{2}{2} = 1 way; probability =16= \frac{1}{6}. Sum from Box A =6= 6.
      • One 2 and one 3: (21)(21)=4\binom{2}{1}\binom{2}{1} = 4 ways; probability =46= \frac{4}{6}. Sum from Box A =5= 5.
    • Secondary draws from Box B (each card has probability 13\frac{1}{3}):
      • Given Two 2s: Total sum can be 4+2=64 + 2 = 6, 4+3=74 + 3 = 7, or 4+4=84 + 4 = 8, each with probability 16×13=118\frac{1}{6} \times \frac{1}{3} = \frac{1}{18}.
      • Given Two 3s: Total sum can be 6+2=86 + 2 = 8, 6+3=96 + 3 = 9, or 6+4=106 + 4 = 10, each with probability 16×13=118\frac{1}{6} \times \frac{1}{3} = \frac{1}{18}.
      • Given One 2 and One 3: No extra card drawn. Sum =5= 5 with probability 46=1218\frac{4}{6} = \frac{12}{18}.
    • Complete PDF of XX:
      • P(X=5)=1218P(X = 5) = \frac{12}{18}
      • P(X=6)=118P(X = 6) = \frac{1}{18}
      • P(X=7)=118P(X = 7) = \frac{1}{18}
      • P(X=8)=118+118=218P(X = 8) = \frac{1}{18} + \frac{1}{18} = \frac{2}{18}
      • P(X=9)=118P(X = 9) = \frac{1}{18}
      • P(X=10)=118P(X = 10) = \frac{1}{18}
    • Since XX assumes a countable set of integer values and the probabilities sum to 11, XX is a DRV.
    • Expected score per game:       E(X)=(5×12)+(6×1)+(7×1)+(8×2)+(9×1)+(10×1)18E(X) = \frac{(5 \times 12) + (6 \times 1) + (7 \times 1) + (8 \times 2) + (9 \times 1) + (10 \times 1)}{18}E(X)=60+6+7+16+9+1018=10818=6E(X) = \frac{60 + 6 + 7 + 16 + 9 + 10}{18} = \frac{108}{18} = 6
    • Payoff Analysis: Cost to play is $6.50\$6.50. The player receives $1.00\$1.00 per point, yielding an expected payout of $6.00\$6.00.
    • Net return per single game: 6.00−6.50=−$0.506.00 - 6.50 = -\$0.50.
    • Expected net loss over 99 games:       Expected Net=9×(−0.50)=−$4.50\text{Expected Net} = 9 \times (-0.50) = -\$4.50

Continuous Random Variables and Probability Density Functions

  • Properties of Continuous Probability Distributions and Density Curves:
    • Measurement: Outcomes of continuous random variables are measured rather than counted.
    • Probability as Area: For continuous variables, probability corresponds to the area under the curve across an interval. The probability of an exact single point is zero: P(X=c)=0P(X = c) = 0.
    • Probability Density Function (PDF): Denoted f(x)f(x), this function defines the height of the curve at any value xx.
    • Cumulative Distribution Function (CDF): Gives the cumulative area under the density curve up to a point xx, representing P(X≤x)P(X \le x).
    • Requirements for a Valid Density Curve:
    • The curve must remain on or above the horizontal axis across its entire support: f(x)≥0f(x) \ge 0.
    • The total area between the curve and the horizontal axis across all possible values equals exactly 11.

Linear density curve example

  • Verification of a Linear Density Function:

    • Consider a linear function defined on [0,1][0, 1] rising from f(0)=0.5f(0) = 0.5 to f(1)=2.5f(1) = 2.5.

    • The graph forms a trapezoid above the horizontal axis with base b=1−0=1b = 1 - 0 = 1 and parallel heights h1=0.5h_1 = 0.5, h2=2.5h_2 = 2.5.

    • Area calculation:       Area=h1+h22×b=0.5+2.52×1=3.02×1=1.5\text{Area} = \frac{h_1 + h_2}{2} \times b = \frac{0.5 + 2.5}{2} \times 1 = \frac{3.0}{2} \times 1 = 1.5

    • Since the total area equals 1.5≠1.01.5 \neq 1.0, this graph is not a valid probability density curve.

    • The Continuous Uniform Distribution:

  • A continuous random variable XX has a uniform distribution over the interval [a,b][a, b], denoted X∼U(a,b)X \sim U(a, b), if its density function is constant:     f(x)=1b−afor a≤x≤bf(x) = \frac{1}{b - a} \quad \text{for } a \le x \le b

The uniform distribution

  • Theoretical Mean and Standard Deviation of a Uniform Distribution:μ=a+b2\mu = \frac{a + b}{2}σ=(b−a)212\sigma = \sqrt{\frac{(b - a)^2}{12}}

  • Uniform Probability as Rectangular Area:P(x1<X<x2)=(base)×(height)=(x2−x1)×1b−aP(x_1 < X < x_2) = (\text{base}) \times (\text{height}) = (x_2 - x_1) \times \frac{1}{b - a}

  • Worked Uniform Exercise (Try It 5.1):

    • Given f(x)=18f(x) = \frac{1}{8} over the interval 0≤x≤80 \le x \le 8.

    • Calculate P(2.5<x<7.5)P(2.5 < x < 7.5):       P(2.5<x<7.5)=(7.5−2.5)×18=5×18=58=0.625P(2.5 < x < 7.5) = (7.5 - 2.5) \times \frac{1}{8} = 5 \times \frac{1}{8} = \frac{5}{8} = 0.625

    • Conditional Probability in Uniform Distributions:

Conditional probability in a uniform distribution

  • Case Study: Donut Consumption Time:

    • Let XX be the time (in minutes) a 9-year-old child takes to eat a donut, where X∼U(0.5,4)X \sim U(0.5, 4).

    • Height of original PDF: f(x)=14−0.5=13.5=27f(x) = \frac{1}{4 - 0.5} = \frac{1}{3.5} = \frac{2}{7}.

    • Calculate the conditional probability P(X>2∣X>1.5)P(X > 2 | X > 1.5).

    • Method 1 (Restricted Sample Space):

      • The condition X>1.5X > 1.5 restricts the support to the new interval [1.5,4][1.5, 4].
      • New base length: 4−1.5=2.54 - 1.5 = 2.5.
      • Adjusted height: 12.5=25\frac{1}{2.5} = \frac{2}{5}.
      • Desired interval: from x=2x = 2 to x=4x = 4, with width 4−2=24 - 2 = 2.       P(X>2∣X>1.5)=(base)×(new height)=(4−2)×(25)=2×25=45=0.80P(X > 2 | X > 1.5) = (\text{base}) \times (\text{new height}) = (4 - 2) \times \left(\frac{2}{5}\right) = 2 \times \frac{2}{5} = \frac{4}{5} = 0.80
    • Method 2 (Conditional Formula):P(X>2∣X>1.5)=P(X>2∩X>1.5)P(X>1.5)=P(X>2)P(X>1.5)P(X > 2 | X > 1.5) = \frac{P(X > 2 \cap X > 1.5)}{P(X > 1.5)} = \frac{P(X > 2)}{P(X > 1.5)}P(X>2)=4−23.5=23.5P(X > 2) = \frac{4 - 2}{3.5} = \frac{2}{3.5}P(X>1.5)=4−1.53.5=2.53.5P(X > 1.5) = \frac{4 - 1.5}{3.5} = \frac{2.5}{3.5}P(X>2∣X>1.5)=23.52.53.5=22.5=45=0.80P(X > 2 | X > 1.5) = \frac{\frac{2}{3.5}}{\frac{2.5}{3.5}} = \frac{2}{2.5} = \frac{4}{5} = 0.80

    • Uniform Distribution Case Study: Vehicle Age Distribution:

  • Scenario: The age of automobiles in a college staff parking lot is uniformly distributed between six months (0.5 years0.5\text{ years}) and 9.5 years9.5\text{ years}.

  • Variable definition: Let XX denote the age of a randomly selected car in years. XX is a continuous variable.

  • Distribution notation: X∼U(0.5,9.5)X \sim U(0.5, 9.5).

  • Support interval: [0.5,9.5][0.5, 9.5].

  • Probability Density Function:     f(x)=19.5−0.5=19.0for 0.5≤x≤9.5f(x) = \frac{1}{9.5 - 0.5} = \frac{1}{9.0} \quad \text{for } 0.5 \le x \le 9.5

  • Average automobile age:     μ=a+b2=0.5+9.52=10.02=5.0 years\mu = \frac{a + b}{2} = \frac{0.5 + 9.5}{2} = \frac{10.0}{2} = 5.0\text{ years}

  • Probability a car is less than four years old:     P(X<4)=(4.0−0.5)×19.0=3.5×19.0=3.59.0=718≈0.3889P(X < 4) = (4.0 - 0.5) \times \frac{1}{9.0} = 3.5 \times \frac{1}{9.0} = \frac{3.5}{9.0} = \frac{7}{18} \approx 0.3889

  • Conditional Probability (cars less than 4 years old, given they are less than 7.5 years old):     P(X<4∣X<7.5)=P(X<4∩X<7.5)P(X<7.5)=P(X<4)P(X<7.5)P(X < 4 | X < 7.5) = \frac{P(X < 4 \cap X < 7.5)}{P(X < 7.5)} = \frac{P(X < 4)}{P(X < 7.5)}P(X<7.5)=(7.5−0.5)×19.0=7.09.0P(X < 7.5) = (7.5 - 0.5) \times \frac{1}{9.0} = \frac{7.0}{9.0}P(X<4∣X<7.5)=3.59.07.09.0=3.57.0=12=0.50P(X < 4 | X < 7.5) = \frac{\frac{3.5}{9.0}}{\frac{7.0}{9.0}} = \frac{3.5}{7.0} = \frac{1}{2} = 0.50

    • Overview of Exponential and Normal Distributions:

Exponential and standard normal density curves

  • The Exponential Distribution:
    • A continuous distribution often used to model wait times or intervals until an event occurs.
    • The density curve starts at its maximum at x=0x = 0 and declines exponentially toward zero as x→∞x \to \infty.
    • The area under the curve between boundaries x=2x = 2 and x=4x = 4 corresponds to the probability P(2<X<4)P(2 < X < 4).
  • The Standard Normal Distribution:
    • A symmetric, bell-shaped continuous probability distribution with a mean of μ=0\mu = 0 and standard deviation of σ=1\sigma = 1.
    • Values along the horizontal axis represent standard deviations (zz-scores).
    • The shaded region between z=1z = 1 and z=2z = 2 represents the probability P(1<Z<2)P(1 < Z < 2).