Interpreting a Distribution to Answer a Question

Turn a Graph Into an Answer

In Rewriting a Weak Description Into a Full-Credit One, you practiced making claims specific, supported, and responsive to a question. Here the task is narrower: use a quantitative graph to answer a practical question such as, “What amount of sleep is most typical?” The key is to decide what the question means by typical and then use a feature the graph can actually show.

“Typical” does not always mean “most frequent.” If someone asks which exact value occurred most often, the relevant feature is the mode. If the question asks for a representative or central value, the median or a central cluster may be more useful. A skewed distribution or one with distinct clusters can make a single typical value misleading, so the answer should acknowledge the pattern when it matters.

Key idea: Match the question to the graph feature. A question about the most frequent exact value calls for the mode when the display preserves exact values. A question about a representative center may call for the median or an estimated central range. State the evidence and the limits of what the graph shows.

A graph’s resolution matters, too. A dotplot can show the exact observed values and their frequencies. A histogram groups values into intervals, so it may support an answer such as “most observations were between 7 and 8 hours,” but it generally cannot reveal the most frequent exact value inside that interval. A boxplot displays the median and quartiles, but not the individual values or detailed peaks.

A Question-to-Feature Routine

Before reading off a number, identify what kind of answer the question needs. “Which value happened most often?” is different from “About how much is a representative observation?” The first asks about frequency; the second asks about center. If the prompt uses the broad word “typical,” use its context and the graph to explain your interpretation.

Clarify the target.
Decide whether the question asks for the most frequent exact value, a central value, or a range containing many observations. Do not assume these are interchangeable.

Check what the display preserves.
Use exact stacks or values in a dotplot to identify a mode. Use a histogram’s intervals for approximate ranges. Use the median line of a boxplot for its displayed center, but do not infer exact modes from a boxplot.

Inspect the distribution around the candidate.
Check for skew, separated clusters, or isolated values, as in Reading Graphs for Shape, Gaps, and Outliers. These features may affect whether one number represents the observations well.

State the answer with evidence and appropriate precision.
Name the group and variable, give the value or range the graph supports, and explain why it answers the question. If the display is grouped or approximate, say so.

This routine is not a reason to calculate every possible summary. It is a way to select the information that answers the question. The SOCS framework and the earlier tutorials on choosing center measures help describe a full distribution; here, use only the features needed for the practical decision.

Worked Examples: Answering “What Is Typical?”

Worked Example: Typical Sleep on School Nights

A dotplot shows the number of hours slept on school nights by 15 fictional students. Reading the stacks from left to right gives these values: 4 hours (1 student), 5 hours (2), 6 hours (3), 7 hours (4), 8 hours (2), 9 hours (1), 10 hours (1), and 12 hours (1). A student asks, “What amount of sleep is most typical for this group?” Give a justified answer.

Solution. The word “typical” could mean the most frequent amount or a representative center. In this dotplot, 7 hours is the most frequent exact value: its stack contains 4 students, more than any other stack. The median also provides a central value. There are 15 observations, so the median is the eighth value in order. Counting the observations from the left, positions 1 through 3 are 4 or 5 hours, positions 4 through 6 are 6 hours, and positions 7 through 10 are 7 hours. Thus the eighth observation is 7 hours.

Both interpretations point to the same amount. The graph also shows that values are concentrated around 6 to 8 hours, while a few students reported higher amounts, including 12 hours. The dotplot supports an answer for the students shown; it does not establish the sleep habits of all students at other schools.

Answer in context. “For these 15 fictional students, 7 hours is a reasonable typical amount on school nights: it is the most frequent exact value, reported by 4 students, and it is also the median. Many of the displayed amounts are near 6 to 8 hours.”

The answer uses the exact frequency available from the dotplot and adds the median as supporting evidence. It does not claim that every student, or even most students, slept exactly 7 hours.

Worked Example: A Histogram Gives a Typical Interval

A frequency histogram summarizes the nightly sleep amounts of 30 fictional students. Its equal-width, one-hour bins have these counts: 4 to less than 5 hours, 2; 5 to less than 6, 4; 6 to less than 7, 7; 7 to less than 8, 8; 8 to less than 9, 5; and 9 to less than 10, 4. The question asks, “Which amount of sleep is most typical?” What can the histogram justify?

Solution. The largest count is 8, in the bin from 7 to less than 8 hours. Because the histogram groups observations into one-hour intervals, it does not show which exact sleep amount within that interval occurred most often. The appropriate evidence-based answer is an interval, not a specific decimal or whole-hour value.

The bins all have equal width. For equal-width bins, the tallest bar identifies the interval with the greatest count. Here, the bar for 7 to less than 8 hours is tallest. The distribution also has substantial counts in the adjacent intervals from 6 to less than 7 and from 8 to less than 9 hours, so it is reasonable to describe sleep amounts as concentrated around 7 to 8 hours rather than implying that all students had the same amount.

Answer in context. “For the 30 fictional students, the most common one-hour interval is 7 to less than 8 hours, with 8 students. The histogram supports describing sleep as typically around 7 to 8 hours, but it does not identify the most frequent exact amount within that interval.”

Notice the careful wording: “most common one-hour interval” is supported by the bar counts, while “most frequent exact amount” is not. A histogram’s bin boundaries and scale determine how specifically the values can be reported.

Worked Example: Unequal-Width Histogram Bins

A fictional community survey records sleep amounts for 32 adults. A density histogram has bins of unequal widths: 0 to less than 2 hours has density 3 observations per hour; 2 to less than 4 hours has density 5 observations per hour; 4 to less than 8 hours has density 3 observations per hour; and 8 to less than 12 hours has density 1 observation per hour. Which interval contains the greatest number of observations?

Solution. With unequal-width bins in a density histogram, bar height represents density, not the count in the interval. The area of a bar gives the interval’s count: multiply its density by its width.

\begin{aligned} 0\text{ to }2 &: 3(2)=6\text{ observations}\\ 2\text{ to }4 &: 5(2)=10\text{ observations}\\ 4\text{ to }8 &: 3(4)=12\text{ observations}\\ 8\text{ to }12 &: 1(4)=4\text{ observations} \end{aligned}

The counts sum to \(6+10+12+4=32\), matching the stated total. The tallest bar is the 2-to-4-hour bin, whose density is 5 observations per hour. But the 4-to-8-hour bin is wider, so its area—and its count—is larger: 12 observations compared with 10. Therefore, the 4-to-8-hour interval contains the greatest number of observations.

Answer in context. “Among the 32 fictional adults, the 4-to-less-than-8-hour interval contains the most observations, with 12. The tallest bar is not the interval with the largest count because the bins have unequal widths; comparing bar areas gives the counts.”

This example illustrates why the display type and its scale matter. For equal-width bins, bar heights can be compared to compare counts. With unequal-width bins in a density histogram, compare areas. Do not describe the tallest unequal-width bar as containing the most observations unless its area supports that claim.

Worked Example: A Boxplot and a Skewed Distribution

A boxplot summarizes fictional one-way travel times, in minutes, for 40 commuters. Its five-number summary is 12, 18, 24, 35, and 70 minutes. A commuter asks, “What is a typical travel time?” Answer using only what this boxplot supports.

Solution. The median is 24 minutes, so half of the displayed travel times are at or below 24 minutes and half are at or above 24 minutes. The median is a defensible measure of a typical travel time, especially since the upper whisker extends much farther from the box than the lower whisker: from 35 to 70 minutes is 35 minutes, whereas from 12 to 18 minutes is 6 minutes. That pattern suggests a longer tail toward higher times.

The boxplot supports reporting the median, but it does not show which exact travel time was most frequent or where any histogram peak occurs. It also does not show the individual commute times. A response should not turn the median into a claim that most commuters take exactly 24 minutes.

Answer in context. “For these 40 fictional commuters, the median one-way travel time is 24 minutes, a reasonable central or typical value. The longer upper whisker suggests that some travel times extend considerably above the median; the boxplot does not identify the most frequent exact time.”

Common Mistakes and AP Exam Tips

Treating “typical” as automatically meaning the mode. The most frequent value and the median answer different questions. State which meaning you are using, and support it with the graph.
Reporting an exact value from a histogram bin. A bar from 7 to less than 8 hours groups values in that interval; it does not reveal which exact amount occurred most. Report the interval or say “around” the interval.
Comparing bar heights in an unequal-width density histogram to find the largest count. With unequal-width bins, calculate or compare bar areas. Height represents density; area represents relative frequency or count, depending on the vertical scale.
Calling the median the most frequent value. The median marks the middle ordered observation or observations. It need not be the mode, and a boxplot does not display the mode.
Forcing one typical value onto a distribution with distinct clusters. If a graph has two separated concentrations, a single center may describe neither group well. Name the clusters or explain that one overall typical value could be misleading.
Overstating what the graph says. A graph of a sample supports statements about the observations shown. It does not, by itself, explain why values differ or prove that the same pattern holds for a broader population.

For a full-credit response, make the link between question and evidence explicit. For example: “The 7-to-less-than-8-hour interval is the most common interval, with 8 observations, so sleep is typically around 7 to 8 hours in this group.” That is stronger than “7 hours is typical” when the graph is a histogram, because it respects the grouping and gives the evidence.

Key takeaway: First decide whether “typical” means most frequent, central, or concentrated in a range. Then use a graph feature that supports that meaning, report only the precision the display allows, and explain the evidence in context.

Check Your Understanding

For each question, identify the graph feature that would support a careful answer and note any limit on what can be concluded.

A dotplot of daily water use has its tallest stack at 120 liters. What does that establish about the displayed observations, and what does it not establish about all households?
A histogram has equal-width bins, and the tallest bin runs from 10 to less than 15 minutes. What is the most precise claim about the most common interval that the graph supports?
A density histogram has unequal-width bins. What should you compare to decide which interval contains the most observations?
A boxplot has a median of 6 days. Write one careful sentence describing a typical value, without claiming that 6 days was the most frequent exact value.
A histogram has two separated peaks. Why might reporting one overall “typical” amount be incomplete?