14.3 Descriptive Statistics, Data Displays, and Probability Theory

14.3 Descriptive Statistics, Data Displays, and Probability Theory

CSET Focus: Statistics and probability on CSET Multiple Subjects Subtest II test your ability to calculate and interpret summary statistics, evaluate how skewness affects mean versus median, apply the $1.5 \times \text{IQR}$ outlier rule, interpret graphical data displays (including boxplots and histograms), solve single and compound probability problems with and without replacement, and apply counting principles (permutations vs. combinations) to classroom scenarios.

---

1. Descriptive Statistics: Central Tendency, Spread, and Distribution Shape

Descriptive statistics synthesize raw numerical datasets into meaningful quantitative summaries characterizing central location, spread, and distributional symmetry.

                                  DESCRIPTIVE STATISTICS
                ┌───────────────────────────┴───────────────────────────┐
       MEASURES OF CENTRAL TENDENCY                            MEASURES OF DISPERSION (SPREAD)
   • Mean (x̄): Arithmetic average                          • Range: Max - Min (Sensitive to outliers)
   • Median (Q₂): Middle value (Resistant)                • IQR: Q₃ - Q₁ (Middle 50% spread, Resistant)
   • Mode: Most frequent value (Categorical)               • Variance (s²): Mean squared deviation
   • Weighted Mean: Value weighting by frequency           • Standard Deviation (s): Root mean squared spread

Measures of Central Tendency

1. Arithmetic Mean ($\bar{x}$ or $\mu$): The sum of all observations divided by the total number of observations: $$\bar{x} = \frac{\sum_{i=1}^n x_i}{n}$$
   • Property: Highly sensitive to extreme values and outliers.
2. Weighted Mean ($\bar{x}_w$): Computes an average where each value is weighted by its relative importance or frequency $w_i$: $$\bar{x}_w = \frac{\sum_{i=1}^k w_i x_i}{\sum_{i=1}^k w_i}$$
   • Example: Course grade calculation with $3$ tests at $20\%$ each ($60\%$ total) and a final exam at $40\%$.
3. Median ($Q_2$ / $\tilde{x}$): The physical middle value in an ordered (sorted) dataset:
   • If $n$ is odd, the median is the unique value at position $\frac{n+1}{2}$.
   • If $n$ is even, the median is the arithmetic mean of the two middle values at positions $\frac{n}{2}$ and $\frac{n}{2} + 1$.
   • Property: Resistant (robust) measure of center; not influenced by extreme outliers.
4. Mode: The data value(s) occurring with the highest frequency. A distribution may be unimodal (one mode), bimodal (two modes), multimodal, or have no mode (if all values occur equally). The mode is the only measure of central tendency appropriate for qualitative/nominal data.

Skewness and the Mean-Median Relationship

   Left-Skewed (Negatively Skewed)           Symmetric (Bell-Shaped)          Right-Skewed (Positively Skewed)
            Long Left Tail                                                            Long Right Tail
              ┌───┐                                   ┌───┐                                 ┌───┐
           ┌──┘   └──┐                             ┌──┘   └──┐                            ┌──┘   └──┐
       ────┘         └────                     ────┘         └────                    ────┘         └────
     Mean < Median < Mode                      Mean ≈ Median ≈ Mode                     Mode < Median < Mean

• Symmetric Distribution: The data are balanced around a central peak. $\text{Mean} \approx \text{Median} \approx \text{Mode}$.
• Right-Skewed (Positively Skewed): Tail extends toward higher values to the right. Extreme high values pull the mean upward: $\mathbf{\text{Mean} > \text{Median} > \text{Mode}}$. (Example: Household income, home sale prices). Use Median and IQR as the preferred summary statistics.
• Left-Skewed (Negatively Skewed): Tail extends toward lower values to the left. Extreme low values pull the mean downward: $\mathbf{\text{Mean} < \text{Median} < \text{Mode}}$. (Example: Age at natural retirement, easy exam scores). Use Median and IQR.

Measures of Dispersion and the Five-Number Summary

1. Range: Difference between the maximum and minimum observations: $\text{Range} = x_{\max} - x_{\min}$.
2. Quartiles and Interquartile Range (IQR):
   • First Quartile ($Q_1$): 25th percentile (median of the lower half of data).
   • Second Quartile ($Q_2$): 50th percentile (median of the entire dataset).
   • Third Quartile ($Q_3$): 75th percentile (median of the upper half of data).
   • Interquartile Range (IQR): The spread of the middle $50\%$ of data: $\mathbf{\text{IQR} = Q_3 - Q_1}$. (Resistant to outliers).
3. The Five-Number Summary: Comprises $[\text{Minimum}, \; Q_1, \; Q_2, \; Q_3, \; \text{Maximum}]$.

Tukey's Outlier Detection Rule ($1.5 \times \text{IQR}$ Criterion)

A data observation $x$ is formally classified as an outlier if it lies outside the inner fences:

$$\text{Lower Fence} = Q_1 - 1.5 \times \text{IQR} \qquad \text{Upper Fence} = Q_3 + 1.5 \times \text{IQR}$$

• Outlier Criterion: $x < \text{Lower Fence}$ OR $x > \text{Upper Fence}$.

Standard Deviation & The Empirical Rule (68-95-99.7%)

• Sample Variance ($s^2$) and Standard Deviation ($s$): $$s^2 = \frac{\sum_{i=1}^n (x_i - \bar{x})^2}{n - 1} \qquad s = \sqrt{\frac{\sum_{i=1}^n (x_i - \bar{x})^2}{n - 1}}$$
• The Empirical Rule (68-95-99.7% Rule for Normal Distributions):
  • Approximately $68.27\%$ of data fall within $\mu \pm 1\sigma$.
  • Approximately $95.45\%$ of data fall within $\mu \pm 2\sigma$.
  • Approximately $99.73\%$ of data fall within $\mu \pm 3\sigma$.
• Standard Score ($z$-score): Measures the signed distance from the mean in standard deviation units: $z = \frac{x - \mu}{\sigma}$.

---

2. Graphical Data Displays and Bivariate Analysis

                                   DATA DISPLAYS
                ┌────────────────────────┴────────────────────────┐
       CATEGORICAL DISPLAYS                              QUANTITATIVE DISPLAYS
   • Bar Charts (Spaces between bars)              • Dot Plots / Line Plots (Discrete points on line)
   • Pie Charts / Circle Graphs (Angle = % × 360°)  • Stem-and-Leaf Plots (Preserves exact raw values)
                                                   • Histograms (Continuous contiguous bins, no gaps)
                                                   • Box-and-Whisker Plots (5-number summary & outliers)
                                                   • Scatter Plots (Bivariate correlation & trend lines)

Graphical Data Displays Matrix

Bivariate Data: Scatter Plots and Correlation vs. Causation

• Scatter Plots: Display paired quantitative data $(x_i, y_i)$ to investigate associative relationships between an independent explanatory variable ($x$) and a dependent response variable ($y$).
• Linear Correlation Coefficient ($r$): Measures the strength and direction of a linear relationship ($-1 \le r \le 1$):
  • $r = +1$: Perfect positive linear correlation.
  • $r = -1$: Perfect negative linear correlation.
  • $r = 0$: No linear correlation.
• Trend Line / Line of Best Fit ($y = mx + b$): Used for interpolation (estimating within the observed data domain) and cautious extrapolation (predicting outside the observed domain).
• Correlation vs. Causation: A strong statistical correlation between two variables does NOT imply a cause-and-effect relationship. Apparent correlations are frequently driven by confounding or lurking variables (e.g., shoe size and reading ability correlate strongly in elementary school children due to the lurking variable of age).

---

3. Probability Theory and Combinatorics

Probability quantifies the likelihood of an event occurring within a well-defined sample space ($S$).

Foundations of Probability

• Theoretical Probability: Based on mathematical analysis of equally likely outcomes in sample space $S$: $$P(E) = \frac{n(E)}{n(S)} = \frac{\text{Number of favorable outcomes}}{\text{Total number of possible outcomes}}$$
• Experimental (Empirical) Probability: Based on observed relative frequency from empirical trials: $$P_{\text{exp}}(E) = \frac{\text{Frequency of event occurrence}}{\text{Total number of experimental trials}}$$
• The Law of Large Numbers: As the number of independent repetitions ($N$) increases, the empirical experimental probability converges toward the theoretical probability: $\lim_{N \to \infty} P_{\text{exp}}(E) = P(E)$.

Fundamental Probability Axioms & Operational Rules

1. Probability Bounds: $0 \le P(E) \le 1$. Impossible event: $P(\emptyset) = 0$; Certain event: $P(S) = 1$.
2. Complement Rule: The probability that event $E$ does not occur ($E'$ or $E^c$): $$P(E') = 1 - P(E)$$
   • "At Least One" Strategy: To find the probability of at least one success across multiple trials: $$P(\text{At least one}) = 1 - P(\text{None})$$
3. Addition Rule (Union of Events):
   • Mutually Exclusive (Disjoint) Events ($A \cap B = \emptyset$): $$P(A \cup B) = P(A) + P(B)$$
   • Non-Mutually Exclusive (Inclusive) Events ($A \cap B \ne \emptyset$): $$P(A \cup B) = P(A) + P(B) - P(A \cap B)$$
4. Conditional Probability & Multiplication Rule (Intersection of Events):
   • Conditional Probability ($P(B|A)$): The probability of $B$ occurring given that $A$ has already occurred: $$P(B|A) = \frac{P(A \cap B)}{P(A)} \quad (P(A) > 0)$$
   • General Multiplication Rule: $$P(A \cap B) = P(A) \cdot P(B|A)$$
   • Independent Events: Two events are independent if $P(B|A) = P(B)$. Then: $$P(A \cap B) = P(A) \cdot P(B)$$
   • Dependent Events (Sampling Without Replacement): Probabilities change on each successive draw because the sample space decreases.

Combinatorics: Counting Principles, Permutations, and Combinations

                                  COUNTING TECHNIQUES
                ┌──────────────────────────┴──────────────────────────┐
        ORDER MATTERS? YES                                    ORDER MATTERS? NO
           PERMUTATIONS (nPr)                                 COMBINATIONS (nCr)
   • Arranging officers (Pres, VP, Sec)               • Choosing committees, teams, hands
   • Seating in a line, race finish order             • Selecting toppings, fruit salads
   • nPr = n! / (n - r)!                              • nCr = n! / [r!(n - r)!]

1. Fundamental Counting Principle: If a multi-step task consists of $k$ independent successive stages with $n_1, n_2, \dots, n_k$ choices, total outcomes $= n_1 \times n_2 \times \dots \times n_k$.
2. Factorial Notation ($n!$): For $n \in \mathbb{Z}^+$, $n! = n(n-1)(n-2)\cdots(1)$, with the formal convention $0! = 1$.
3. Permutations ($nPr$ — Order Matters): The number of ordered arrangements of $r$ items selected from a set of $n$ distinct items: $$_n P_r = \frac{n!}{(n - r)!}$$
   • Example: Choosing 1st, 2nd, and 3rd place winners from 8 runners: $$_8 P_3 = \frac{8!}{(8-3)!} = \frac{8!}{5!} = 8 \times 7 \times 6 = 336$$
4. Combinations ($nCr$ or $\binom{n}{r}$ — Order Does NOT Matter): The number of unordered subsets of $r$ items chosen from a set of $n$ distinct items: $$_n C_r = \binom{n}{r} = \frac{n!}{r!(n - r)!} = \frac{_n P_r}{r!}$$
   • Example: Selecting a committee of 3 teachers from a department of 8 teachers: $$_8 C_3 = \frac{8!}{3!(8-3)!} = \frac{8 \times 7 \times 6}{3 \times 2 \times 1} = 56$$

Probability & Combinatorics Reference Matrix