Skip to content
LeanSuite - Lean Manufacturing Software

Free template

Updated

Histogram template

A histogram is a bar chart of how many readings fall into each of a row of equal ranges, called bins, so the shape of a process's variation shows at a glance: where it is centered, how widely it spreads, and whether it has one peak, two, or a long tail. Drawing the specification limits on it shows how many parts fall outside them. It is one of the seven basic quality tools. Page 1 of the PDF has the header fields, a table for 100 readings, a tally table for up to 12 bins, a box for the numbers that set the bins (n, largest, smallest, range, number of bins, bin width and first edge) and a grid to draw the bars and the limits. Page 2 has the steps for setting the bins, six common shapes and what each often means, and a worked example drawn to scale. The Excel version takes up to 125 readings, works out n, the largest, the smallest, the range, the mean and the standard deviation, sets the bins by the square root rule (with Sturges' rule shown for reference and optional overrides for the number of bins, the width and the first edge), counts each bin with its % and cumulative %, counts the readings below the LSL and above the USL, and draws the histogram with the readings outside the limits in red. It has a worked example sheet and a shapes sheet.

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Page 1 of the histogram template: fields for part, characteristic, units, LSL, USL, gauge and resolution, a table for 100 readings, a tally table for 12 bins beside a box for n, range, bins, bin width and first edge, and a grid to draw the bars.
100 readings (125 in Excel)Bins by the square root ruleTally and countsBar grid with spec limitsShare out of specSix common shapesWorked example

When to use it

When to draw a histogram

  • To see where a measured characteristic sits against its specification and how widely it spreads, before deciding whether to re-center the process or reduce its variation.
  • As a first look before a capability study, to check that the readings have one peak and roughly a bell shape.
  • When parts from two machines, cavities, shifts or material lots may be mixed: two peaks or a flat top show it.
  • To check a supplier's parts against the specification: a side that ends in a sharp cliff often means the lot was sorted.

How to fill it in

  1. 1

    Collect the readings

    At least 50, ideally 100, from one period of normal running. Write them in the order taken and note the gauge and its resolution.

  2. 2

    Find the range

    Count the readings (n), find the largest and the smallest, and work out range = largest − smallest.

  3. 3

    Set the bins

    Number of bins = √n, rounded, as a starting point. Width = range ÷ number of bins, rounded up to a whole number of resolution steps. First edge = smallest − half the resolution, so no reading sits on an edge.

  4. 4

    Tally and count

    Put one stroke per reading in its bin, then count each bin. The counts must add up to n.

  5. 5

    Draw it

    One bar per bin, with no gaps, as high as its count. Draw the LSL, the USL and the target, and count the readings outside the limits.

  6. 6

    Read the shape

    Compare it with the common shapes on page 2, then check the likely cause at the process before acting.

Shapes

Common histogram shapes and what they often mean

A shape is a clue, not a verdict: check the cause at the process before acting on it.

  • Bell (normal)

    What it looks like
    One peak in the middle, tails falling off evenly on both sides
    What it often means
    Usual variation only: check where it sits against the limits
  • Skewed

    What it looks like
    Peak off center, a long tail on one side
    What it often means
    A natural limit on the short side (zero, a stop, flatness, runout), or a process adjusted from one side
  • Two peaks (bimodal)

    What it looks like
    Two humps with a dip between them
    What it often means
    Two sets of data mixed: machines, cavities, shifts, operators or material lots. Split the readings by source
  • Cliff (cut off)

    What it looks like
    One side ends sharply
    What it often means
    Parts sorted at inspection, readings past a limit not written down, or the end of the gauge's range
  • Plateau

    What it looks like
    A flat top over several bins
    What it often means
    Several streams with slightly different means mixed, or a process drifting (tool wear) during the period
  • Isolated peak

    What it looks like
    A small group apart from the rest
    What it often means
    A short abnormal event: a wrong setting, another material lot, a measuring error or a mix-up

A filled-in example

Illustrative, not a benchmark

An example: the diameter of a turned shaft, specification 25.00 ± 0.10 mm, 100 readings to 0.01 mm (illustrative numbers). The Excel version's Worked example sheet has all 100 readings.

  • n = 100, largest 25.12 mm, smallest 24.94 mm, range 0.18 mm. Mean 25.030 mm, standard deviation 0.0338 mm.
  • Bins: √100 = 10. Width 0.18 ÷ 10 = 0.018, rounded up to 0.02 mm. First edge 24.94 − 0.005 = 24.935 mm.
  • Counts from 24.935 mm up, bin by bin: 1, 3, 12, 17, 22, 20, 17, 5, 2, 1.
  • No reading below 24.90 mm and 2 above 25.10 mm (25.11 and 25.12): 2.0% out of specification.

One peak and close to a bell, so nothing points to mixed streams. But the process is centered 0.030 mm above nominal and its spread nearly fills the 0.20 mm tolerance, so the top tail runs past the USL. The action was to re-center the process on 25.00 mm, then draw a new histogram to check.

Common mistakes

  • Too few readings

    With 20 or 30 readings the shape owes a lot to chance. Use at least 50, ideally 100.

  • Bins that do not suit the data

    Too few bins hide the shape and too many make it ragged; a width that does not match the resolution gives a comb of alternating high and low bars. Start from √n and round the width to the resolution.

  • Mixing streams without noting it

    Readings from two machines or cavities in one histogram can look like one wide process. Note where each reading came from and draw the groups apart when the shape looks odd.

  • Reading a histogram as a trend

    It hides the order of the readings, so a drift looks like a wide spread. Use a control chart to see change over time.

Download the template

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Run this template in LeanSuite

In LeanSuite defects are captured as they happen and traced to their root causes with 5 Whys, 4M1D and fishbone, and KPI Builder tracks quality KPIs against targets and shows where defects happen.

FAQ

Histogram template: common questions

More free templates

All templates

Can't find the template you need?

Tell us what you're looking for. We'll make it and send it to your inbox, free.

A shift handover sheet, a red tag log, a changeover checklist for one machine: if your team uses it on the floor, ask.

By sending this form you agree to our Terms of Service and Privacy Policy.