Cumulative graphs organize quantitative data by showing how totals build across values, helping students see overall patterns, estimate proportions below cutoffs, and answer questions about accumulated counts.

What a Cumulative Graph Shows

A cumulative graph is useful when interest centers on totals up to a value rather than frequencies in separate intervals. The graph builds running totals as the quantitative variable increases. On the horizontal axis, values or class boundaries appear in order. On the vertical axis, the graph shows either a cumulative count or a cumulative proportion. Because the totals accumulate, each point represents all observations at or below that location.

Relative cumulative frequency (ogive) for a quantitative variable, with cumulative percentile/proportion on the vertical axis and x-values on the horizontal axis. The monotone increasing shape shows how the running total builds as $x$ increases, with steeper segments indicating ranges where observations accumulate quickly and flatter segments indicating few observations in that range. Source

A cumulative graph emphasizes how a distribution grows. Early parts of the graph describe the lower end of the data, while later parts represent larger values. A steeper section means observations are being added quickly over that range. A flatter section means relatively few observations lie there. This makes the display especially helpful for locating cutoffs and judging how data accumulate across the scale.

Counts and Proportions

Some cumulative graphs use counts, and others use proportions. Both communicate the same ordering idea, but proportions are often easier when sample sizes differ or when results need to be interpreted as percentages. In a proportion-based graph, the final value is 1, or 100%, because all observations have been accumulated by the largest value.

When reading either version, always connect the vertical value to the phrase less than or equal to. A cumulative value of 0.68 at 12 does not mean 68% are exactly 12. It means 68% of observations are 12 or less. That distinction is essential.

How to Read a Cumulative Graph

Reading a cumulative graph requires attention to both axes and to the direction of accumulation. Start with a specific x-value, move upward to the graph, then move across to the vertical axis. The vertical coordinate gives the number or proportion of observations at or below that value. Reversing the process is also useful: start with a cumulative count or proportion and find the corresponding value on the horizontal axis.

Ogive annotated with horizontal-then-vertical guide lines that show how to translate between a cumulative proportion (y) and its corresponding cutoff value (x). This is the graphical workflow for locating percentiles: pick a cumulative level on the y-axis, move to the curve, then read the associated x-value. Source

Common interpretations include:

• finding how many observations are less than or equal to a cutoff

• finding what proportion of the data fall below a benchmark

• estimating the value where a chosen cumulative percentage is reached

• comparing how rapidly the data accumulate across different ranges

A cumulative graph should never decrease as you move from left to right.

Cumulative count over time (an empirical cumulative total), illustrating that cumulative graphs are nondecreasing because totals only add as time advances. Flat portions indicate little or no additional accumulation over that interval, while steep portions indicate rapid accumulation. Source

Counts and proportions can stay the same over an interval if no observations are added, but they cannot go down. If a graph drops downward, it has been constructed incorrectly or read incorrectly.

Constructing and Interpreting from Data

To create a cumulative graph, the data must first be ordered from smallest to largest, or grouped into ordered intervals. Then running totals are formed.

Basic steps are:

• order the values or class intervals from low to high

• calculate the cumulative count or cumulative proportion at each value or upper class boundary

• plot each boundary with its accumulated total

• connect the plotted points in a way that reflects the data structure used

With grouped data, cumulative graphs are typically plotted at upper class boundaries because each point represents observations less than or equal to the top of that interval. This matches the meaning of cumulative frequency. If the data have been grouped, the graph provides an approximation rather than a perfect record of individual observations.

That approximation matters. A grouped cumulative graph can answer questions about overall accumulation, but exact values inside an interval may not be known precisely. Any estimated cutoff taken from the graph should therefore be described as approximate, especially when intervals are wide.

Other Quantitative Displays and Choosing a Graph

The syllabus notes that other quantitative graphs are possible. This matters because no single display is best for every statistical question. A cumulative graph is specialized: it highlights accumulation. If the goal is to know how many values fall at or below a threshold, it is often the clearest choice. If the goal is different, another type of quantitative display may be more informative.

A cumulative graph is especially strong for:

• answering at most questions

• estimating the share of observations below a standard

• locating approximate cut points in the ordered data

• showing whether data accumulate gradually or rapidly across regions

A cumulative graph is less effective for:

• showing exact frequencies in isolated intervals

• displaying individual observations

• revealing local features that do not strongly affect the running total

Because the cumulative totals only move upward or stay level, some details of distribution shape are compressed. Two data sets can have similar cumulative patterns even if their internal interval-by-interval frequencies differ. For that reason, statisticians choose graphs based on the question being asked, not just on habit. When the question is about the number or proportion of observations less than or equal to a value, a cumulative graph directly answers that question.