A histogram has a conspicuous empty stretch between two groups of bars. Has the student forgotten to draw a bar, or did no observation fall into that interval? The answer cannot be decided from the white space alone. The first job is to find the bin definitions and the underlying counts.
A zero-frequency interval is a defined interval containing no observations. In an ordinary equal-width frequency histogram its bar has height zero, so it leaves a gap along the baseline. A missing interval is different: its frequency is unknown or its coverage has not been shown. Check the data table before calling a blank interval an error, and never invent a count to make the picture look continuous.
- Did you know? Adjacent bins can be consecutive even though one has no bar because its frequency is zero.
- Immediate method: define non-overlapping intervals, tally every valid value once, include zero-count intervals, then plot.
- Main warning: a gap is not proof of zero if the chart has skipped a time range, filtered records or omitted a category.
Start with the question the chart must answer
Histograms group numerical observations into intervals, often called bins. They are not bar charts of unrelated names. A histogram of journey times might use 0 to under 5 minutes, 5 to under 10, 10 to under 15 and 15 to under 20. Each interval covers a stretch of a numerical scale. The bins should meet without overlapping or leaving an undefined region if they are intended to cover the full measurement range.
Ask the learner to state the horizontal quantity, its unit and whether the boundary is included. A convention such as 5 ≤ t < 10 minutes is precise: a 5.0-minute journey belongs there, while a 10.0-minute journey belongs in the next interval. A class that alternates between including both endpoints can accidentally count boundary measurements twice.
A complete, hand-checked worked example
The following numbers are invented solely for teaching. Twelve journeys to a community library, measured to the nearest minute, took 2, 3, 4, 5, 7, 9, 15, 16, 17, 18, 19 and 19 minutes. These are not real pupils or school records. Assume all measurements were recorded and the intervals are left-closed and right-open.
- 0 ≤ t < 5 minutes: 2, 3, 4 → frequency 3.
- 5 ≤ t < 10 minutes: 5, 7, 9 → frequency 3.
- 10 ≤ t < 15 minutes: no values → frequency 0.
- 15 ≤ t < 20 minutes: 15, 16, 17, 18, 19, 19 → frequency 6.
Check the total: 3 + 3 + 0 + 6 = 12 journeys, exactly matching the twelve values. A histogram using equal five-minute widths would show heights 3, 3, 0 and 6 on a frequency axis. The interval from 10 to under 15 remains part of the declared range even though there is no rectangle rising above its baseline.
Why the blank region matters
If a student deletes the empty interval and draws the last bar immediately next to the 5–10 bar, the chart visually suggests that 10–15 minutes did not exist as a possible range. That changes the horizontal scale. It also hides an interesting feature of the observations: the journeys were clustered in two separated groups. Preserving the zero bin reveals the pattern without pretending to explain what caused it.
Checking proportions without changing the denominator
The zero bin contains 0/12 = 0% of the recorded journeys. The 15–20 bin contains 6/12 = 50%. The denominator is the total number of valid journeys, not the number of non-empty bars. A learner who divides by three because only three visible bars rise above the axis has confused a count of categories with a count of observations.
What if the class intervals have different widths?
This worked example deliberately uses equal-width bins. For a histogram with unequal widths, a correctly constructed frequency-density histogram uses area to represent frequency, so height is frequency divided by class width. A zero-frequency bin still has zero area, but the heights of other bins cannot simply be compared as raw counts. Learn that separate rule before interpreting irregular-width histograms.
For an existing related example, eduKate explains how frequency density works when histogram widths differ. The present lesson focuses on why a defined interval can correctly have no observations.
Diagnose the misunderstanding rather than correcting the drawing for the student
Show a learner a frequency table with four consecutive bins: 4, 2, 0 and 5. Ask, ‘What must appear on the horizontal axis between the second and fourth bins?’ Then present a chart with a silent gap but no table and ask, ‘Can you prove the missing frequency is zero?’ A strong answer distinguishes the first question, where zero is documented, from the second, where the chart alone does not establish whether the data are missing.
Three common wrong answers reveal three different problems. ‘Every bin must have a bar’ indicates confusion about zero. ‘Just join the bars across the gap’ indicates failure to preserve the horizontal scale. ‘The journey of exactly ten minutes belongs in both bins’ indicates an endpoint problem. These need different explanations and therefore different practice.
Guided practice, then independent transfer
Guided task: draw from the table
Supply five equally spaced intervals of four units each, with counts 1, 0, 4, 0 and 2. Have the learner mark every interval boundary first, then create the non-zero rectangles. The second and fourth intervals stay labeled even when their bars vanish. Ask for the total frequency: 1 + 0 + 4 + 0 + 2 = 7 observations.
Independent task: decide whether the gap is genuine
A different graph shows 0–5 minutes with frequency 8 and 15–20 minutes with frequency 4 but provides no counts for 5–10 and 10–15. Is it safe to say nobody took between 5 and 15 minutes? No. Those intervals were not documented. Before answering, obtain the original data or a complete frequency table. If there were genuinely zero observations there, show the bins and their zero counts explicitly.
Extension: compare the same observations under a new binning
Change the first example to 0–10 and 10–20 minutes. The new frequencies become 6 and 6. The blank 10–15 interval disappears from the picture because the new grouping combines it with 15–20. The data did not change; the bins did. Ask whether a two-bin graph still reveals the separation between groups. It does not. This teaches that plotting choices influence what a viewer can notice, even when every plotted count is correct.
What parents and teachers can do next
- Parent: ask the child to explain the zero interval using the original observations rather than telling them how many bars to draw.
- Teacher: require a small frequency table and interval convention beside early histogram exercises.
- Student: count all values back to the original dataset, label the axis with units and ask whether every possible interval has been covered.
- Tutor: vary one feature at a time—empty bins, boundary observations, then unequal bin widths—to locate the first unsupported inference.
Progress is visible when the learner can build a table from new measurements, keep a zero-count interval on the axis without prompting, and explicitly say what cannot be concluded when the data source is absent. That is more dependable than copying an attractive picture.
Limitations: blank space alone is weak evidence
The chart might be cropped, the data might be filtered, the sample might be incomplete, or the author might have chosen non-contiguous ranges. A zero-frequency interval is a fact about a defined dataset and binning rule, not automatically about the wider population. With small datasets, an empty bin can arise by chance. Do not turn ‘none of these twelve journeys took 10–15 minutes’ into ‘nobody ever takes 10–15 minutes’.
Frequently asked questions
Is it wrong to have an empty interval?
No. It is correct when the interval was defined and the verified frequency is zero. Retain its place on the numerical axis.
Must bars in a histogram always touch?
For consecutive non-empty intervals they normally meet at their shared boundary. If an intervening interval has zero frequency, the gap is meaningful. Do not slide bars together and change the measurement scale.
Can a blank interval mean data were not collected?
Yes. The graphic alone may not tell you. Check the title, notes, axis and original counts; state uncertainty if the source does not resolve it.
Why not just use a bar chart?
A bar chart can compare named categories, whereas a histogram encodes numeric intervals and their widths. Choosing the wrong chart can disguise the meaning of a gap.
Useful reading and sources
For more on honest bins, see the Core Aim guide to non-overlapping age intervals. The eduKate Education Hub connects this data-literacy lesson to wider learning, and the Clementi Mathematics teaching reference shows the value of identifying the broken reasoning step before increasing the amount of practice.
A histogram is useful not because every region is filled, but because it preserves what was counted and what was not. Sometimes the most important bar is the one whose height is exactly zero.