What are the four dimensions of is / is-not analysis?

What (which object and which defect), where (location on the object and geographically), when (first seen, point in cycle, shift, trend), and extent (how many, how big, the trend). You answer each dimension in both an IS column and an IS-NOT column, where IS-NOT holds plausible comparisons that stay clean.

Why do the is-not entries matter?

Because a problem's cause lives in the differences between IS and IS-NOT. If line 3 has the defect and line 4, running the same part on the same shift, does not, the IS-NOT makes that contrast a lead rather than a stray fact. A true cause must explain not only where the problem appears but why it is absent everywhere the IS-NOT says it is absent.

How do distinctions and changes point to the root cause?

A distinction is anything true of the IS but not the IS-NOT, such as line 3 using a particular die. A change is something recently modified around a distinction. Root causes cluster where a distinction meets a recent change and the timing fits, for example a reground die installed on line 3 just before the first bad part. That intersection names the prime suspect.

When should you use is / is-not analysis?

Use it at the very start of an investigation, before brainstorming causes. It is a front-end tool that turns a vague complaint into a sharp specification, so cause-finding tools work on a bounded target. It also right-sizes containment, because the IS-NOT column shows how limited the problem really is.

How is is / is-not analysis different from the 5 Whys and fishbone diagrams?

Is / is-not defines and bounds the problem; the 5 Whys and fishbone find and drill the cause. Run is / is-not first to narrow the field, use a fishbone to organize the surviving candidate causes by category, then use 5 Whys to dig a confirmed cause to its root. They are complementary stages, not competitors.

Where did is / is-not analysis come from?

It is part of the Problem Analysis process developed by Charles Kepner and Benjamin Tregoe, which specifies a problem across what, where, when, and extent and compares what the problem is with what it is not. It is widely taught as a rational problem-solving and quality tool alongside root-cause methods.

Is / Is-Not Analysis for Root Cause

Bound a problem by what it IS and IS NOT to shrink the suspect list before you chase causes.

Is / is-not analysis is a problem-definition tool that bounds a problem by describing what it IS and what it could be but IS NOT, across four dimensions, what, where, when, and extent. Comparing the two columns exposes the distinctions that point to a cause, shrinking the suspect list before you start chasing why.

Most failed investigations fail at the start, not the end. A team hears "we have a quality problem" and immediately jumps to causes, brainstorming, blaming, and testing fixes for a problem nobody has actually defined. Is / is-not analysis, a core piece of the Kepner-Tregoe problem-analysis method, forces the discipline of describing the problem precisely first. By pinning down exactly what is and is not happening, you turn a vague complaint into a sharp specification, which is the foundation of any real root cause analysis and a habit worth building in any lean manufacturing operation.

What Is Is / Is-Not Analysis?

Is / is-not analysis is a structured way to specify a problem by contrasting where it appears with where it does not. For each dimension of the problem, you write two columns: the IS column captures the facts of where the problem shows up, and the IS-NOT column captures the closely related places, times, or conditions where you might reasonably expect the problem but it does not appear. The gap between the two columns is where the cause hides.

The key insight is that a problem's cause almost always lives in the differences between IS and IS-NOT. If line 3 has the defect and line 4, running the same part on the same shift, does not, the cause is something different about line 3. By building the two columns carefully, you convert a fuzzy problem into a short list of distinctive features that any true cause must explain. It is a filter you run before, not instead of, cause analysis.

Four dimensions down the side, two columns across the top. The IS-NOT column holds only plausible comparisons, things you would expect to be affected but are not.

Why Bound a Problem Before Chasing Causes?

Jumping straight to causes wastes the most expensive resource in an investigation: the team's attention. Without boundaries, brainstorming generates a cloud of possible causes, most irrelevant, and the team burns days testing theories that a two-minute look at the facts would have ruled out. A tightly bounded problem does the opposite: it hands you a small set of features the true cause must explain, so you can dismiss most candidate causes on sight.

Bounding also stops two classic errors. It stops the team from solving the wrong problem, because writing the IS column forces agreement on what the problem actually is. And it stops over-containment and over-reaction, because the IS-NOT column shows how limited the problem really is, which directly informs how wide an interim containment action needs to be. You cannot right-size a response to a problem you have not bounded.

What Are the Four Dimensions?

Kepner-Tregoe specifies a problem across four dimensions. Answering all four in both columns produces a complete picture.

Dimension	IS (the facts)	IS-NOT (could be, but isn't)
What	Which specific object and which specific defect	Similar objects that are fine; other defects not seen
Where	Where on the object, and where geographically (line, cell, plant)	Comparable locations that stay clean
When	When first seen, at what point in the cycle, which shift, trend over time	Times or points in the cycle it does not occur
Extent	How many units, how big the defect, how many per unit, the trend	How many are unaffected, the size it does not exceed

The discipline is filling in the IS-NOT column with genuine near-misses, not random opposites. A good IS-NOT is something you would reasonably expect to be affected.

How Do You Build an Is / Is-Not Table? A 6-Step Method

Write one clear problem statement. Name the object and the deviation in one line: "part 4471 shows a burr on the left flange." Vague statements ("quality is bad") cannot be specified. This single sentence anchors the whole grid.
Fill the IS column with facts, dimension by dimension. Work through what, where, when, and extent using observed data, not guesses. Go to the floor and the defect records rather than filling it from memory in a conference room.
Fill the IS-NOT column with plausible comparisons. For each IS, ask "what closely related thing could reasonably show this problem but does not?" Line 4 running the same part; the right flange; the day shift; the other supplier's lot. Only meaningful near-misses belong here.
Identify the distinctions. For each row, ask what is different or unique about the IS versus the IS-NOT. What does line 3 have that line 4 does not? These distinctions are the raw material for finding the cause.
Note any changes tied to those distinctions. For each distinction, ask what changed, and when, that could connect to it: a new die installed on line 3 last Tuesday, a coolant batch swapped, a fixture replaced. A distinction plus a recent change is a strong cause candidate.
Test candidate causes against every fact. A true cause must explain the entire grid, both the IS and the IS-NOT, without contradiction. Discard any cause that cannot explain why the problem is absent where the IS-NOT says it is absent.

The two columns act as a filter. Any cause that cannot explain why the problem is absent where the IS-NOT says it is absent gets dropped, leaving a short, testable list.

How Do Distinctions and Changes Point to the Cause?

This is the engine of the method. A distinction is anything true of the IS but not the IS-NOT: line 3 uses die #2, the affected units all ran after 2 p.m., the burr only appears on the left flange. A change is something that was modified in or around a distinction, and when. Root causes overwhelmingly live at the intersection: a distinctive feature where something recently changed. The burr appears only on line 3 (distinction), and line 3 got a reground die last Tuesday (change), and the first bad part shipped Wednesday (timing fits), so the die is the prime suspect.

Writing the IS-NOT column is what makes the distinctions visible. Without it, "line 3 uses die #2" is just a fact; with the IS-NOT showing line 4 uses die #5 and is clean, it becomes a lead. This is also why the method guards against premature blame: a favorite theory that cannot explain a single IS-NOT fact is disproven on the spot, no argument needed.

Root causes cluster where a distinction meets a recent change and the timing lines up. The grid surfaces all three so the suspect names itself.

How Does Is / Is-Not Fit With 5 Whys and Fishbone?

Is / is-not analysis is a front-end tool; it defines and bounds the problem so the cause-finding tools work on a sharp target. Run it first, then feed its output into whichever analysis fits. A fishbone diagram organizes candidate causes by category, and is / is-not tells you which branches are even worth exploring, because a cause category that cannot explain the IS-NOT is a dead branch. The 5 Whys then drills a confirmed cause down to its root. Used together, is / is-not narrows the field, fishbone organizes the survivors, and 5 Whys digs to the bottom.

Once the true cause is confirmed, the finding flows into your corrective and preventive action process, and the tight boundary you established pays off again: knowing exactly what is and is not affected keeps both the containment and the permanent fix correctly sized. A problem defined loosely gets fixed loosely.

Is / is-not analysis by the numbers

The method has a documented origin. Is / is-not analysis is part of the Problem Analysis process developed by Charles Kepner and Benjamin Tregoe, which specifies a problem across four dimensions, what, where, when, and extent (magnitude), and compares what the problem IS with what it IS NOT to expose distinctions that point to a cause (Kepner-Tregoe, Problem Analysis). The approach is a recognized problem-solving and quality tool taught alongside root-cause methods by bodies such as the American Society for Quality (ASQ, Root Cause Analysis). Its power comes from a simple rule: a true cause must explain every fact in the grid, both where the problem is and where, plausibly, it is not.

The whole method rests on accurate facts, exactly which units, which lines, which times, which lots. When that data lives on scattered paper and in memory, the IS and IS-NOT columns fill with guesses, and a grid built on guesses points at the wrong die. Plants that capture production, defect, and lot data live at the station can populate an is / is-not grid in minutes with facts they trust, and confirm the timing of a suspected change against the real record, which is the practical value of live floor data over your existing systems no rip-and-replace. See how tighter floor data sharpened problem-solving in our CLS case study.