How to make a histogram
- Copy your measurement column from Excel together with its header.
- Click the column header (A) in the data table and paste (Ctrl+V) - the first line becomes the column name. You can also open a CSV or Excel file.
- Select the columns to plot and, if needed, enter the lower (LSL) and upper (USL) spec limits.
- Press Draw and check the chart, statistics and automatic interpretation. Right-click the chart to save or copy it as an image.
What is a histogram?
A histogram splits measurements into equal-width intervals (bins) and shows how many values fall in each bin as the bar height. It shows at a glance where the data are centered, how widely they are spread, and whether they lean to one side (skewness).
In quality control it is the first step in checking whether process data follow a normal distribution. Overlaying a normal curve with the same mean and standard deviation makes it easy to compare the actual distribution with the theoretical one.
How is the bin width chosen?
By default the tool uses the finer of Sturges' rule and the Freedman-Diaconis rule (NumPy's 'auto' method), which gives a sensible number of bars for most data. When several columns are overlaid it uses the narrowest width among them so a tightly clustered group is not squashed into one or two bars. You can also type your own bin width.
How Pp and Ppk are calculated
Enter the lower (LSL) and upper (USL) spec limits and the capability indices are computed with the overall standard deviation (σ): Pp = (USL − LSL) ÷ 6σ and Ppk = min(USL − mean, mean − LSL) ÷ 3σ. With only one limit, the one-sided index (Ppu or Ppl) is shown as Ppk.
The sigma multiplier is entered as the total of both sides. The default of 6 means a process spread of ±3σ, i.e. 6σ in total.
How to read a histogram
- Bars that peak in the middle and fall off evenly on both sides suggest a stable, roughly normal process.
- A long tail on one side (high skewness) means the normal assumption may not hold - interpret Ppk with care.
- Two or more peaks often mean data from different conditions (machines, materials, operators) are mixed.
- Ppk of 1.33 or more is excellent, 1.0-1.33 is adequate with room for improvement, and below 1.0 carries a risk of out-of-spec output.
Frequently asked questions
Can I paste data straight from Excel?
Yes. Copying from Excel or Google Sheets gives a tab-separated table that is pasted as is. Click a column header before pasting to use the first line as column names. Blank and non-numeric cells are left out of the calculation.
Is my data sent to a server?
No. All calculations and charts are produced inside your browser. The data is never stored or transmitted anywhere.
How is Ppk different from Cpk?
Ppk uses the standard deviation of all data (long-term variation), while Cpk uses the within-subgroup standard deviation (short-term variation). This tool calculates Pp and Ppk. Cpk and control-chart based analysis are available in the Qstat desktop program.
Can I compare the distributions of two groups?
Yes. Select several columns to overlay them on one chart, and the automatic interpretation compares their means and standard deviations. You can also pick one data column and a group column to split by group.
Other chart tools
Need process capability, control charts or DOE?
The Qstat desktop program offers about 250 analysis features: histograms and box plots plus process capability (Cp, Cpk), control charts, measurement system analysis (MSA), design of experiments (DOE), regression and machine learning.
Try Qstat free for 30 days