Histogram
Shows the distribution of a measurement: where values gather, whether the spread is symmetric, whether anything sits at the edges. You give raw measurements, it bins them, and with specification limits it also computes process capability.
Register nowWhen to use it?
- To see what the average hides: the same mean can come from very different distributions.
- To judge how consistently a process runs — a narrow spread means consistent, a wide one means erratic.
- To see how often measurements fall outside specification and whether the process can meet demand.
Step by step
- 1Create the analysis; switch between analyses in the same project from the list at the top.
- 2Type the measurements into the box on the left, or paste a list as is. One value per line; decimals may use a comma or a dot.
- 3The tool picks the number of classes itself (Sturges rule). Override it with your own number if you prefer.
- 4Enter the lower and upper specification limits. They appear as red dashed lines and out-of-spec columns turn red.
- 5Below you get count, mean, standard deviation and range; with both limits entered, Cp and Cpk as well.
Keyboard shortcuts (desktop)
Tips
- The grey curve is a normal distribution with the same mean and deviation. Columns departing clearly from it point to a special cause.
- A two-peaked distribution usually means two different processes (two shifts, two machines) got mixed into one table.
- Cpk of 1.33 and above is generally considered capable; below 1 means the process cannot hold the limits.
- Good Cp with poor Cpk means the spread is tight but the mean has drifted — an adjustment fixes it, no need to narrow the distribution.
Worked example
Example: delivery times
Average delivery time is reported as 3 days, which looks fine. Complaints keep coming anyway. Grouping the individual times reveals what the average was hiding.
Delivery time distribution (500 orders)
- 1 day — 140 orders
- 2 days — 165 orders
- 3 days — 95 orders
- 4 days — 30 orders
- 5 days — 12 orders
- 6 days or more — 58 orders
Reading it
- 60% of orders arrive within two days
- A small but distinct cluster sits at six days and above
- The shape has two peaks, not one
- Three days — the average — is one of the least common outcomes
The average says three days, but there are really two different customer experiences: most people get it in two days, some wait a week. A two-peaked distribution always means the same thing — this is not one process, it is two. The next question is which region or which warehouse those 58 orders came from.
Frequently asked questions
What is a histogram?
A chart that splits measurements into intervals and shows how many fall into each one. It makes visible the thing an average hides: how the values are spread out.
What is the difference between a histogram and a bar chart?
A bar chart shows categories, and you can reorder them — cities, products and so on. A histogram has a numeric axis, the order is fixed, and the bars touch. What decides which one you need is the type of the data.
How many bins should I use?
A common starting point is roughly the square root of the number of measurements — about ten for 100 readings. Too few bins erase the shape; too many turn noise into apparent structure. Try a couple of values and keep the one where the shape stays stable.
What does a histogram with two peaks mean?
Almost always that the data did not come from a single process — two shifts, two machines, two regions. When you see that shape, the first thing to do is split the data and look at each part separately.
Is it free?
Yes. Klarsti is free and ad-free right now, and you do not need an account to build a histogram.