Statistical Data Analysis, Box Plots, and Frequency Distributions

Measures of Position and Data Interpretation

Measures of position divide ordered data into equal parts: the median splits data into two equal halves (50%50\% below, 50%50\% above), quartiles into four parts (25%25\% each), deciles into ten parts (10%10\% each), and percentiles into one hundred parts (1%1\% each). The Interquartile Range (IQR) measures the central spread of the middle 50%50\% of data between Q1Q_1 and Q3Q_3. Outliers are extreme values outside the main group that impact data interpretation, while percentile rank indicates an individual score's position relative to the entire group.

Drawing conclusions from position measures requires distinguishing between direct and indirect proportion variables. For direct proportion variables—such as speed, test scores, income, height, strength, production output, bandwidth, grades, power, and literacy rate—higher values indicate better performance, making higher percentiles preferable. For indirect proportion variables—such as race completion time, reaction time, error rate, defect count, pollution level, blood pressure, cholesterol level, energy consumption, noise level, and cost—lower values represent superior performance, making lower percentiles preferable.

Box-and-Whisker Plots

A box-and-whisker plot visually summarizes a dataset using the Five-Number Summary (FNS), which consists of the minimum value, lower quartile (Q1Q_1 / 25th percentile), median (Q2Q_2 / 50th percentile), upper quartile (Q3Q_3 / 75th percentile), and maximum value. To construct a box plot, ordered data is used to draw a box extending from Q1Q_1 to Q3Q_3 with a vertical line marking Q2Q_2, while whiskers extend from the box to the minimum and maximum values.

Interpreting a box plot involves analyzing the median location, box size, and whisker lengths. A median near the center indicates a symmetric distribution, whereas a median closer to Q1Q_1 signifies positive skewness (right skew) and a median closer to Q3Q_3 signifies negative skewness (left skew). Box size reflects the IQR: a larger box shows greater variability among the middle 50%50\% of values, while a smaller box indicates tight clustering. Long or unequal whiskers highlight variation outside the quartiles and indicate the direction of the distribution's tail.

Frequency Distribution Table Components

Grouping raw data into class intervals facilitates analysis for large datasets through a frequency distribution table. The primary components include class intervals defined by lower and upper limits, with width calculated as Class Width=(Upper Limit−Lower Limit)+1\text{Class Width} = (\text{Upper Limit} - \text{Lower Limit}) + 1. Class boundaries form seamless continuous ranges where Lower Boundary=Lower Limit−0.5\text{Lower Boundary} = \text{Lower Limit} - 0.5 and Upper Boundary=Upper Limit+0.5\text{Upper Boundary} = \text{Upper Limit} + 0.5.

The class mark serves as the midpoint representing the interval, computed using Class Mark=Lower Limit+Upper Limit2\text{Class Mark} = \frac{\text{Lower Limit} + \text{Upper Limit}}{2}. Frequency counts the observations within an interval, while cumulative frequency calculates the running total of frequencies up to a specified class (Cfn=Frequency1+Frequency2+⋯+FrequencynCf_n = \text{Frequency}_1 + \text{Frequency}_2 + \dots + \text{Frequency}_n), maintaining a non-decreasing value useful for locating position measures.

Histograms, Polygons, and Ogives

A histogram displays class interval frequencies as vertical bars plotted with class boundaries on the x-axis and frequencies on the y-axis. A cumulative frequency histogram plots class boundaries on the x-axis against progressive cumulative frequencies on the y-axis, resulting in bar heights that always rise or remain flat.

A frequency polygon represents data shape as a line graph by plotting frequency values against class marks on the x-axis, anchored to the x-axis at both ends before the first interval and after the last interval. A cumulative frequency polygon (ogive) tracks cumulative trends across boundaries. A Less Than Ogive plots cumulative frequencies against upper class boundaries as a rising curve to determine medians, quartiles, and percentiles. A Greater Than Ogive plots cumulative frequencies against lower class boundaries as a falling curve to evaluate values exceeding specific thresholds.