AP Statistics Unit 1 Summary: Exploring One Variable Data

Foundations of Data Analysis: Individuals, Variables, and Scales

Statistics vs. Parameters: * Statistic: Any summary information learned from a sample. Memorization tip: "Sample" and "Statistic" both start with the letter "S." * Parameter: Any information collected from an entire population. Memorization tip: "Population" and "Parameter" both start with the letter "P."
Individuals and Variables: * Individuals: The objects described by a set of data. These can be people, animals (e.g., lemurs), objects (e.g., chairs), or even abstract concepts like days or states. * Variable: Any characteristic that can change from one individual to another (e.g., eye color, height, weight).
Types of Variables: * Categorical Variables: These take on values that are category names or group labels (e.g., eye color, hair color). They are typically words. * Exceptions: Zip codes are numbers but are technically categorical because they serve as group labels for post offices rather than measured values. * Quantitative Variables: These take on numerical values that are counted or measured (e.g., weight of a frog, number of candies in a bag). * Discrete Quantitative: Countable and finite values, typically whole numbers (e.g., goals scored in a soccer game: $0, 1, 2, \dots$ ). * Continuous Quantitative: Values that are not countable and can be theoretically infinite due to precision of measurement (e.g., height or weight, where values can exist between any two points like $5.15234\,lb$ ).

Analyzing Categorical Data

Frequency and Relative Frequency: * Frequency: A fancy word for "counts." A frequency table lists categories and how many individuals fall into each. * Relative Frequency: The proportion (or percentage/rate) of the total. Formula: $\text{Relative Frequency} = \frac{\text{Count in Category}}{\text{Total Sample Size}}$ . * Proportions: Preferred for comparing samples of different sizes to ensure a fair comparison.
Graphical Representations: * Pie Charts (Circle Graphs): Each slice represents a proportion of the whole circle. Useful for showing parts of a whole. * Bar Graphs: Bars represent counts or proportions for each category. * Relative Bar Graphs: The height of the bars represents the proportion or percentage rather than the raw count.
Distribution of Categorical Data: Describes which values the data takes and how often. Analysis is usually limited to identifying the most/least common categories or comparing proportions across groups (e.g., comparing lemur types in Forest 1 vs. Forest 2).

Analyzing Quantitative Data: Tables and Graphs

Frequency Tables for Numbers: Requires the creation of Bins (or intervals). * Bins must be equal in size. * Left-handed Bins: A common convention where a value equal to the left boundary is included in that bin, but a value equal to the right boundary moves to the next bin (e.g., a bin of $20\text{ to }30$ includes $20$ up to $29.999\dots$ ).
Types of Quantitative Graphs: * Stem-and-Leaf Plot: Allows the viewer to see every individual data value while showing the shape of the distribution. * Dot Plot: Places dots for each individual value along a number line. * Histogram: The preferred graph for quantitative data. Groups data into bins. * Crucial Distinction: A histogram is NOT a bar graph. Bar graphs are for categorical data; histograms are for quantitative data. * Cumulative Graph (Ogive): * X-axis represents the variable value; Y-axis represents the proportion of data at or below that value. * A dot at $(80\,ft, 0.45)$ means $45\%$ of trees are below $80\,ft$ . * Slope: Steeper slopes indicate a higher density of data in that range. Horizontal lines indicate no data in that bin.

Describing Distributions of Quantitative Variables (SCSO)

Shape: * Unimodal: One clear peak. * Bimodal: Two clear peaks (may indicate two distinct clusters in the data). * Symmetric: The left and right sides are roughly mirror images. * Uniform: Data is spread evenly across the range (a flat shape). * Skewed Right: The tail extends toward higher values (the right). * Skewed Left: The tail extends toward lower values (the left). * Gaps and Clusters: Notable spaces with no data or groups where data is bunched together.
Center: One value that summarizes the "middle" of the data (Mean or Median).
Spread: Analysis of how the data varies (Range, IQR, or Standard Deviation).
Outliers: Data values far away from the rest of the distribution.

Measures of Center: Mean and Median

Mean ( $\bar{x}$ ): Typically the average. Found by adding all values and dividing by the count ( $n$ ). * Sensitivity: The mean is non-resistant; it is easily influenced by outliers and skewness. It acts as a balance point.
Median ( $M$ ): The physical middle value when data is ordered from least to greatest. * Resistance: The median is resistant; it is not affected by extreme outliers. * Location Formula: $\frac{n + 1}{2}$ . This tells you the position, not the value. If $n = 20$ , the position is $10.5$ , meaning you average the $10$ th and $11$ th values.
Relationship to Shape: * Symmetric: $\text{Mean} \approx \text{Median}$ . * Skewed Left: $\text{Mean} < \text{Median}$ (Mean is pulled down by the tail). * Skewed Right: $\text{Mean} > \text{Median}$ (Mean is pulled up by the tail).

Measures of Position: Percentiles and Quartiles

Percentile: The percentage of data at or below a score. The $95$ th percentile means $95\%$ of scores are at or below yours.
First Quartile ( $Q1$ ): The $25$ th percentile. The middle of the bottom half of data.
Median ( $Q2$ ): The $50$ th percentile.
Third Quartile ( $Q3$ ): The $75$ th percentile. The middle of the top half of data.

Measures of Spread

Range: $\text{Maximum} - \text{Minimum}$ . Highly sensitive to outliers.
Interquartile Range (IQR): $Q3 - Q1$ . Represents the spread of the middle $50\%$ of the data. Resistant to outliers.
Standard Deviation ( $\sigma$ or $s$ ): The typical distance of data values from the mean. * Small Standard Deviation: Data is clustered closely around the mean. * Large Standard Deviation: Data is spread far from the mean.

Identifying Outliers

The Fence Method: * IQR = $Q3 - Q1$ . * Upper Fence = $Q3 + 1.5 \times IQR$ . Any value above this is an outlier. * Lower Fence = $Q1 - 1.5 \times IQR$ . Any value below this is an outlier.
The Standard Deviation Method: Any value more than $2$ standard deviations above or below the mean ( $\mu \pm 2\sigma$ ) is considered an outlier.

Transforming Data

Adding/Subtracting a Constant: Affects measures of center (Mean, Median) and position (Percentiles, Quartiles), but does not affect measures of spread (Range, IQR, Standard Deviation).
Multiplying/Dividing by a Constant: Affects everything (Center, Position, and Spread).
Adding/Removing Data Points: * Adding an extreme outlier to the right will significantly increase the Mean but will barely affect the Median. * Adding a value equal to the center will likely decrease the measures of spread.

Five-Number Summary and Box Plots

Five-Number Summary: Consists of the Minimum, $Q1$ , Median, $Q3$ , and Maximum.
Box Plot Construction: * A box is drawn from $Q1$ to $Q3$ with a line for the Median. * Modified Box Plot: Outliers are identified using the Fence Method and marked with asterisks. Whiskers extend only to the highest/lowest values that are not outliers.
Interpretation: Each section of the box plot (each whisker and each half of the box) contains exactly $25\%$ of the data. A wider section indicates the data is more spread out, not that there is more data.

Comparison of Distributions

When comparing two distributions, use comparative language (e.g., greater than, less than, similar to).
Always compare four components in context: 1. Centers: Compare medians or means. 2. Shapes: Compare skewness and modality. 3. Spreads: Compare IQRs or standard deviations. 4. Outliers: Note the presence or absence of unusual values in both sets.

The Normal Distribution and Z-Scores

Density Curves: Models used to describe a population. The area under any density curve is exactly $1$ ( $100\%$ ).
Normal Distribution: A specific density curve that is unimodal, mound-shaped, and symmetric. * Defined by parameters: Mean ( $\mu$ ) and Standard Deviation ( $\sigma$ ).
The Empirical Rule ( $68-95-99.7$ Rule): * $68\%$ of data falls within $1\sigma$ of $\mu$ . * $95\%$ of data falls within $2\sigma$ of $\mu$ . * $99.7\%$ of data falls within $3\sigma$ of $\mu$ .
Standardized Score (Z-score): Measures how many standard deviations a value is above or below the mean. * Formula: $z = \frac{x - \mu}{\sigma}$ * Standard Normal Model: A normal distribution with $\mu = 0$ and $\sigma = 1$ . * Z-scores allow for comparison across different units (e.g., comparing a tree's height to a bear's weight).

Normal Distribution Calculations

Finding Proportions: * Using the Calculator (NormalCDF): Requires lower bound, upper bound, $\mu$ , and $\sigma$ . Use $-99$ or $99$ for infinity approximations. * Using Z-Tables: Look up the Z-score to find the area to the left (below). For area above, calculate $1 - \text{Table Value}$ .
Working Backward (Finding Values): * Given an area or percentile, find the Z-score and solve for $x$ using the formula: $x = \mu + z(\sigma)$ . * Using the Calculator (InvNorm): Requires the area to the left. * Example: To find the $80$ th percentile, input an area of $0.8$ . The resulting Z-score ( $z \approx 0.842$ ) is then used to find the specific value.