A histogram groups numerical values into continuous intervals. Use actual boundaries on the horizontal axis. Rectangles for adjacent occupied intervals share an edge. If there is an empty class, show its width with no rectangle.
For classes of equal width, a histogram can use frequency as the height. If widths differ, use frequency density, \(\text{density}=\text{frequency}/\text{class width}\). On a density scale each area gives the class frequency.
Read the boundaries to find each class width. Then multiply width by density, or read the marked frequency directly for an equal-width frequency-scale graph. The sum gives the total frequency.
A less-than cumulative frequency counts observations below each upper class boundary. Add one class frequency at a time. The final running total is the number of observations \(N\).
Put the upper class boundaries on the horizontal axis and their cumulative frequencies on the vertical axis. Start from the lowest class boundary with cumulative frequency zero. Join the points in increasing boundary order to show the cumulative curve.
Locate \(N/2\) on the vertical axis. Move horizontally to the ogive and then down to the measurement axis. That horizontal reading is the graphical median estimate.