IS / IS NOT analysis online
An IS / IS NOT analysis describes a problem by contrast: for what, where, when and how much, it records where the problem is and, next to it, where it could reasonably be but is not. The comparison removes guesses early and leaves a short list of causes that fit the facts.
Fill in the IS and IS NOT for each row, note what is special about the IS (the distinction) and what changed in or around it (the change). Then list the possible causes and test each one against every row.
01Method
A row counts only when both the IS and the IS NOT are filled. For each possible cause you mark whether it explains the pair (yes), explains it only with an assumption (only if) or cannot explain it (no). A single no rejects the cause.
Among the causes that remain and are tested against every row, the one with the fewest assumptions is shown as the most probable. It is still a hypothesis: verify it with a trial before you call it the root cause.
02When to use it
Use IS / IS NOT for D2 of an 8D, when a defect appears on some parts, lines, cavities or dates but not on others, and before a fishbone or 5 Why when the team has too many opinions and too few facts. It works best for a deviation from a state that used to be good, because then a change can be found.
PDF report in one click
Export the analysis as a landscape PDF with your header, the problem, the full IS / IS NOT table, the cause test matrix with results, and the assumptions to verify.
Private: runs in your browser
The analysis is saved only in your browser. Nothing is uploaded and no account is needed.
How to run an IS / IS NOT analysis
- One object, one defect. If several defects are mixed, split them into separate analyses.
- Fill the IS with facts you have seen or measured: part numbers, batches, dates, positions, counts.
- Find the closest IS NOT for each row. Good comparisons are the same part from another cavity, line or date.
- Write the distinctions: what is true only for the IS. Then the changes: what changed in or around each distinction, and when.
- List possible causes from the distinctions and changes, and test each against every row. A cause that cannot explain one row is out.
- Verify the most probable cause with a trial before the actions.
In an 8D the specification belongs in D2, so the tool can send it to the 8D report. After that, take the most probable cause into a 5 Why analysis for the occurrence and the escape root cause. When the comparison gives no clear distinction, collect candidates with the fishbone diagram or look at the variation with a multi-vari study.
Common mistakes
- Leaving the IS NOT empty or writing only “everything else”: the row then tests nothing.
- Writing opinions or causes in the IS column instead of observations.
- Choosing a far comparison (another plant, another product family) when a close one exists.
- Accepting a cause with many “only if” assumptions without checking them.
Frequently asked questions
›What is the difference between IS / IS NOT and 5W2H?
5W2H describes only where the problem is. IS / IS NOT adds the comparison: where the problem could be but is not. That comparison is what excludes causes; a description without it fits almost any cause.
›What should go into the IS NOT column?
The closest comparison that does not have the problem: the sister part made on the same machine, the other cavities, the parts made the week before, the other side of the part. It must be close enough that a cause could reasonably affect it too.
›What do the distinction and change columns mean?
The distinction is what is true only of the IS and not of the IS NOT, such as a different dryer, material lot or radius. The change is what changed in or around that distinction, usually just before the problem started. Most real causes are a change in a distinction.
›Is the most probable cause the root cause?
No. It is the hypothesis that fits the facts best. Confirm it by switching the cause on and off in a trial, then drill to the systemic root cause with a 5 Why for occurrence and escape.