6.3 Descriptive Statistics, Data Interpretation & Probability

Descriptive Statistics, Data Interpretation & Probability

Statistical reasoning, data interpretation, and probability represent essential competencies tested on the FTCE General Knowledge Mathematics Subtest. Educators must understand how to summarize datasets using measures of central tendency and spread, identify outliers, interpret graphical displays (histograms, stem-and-leaf plots, box plots, scatter plots), evaluate normal distributions, calculate compound probabilities, and apply combinatorial counting principles.

---

1. Measures of Central Tendency & Distribution Skewness

Measures of central tendency identify a single central value that represents the typical score in a numerical distribution.

Definitions of Central Tendency Measures

• Arithmetic Mean ($\bar{x}$): The sum of all observations divided by the total number of observations ($n$): $\bar{x} = \frac{\sum x_i}{n}$ Sensitivity: Non-resistant (highly sensitive) to extreme values and outliers. High outliers pull the mean upward; low outliers pull the mean downward.
• Median ($M$ or $Q_2$): The middle value when data observations are arranged in ascending order.
  • If $n$ is odd, the median is the exact middle observation at position $\frac{n + 1}{2}$.
  • If $n$ is even, the median is the arithmetic mean of the two middle observations at positions $\frac{n}{2}$ and $\frac{n}{2} + 1$. Sensitivity: Resistant (robust) to extreme values and outliers. It represents the preferred measure of center for skewed distributions (such as household incomes or house prices).
• Mode: The value that occurs with the highest frequency in the dataset. A dataset may have no mode, one mode (unimodal), two modes (bimodal), or multiple modes.

Distribution Skewness and Shape Relationships

+-------------------------------------------------------------------------+
|                  DISTRIBUTION SHAPE & SKEWNESS MATRIX                   |
+---------------------+-------------------------+-------------------------+
| Distribution Shape  | Mean vs. Median vs Mode | Visual Characteristics  |
+---------------------+-------------------------+-------------------------+
| Symmetrical         | Mean = Median = Mode    | Bell-shaped / balanced  |
| (Normal)            |                         | identical tails         |
+---------------------+-------------------------+-------------------------+
| Positively Skewed   | Mode < Median < Mean    | Long tail extends to    |
| (Right-Skewed)      | (Mean pulled right/high)| the right (high values) |
+---------------------+-------------------------+-------------------------+
| Negatively Skewed   | Mean < Median < Mode    | Long tail extends to    |
| (Left-Skewed)       | (Mean pulled left/low)  | the left (low values)   |
+---------------------+-------------------------+-------------------------+```

---

## 2. Measures of Dispersion & Outlier Detection

**Measures of dispersion (spread)** quantify the degree of variability or scatter in a dataset.

### Core Dispersion Measures
- **Range:** The difference between the maximum and minimum observations:
  $$\text{Range} = \text{Maximum} - \text{Minimum}$$
- **Interquartile Range (IQR):** The spread of the middle $50\%$ of the data:
  $$\text{IQR} = Q_3 - Q_1$$
  *Sensitivity:* Highly resistant to outliers.
- **Standard Deviation ($\sigma$ or $s$):** The typical distance between data points and the arithmetic mean:
  $$s = \sqrt{\frac{\sum (x_i - \bar{x})^2}{n - 1}}$$

### The Five-Number Summary & Box-and-Whisker Plots
The **five-number summary** partitions an ordered dataset into four equal quartiles, each containing exactly $25\%$ of the observations:
$$\{\text{Minimum}, Q_1, \text{Median } (Q_2), Q_3, \text{Maximum}\}$$

               |--------- BOX ---------|

|-----------|---| |---|-----------| Min Q1 Median (Q2) Q3 Max |<-- 25% -->|<-------- 50% (IQR) ---------->|<-- 25% -->|

### Tukey's 1.5 $\times$ IQR Outlier Detection Rule
An observation is formally classified as a **statistical outlier** if it lies beyond either boundary fence:
$$\begin{aligned}
\text{Lower Outlier Fence} &= Q_1 - 1.5 \times \text{IQR} \\
\text{Upper Outlier Fence} &= Q_3 + 1.5 \times \text{IQR}
\end{aligned}$$
- Any data point $x < \text{Lower Fence}$ is a low outlier.
- Any data point $x > \text{Upper Fence}$ is a high outlier.

---

## 3. The Normal Distribution & The Empirical Rule

A **normal distribution** is a continuous, symmetrical, bell-shaped probability distribution defined completely by its mean ($\mu$) and standard deviation ($\sigma$).

### The 68–95–99.7 Empirical Rule
For any normally distributed dataset:
- Approximately **$68.26\%$** of all observations lie within $1$ standard deviation of the mean: $[\mu - \sigma, \mu + \sigma]$. (Exactly $34.13\%$ on each side).
- Approximately **$95.44\%$** of all observations lie within $2$ standard deviations of the mean: $[\mu - 2\sigma, \mu + 2\sigma]$. (Adds $13.59\%$ in each second-sigma interval).
- Approximately **$99.74\%$** of all observations lie within $3$ standard deviations of the mean: $[\mu - 3\sigma, \mu + 3\sigma]$. (Adds $2.15\%$ in each third-sigma interval).
- Only **$0.26\%$** (roughly $0.13\%$ in each extreme tail) lie beyond $3$ standard deviations ($|z| > 3$).

                  Normal Distribution (Empirical Rule)
                               Mean
                                |
                              .---.
                             /  |                                  / 34%|34%                              .'    |    `.
                       .-' 13.5%|13.5% `-.
                    .-' 2.1%    |    2.1% `-.
               .---' 0.13%      |      0.13% `---.
        -------|--------|-------|-------|--------|-------
              -3σ      -2σ     -1σ      +1σ     +2σ     +3σ
               |<------- 68% ---------->|
               |<--------------- 95% ----------->|
               |<----------------------- 99.7% --------->|

---

## 4. Graphical Displays & Bivariate Scatter Plots

### Common Statistical Displays
- **Histograms:** Display continuous numerical data grouped into equal class intervals (bins). The bar height represents frequency; there are no gaps between bars unless a bin has a frequency of zero.
- **Stem-and-Leaf Plots:** Retain individual data values while displaying distribution shape. Stems represent leading digits; leaves represent final digits in ascending order.
- **Circle Graphs (Pie Charts):** Display categorical proportions of a whole ($100\% = 360^\circ$). Sector angle $= (\text{percentage} / 100) \times 360^\circ$.

### Bivariate Data & Scatter Plots
Scatter plots display relationships between two quantitative variables ($x = \text{independent}$, $y = \text{dependent}$).
- **Positive Correlation:** As $x$ increases, $y$ tends to increase ($m > 0$).
- **Negative Correlation:** As $x$ increases, $y$ tends to decrease ($m < 0$).
- **No Correlation:** Points are randomly scattered ($r \approx 0$).
- **Line of Best Fit (Linear Regression):**
  $$\hat{y} = mx + b$$
  - Slope $m$: Predicted rate of change in $y$ for each 1-unit increase in $x$.
  - Intercept $b$: Predicted value of $y$ when $x = 0$.
  - **Interpolation:** Predicting values within the range of observed data (reliable).
  - **Extrapolation:** Predicting values outside the range of observed data (unreliable).
  - **Causation Fallacy:** Correlation does NOT imply causation (confounding variables may drive both).

---

## 5. Principles of Probability

The **probability** of an event $E$, denoted $P(E)$, measures its likelihood:
$$0 \le P(E) \le 1$$
- **Theoretical Probability:** $P(E) = \frac{\text{Number of Favorable Outcomes}}{\text{Total Number of Equally Likely Outcomes in Sample Space } S}$.
- **Experimental Probability:** $P(E) = \frac{\text{Number of Times Event Occurred}}{\text{Total Number of Experimental Trials}}$.
- **Complement Rule:** The probability that event $A$ does not occur:
  $$P(A') = 1 - P(A)$$

### Compound Probability Rules

#### 1. The Addition Rule ("OR" Events)
Calculates the probability that event $A$ OR event $B$ occurs:
$$P(A \cup B) = P(A) + P(B) - P(A \cap B)$$
- If events $A$ and $B$ are **mutually exclusive (disjoint)**, they cannot occur simultaneously ($P(A \cap B) = 0$), simplifying to:
  $$P(A \text{ or } B) = P(A) + P(B)$$

#### 2. The Multiplication Rule ("AND" Events)
Calculates the probability that event $A$ AND event $B$ both occur:
- **Independent Events:** Occurrence of $A$ has no effect on the probability of $B$:
  $$P(A \text{ and } B) = P(A) \times P(B)$$
- **Dependent Events (Without Replacement):** Occurrence of $A$ changes the sample space for $B$:
  $$P(A \text{ and } B) = P(A) \times P(B|A)$$
  *(where $P(B|A)$ is the conditional probability of $B$ given that $A$ has occurred)*.

---

## 6. Combinatorics: Counting Principles, Permutations & Combinations

### The Fundamental Counting Principle
If a multi-stage task consists of $k$ independent sequential choices where choice 1 has $n_1$ options, choice 2 has $n_2$ options, ..., and choice $k$ has $n_k$ options, the total number of distinct outcomes is:
$$\text{Total Outcomes} = n_1 \times n_2 \times n_3 \times \dots \times n_k$$

### Permutations vs. Combinations

+-------------------------------------------------------------------------+ | PERMUTATIONS VS. COMBINATIONS TAXONOMY | +-------------------+----------------------------+------------------------+ | Feature | Permutations (nPr) | Combinations (nCr) | +-------------------+----------------------------+------------------------+ | Does Order Matter?| YES (Order Matters) | NO (Order Unimportant)| +-------------------+----------------------------+------------------------+ | Keywords | Arrange, rank, sequence, | Choose, select, group, | | | lineup, officers (Pres/VP),| committee, hand of | | | codes, passwords | cards, team roster | +-------------------+----------------------------+------------------------+ | Formula | nPr = n! / (n - r)! | nCr = n! / (r!(n-r)!) | +-------------------+----------------------------+------------------------+ | Example (4 items, | 4P2 = 4! / 2! = 12 | 4C2 = 4! / (2!2!) = 6 | | pick 2: {A,B,C,D})| (AB != BA) | (AB = BA) | +-------------------+----------------------------+------------------------+```

---

Realistic FTCE Mathematical Worked Examples

Worked Example 1: Outlier Detection via Tukey's Rule

Problem: A dataset of 12 test scores is given: $52, 60, 64, 68, 70, 72, 76, 78, 82, 84, 88, 98$. Determine whether any scores are statistical outliers.

Step-by-Step Solution:

1. Find the Median ($Q_2$): $n = 12$ is even; average the 6th and 7th scores: $Q_2 = \frac{72 + 76}{2} = 74$
2. Find $Q_1$ (median of lower 6 scores: $52, 60, 64, 68, 70, 72$): $Q_1 = \frac{64 + 68}{2} = 66$
3. Find $Q_3$ (median of upper 6 scores: $76, 78, 82, 84, 88, 98$): $Q_3 = \frac{82 + 84}{2} = 83$
4. Compute the Interquartile Range: $\text{IQR} = Q_3 - Q_1 = 83 - 66 = 17$
5. Compute Outlier Boundary Fences: \text{Lower Fence} &= Q_1 - 1.5(\text{IQR}) = 66 - 1.5(17) = 66 - 25.5 = 40.5 \\ \text{Upper Fence} &= Q_3 + 1.5(\text{IQR}) = 83 + 1.5(17) = 83 + 25.5 = 108.5 \end{aligned}$$
6. Evaluate Observations:
   • Minimum score is $52 > 40.5$ (no low outliers).
   • Maximum score is $98 < 108.5$ (no high outliers).
7. Conclusion: There are no statistical outliers in this dataset.

Worked Example 2: Combinatorial Committee Selection with Subgroups

Problem: A department has 6 math teachers and 8 science teachers. A district committee of 5 teachers is to be formed consisting of exactly 2 math teachers and 3 science teachers. How many unique committees can be formed?

Step-by-Step Solution:

1. Identify selection type: Committee members share equal status $\implies$ Order does not matter (Combinations).
2. Choose 2 math teachers from 6: ${}_6C_2 = \frac{6!}{2!(6 - 2)!} = \frac{6 \times 5}{2 \times 1} = 15$
3. Choose 3 science teachers from 8: ${}_8C_3 = \frac{8!}{3!(8 - 3)!} = \frac{8 \times 7 \times 6}{3 \times 2 \times 1} = 56$
4. Apply the Fundamental Counting Principle: $\text{Total Committees} = {}_6C_2 \times {}_8C_3 = 15 \times 56 = 840$
5. Conclusion: Exactly $840$ unique committee configurations can be chosen.