Describing a Histogram With Real Numbers

Read the Scales Before Describing the Bars

In Writing a Complete Description of a Distribution, you learned to connect a distribution’s visible pattern with numerical summaries. This tutorial focuses on one practical skill for histogram descriptions: using the axis scales to report value ranges, locate a peak, and estimate counts. The bars only become useful evidence when you can tell what their horizontal positions and vertical heights represent.

A histogram’s horizontal axis shows intervals of values for a quantitative variable. The vertical axis might show frequency, or count: how many observations fall in each interval. To describe the graph with numbers, first locate the interval boundaries on the horizontal scale. Then read each relevant bar’s height against the vertical scale. A reading from a graph is often approximate, even when the axis labels are exact.

Keep the two axes’ jobs separate. If a bar covers values from 10 to 15 minutes and reaches a count of about 17, the interval comes from the horizontal axis and the estimated number of observations comes from the vertical axis. The height does not tell you that the most common individual value is 17, or that the observations all equal the bar’s midpoint.

Translate Tick Marks Into Values

Tick marks are reference points on an axis. They may be labeled at every interval boundary, or only at regular wider steps. For example, an axis labeled 0, 10, 20, 30 might have bars that are each 5 units wide. In that case, use the spacing between the labeled values to locate the unlabeled boundaries: halfway from 10 to 20 is 15.

Check that the scale is evenly spaced before estimating. If adjacent vertical tick marks represent 5 observations, then a bar halfway between the 10 and 15 marks is about 12 or 13 observations high. If the graph’s bars line up with tick marks, read the marked value directly. If a bar falls between marks, estimate its height based on its position and report it as approximate.

A histogram’s peak is the interval or region with the greatest bar height. Since each bar represents a range of values, describe the peak as “the tallest bin is from 10 to 15,” rather than saying “the peak is 12.5.” The midpoint of a bin is not necessarily an observed value or the most frequent individual value. When neighboring bars are similarly high, it can be more accurate to describe a broad region of high concentration than to overstate one bar as a uniquely clear peak.

The first and last nonempty bars can help you describe the range of values displayed. But their boundaries do not usually reveal the exact smallest and largest observations. If the first nonempty bar covers 0 to 5, the smallest observation is somewhere in that interval, not necessarily 0. Say that the histogram’s nonempty bars extend from 0 to 30, or that the values are shown across bins from 0 to 30. Do not claim the exact minimum is 0 unless the problem gives that information.

Estimate Counts Across an Interval

To estimate the count in a range that covers whole bins, read each relevant bar’s height and add those heights. As covered in Reading Counts and Percents From a Histogram, adding the frequencies of complete bins gives the count in their combined range. Here, because heights may be read visually rather than from a frequency table, describe the sum as approximate.

Be cautious when a question asks for a count in a range that cuts through a bin. A histogram shows how many observations are in that whole bin, but it does not show where those observations lie inside it. Without more information, you cannot know exactly what fraction of the bar’s count belongs to only part of the bin. You may give an estimate if the question invites one, but state that it is an estimate and do not present it as an exact count.

Also check what the vertical axis measures. If it is labeled “count” or “frequency,” a bar’s height can be read as an estimated number of observations. Relative-frequency and density histograms use different vertical scales, as discussed in Relative Frequency and Density Histograms. Do not call a height a count unless the axis supports that reading. With a density histogram, for example, count information is represented by bar area rather than height alone.

Worked Examples

Make Numerical Descriptions Precise

A useful histogram description connects each number to the feature it measures. State the variable and units, then give the relevant interval and an approximate count when appropriate. For example, “The tallest bin is from 20 to 30 centimeters and contains about 35 locations” is clearer than “The peak is 35.” The latter does not say whether 35 is a depth, a count, or something else.

When reporting the range shown by the graph, distinguish bin boundaries from observed values. If bars extend from 0 to 60 centimeters, say that the nonempty bins span 0 to 60 centimeters. Do not say the exact observed depths range from 0 to 60 unless the graph or problem provides the exact minimum and maximum. This distinction matters because every bar combines values within an interval.

Likewise, do not overstate what the tallest bar tells you. It identifies the bin with the largest count, not a single value that occurred most often. In Unimodal, Bimodal, and Multimodal Distributions, you learned to identify peaks as regions of concentration rather than treating every bar-to-bar change as a separate peak. Apply that same care when reporting numerical locations: a broad concentration can occupy several neighboring bins.

If the graph has count ticks every 10 observations, do not assume every bar height is a multiple of 10. Estimate a bar’s position between ticks. For instance, a top a little above the 20 mark might represent about 23, not exactly 20 or 30. An estimate should reflect the scale and the visual position, not a guess detached from the tick marks.

Finally, be clear about whether a requested range matches complete bins. If the range is 10 to 20 and the bins are 5 units wide, you can add the two complete bin counts. If the question asks for 12 to 18, those boundaries cut through bins. The histogram alone does not reveal how many observations fall on each side of the cut, so an exact count is unavailable from the display.

Common Mistakes and AP Exam Tips

• Mixing up the axes. Horizontal values define a bin’s range; vertical values give its count when the axis is labeled with frequencies. A full-credit statement names the variable’s interval and the estimated number of observations separately.
• Calling the peak an exact data value. “The peak is 15 minutes” can confuse a bin boundary or midpoint with an observed value. Say, “The tallest bin is 10–15 minutes.”
• Reporting estimated heights as exact counts. When a bar lies between tick marks, use “about” or “approximately.” Explain that the count is estimated from the scale.
• Treating bin limits as the exact minimum and maximum. State that the nonempty bins extend across a range. Do not claim an exact extreme observation unless it is given separately.
• Adding bars that do not match the requested interval. Check the horizontal boundaries before summing heights. Add complete bins only; when a range cuts through a bin, the graph does not reveal the exact partial-bin count.
• Reading a count from the wrong kind of vertical scale. A frequency histogram’s height is a count. If the axis shows relative frequency or density, describe that scale correctly rather than calling the height a number of observations.
• Ignoring units and context. “About 35” is incomplete on its own. A strong response says what was measured, which interval is involved, and what the estimated count represents.