How do I spot a graph that leaves out relevant categories?

Start by identifying exactly what the graph claims to show. Read its title, axis labels, legend, units, date range, and any footnotes. Then ask: “Who or what could be counted here, and what categories would I expect to see?” A graph titled “Survey responses” might list only named response options; check whether there is also a category such as “Other,” “Unknown,” or “No response.” These are prompts for inspection, not proof that anything was deliberately excluded. The Library of Congress recommends asking what seems important and what might be hidden, and comparing related charts to see what each highlights or leaves out [S2].

If the graph provides counts and a full-group total, add the displayed counts and compare their sum with that total. If it shows percentages, check whether they account for the whole, allowing for rounding where relevant. For example, suppose a hypothetical class survey has 100 responses, but a graph shows 40 “A,” 35 “B,” and 15 “C.” Those bars account for 90 responses, so ask what the remaining 10 represent and whether the graph has a note explaining them. Do not assume the answer: it might be an unlisted response, missing information, or something else. A textbook example shows why the check matters: the listed student categories totaled 90.4% of the full group, and the omitted “Other/Unknown” category was 9.6% [S1]. That example supports checking totals and unlisted categories; it does not tell you why a category was left out in another graph.

Check the first and last dates, the spacing between points or bars, and whether the labels suggest regular intervals. Ask whether the stated period matches the question you are trying to answer—for instance, whether a graph labeled “annual results” shows every year in the interval it names. Look for a note that explains gaps or changes in coverage, and compare with a related chart if one is available. If dates are missing or unevenly spaced, describe that as an observation rather than guessing the reason. The cited textbook example concerns categories, not a general method for detecting missing time periods [S1]; the practical date checks here are questions to guide your reading, not a guarantee that every omission will be visible.

Separate what you can see from what you infer. State the graph’s stated scope, identify the apparent gap, and explain how it limits your conclusion. For example: “The chart shows three response categories. Their counts total 90, while the caption says there were 100 responses. The remaining 10 are not identified here, so this chart alone does not show how all responses were distributed.” If you cannot verify the total or the intended scope, say so. Comparing related charts can help you notice differences in what is shown, but a difference does not establish the reason for it [S2].

Reusable note:

1. The graph says it covers: [group, subject, and period].
2. Categories or dates shown: [list].
3. Stated full total, if available: [number].
4. Displayed values account for: [number or percentage].
5. Possible gap to clarify: [category, dates, or “unclear”].
6. What the graph supports: [limited conclusion].
7. What it does not establish: [reason for any absence].