Types vs Traits: Why Some Tests Sort People and Others Place Them

Three shapes, not two

A type sorts somebody into one of a fixed set of categories. Nine of these, sixteen of those. The result is a label, and the label is the whole answer.

A trait places somebody on a range between two ends. There is no category to belong to, only a position, and almost everybody is somewhere in the middle of several ranges at once.

An order is the third and it gets forgotten. Here the result is a list of things ranked within one person, from what they reach for first down to what they reach for least. There is no position on any of them and no category anywhere.

Why a type feels better and a range measures better

Types win on memorability by a distance. A person can say their number at dinner, find other people with the same one, and read a description written for them rather than about a position on a scale. None of that is available from a sentence about sitting toward one end of five ranges.

The cost is what happens at the boundaries. Sorting into categories means drawing lines, and real people are spread smoothly rather than clustered into groups. Somebody just on one side of a line gets a completely different label from somebody just on the other, despite answering almost identically, and both are handed the label with equal confidence.

That is also why type results are less stable on a retake than they look. Nothing about the person needs to change; a handful of answers near a boundary is enough to move the label, which then reads as a much bigger change than it was.

Why ranges are harder to read

A trait result asks more of a reader, which is the honest cost of the extra accuracy.

There is no label to carry, five separate answers rather than one, and the answer to most of them is somewhere in the middle. A well-built trait report has to spend real effort making a middling position feel like information rather than a shrug. On most traits, that is where the true answer sits for a great many readers.

It also has to name both ends in a way that neither one is the target. That sounds like a presentation detail and it is the whole ethical load of the format: a range with a good end is a scoreboard wearing a lab coat.

What an order can and cannot say

An order is the most modest of the three shapes and the most often misread. The list is ranked within one person, so the thing at the bottom is what that person reaches for least, not something they do not have.

Read correctly, that is a genuinely useful result and it needs no reference sample at all: a ranking of somebody's own answers against each other claims nothing about the population. Read as a scoreboard, it is a list of two dozen things somebody is apparently bad at, which is exactly wrong.

The visual conventions all argue for the wrong reading, which is why a report in this shape has to say what it is before it shows the list rather than after.

How to tell which shape is in front of a reader

The shape is usually obvious from the output, and worth checking before deciding how much weight a result deserves.

• A label, a number or a letter code, and a description written for it: a type.
• Several separate results, each a position between two named ends: traits.
• One ranked list, top to bottom, within the person: an order.
• A single overall score for something broad: usually several of the above added together, and worth asking what the sum is meant to mean.

None of the three is the right one

The shape is a choice about what a framework is trying to do, and each buys something the others pay for. Types buy memorability and pay in boundary cases. Ranges buy accuracy and pay in readability. Orders buy honesty without a reference sample and pay by being easy to misread.

What is worth refusing is a result that takes the license of one shape while presenting as another. A type derived by cutting ranges at a line, then reported as a category with no mention of how close the line was, is the common version and it is doing both jobs badly.

Common questions

Is the Enneagram a type system or a trait system?

A type system in how it is described, and in most instruments a set of ranges underneath with the top one reported as the type. That is why a careful Enneagram result says how close the runner-up was rather than handing over the number alone.

Why do so many popular tests use types?

Because a label is shareable and a position on five ranges is not. That is a real advantage for getting people interested and it is not an argument about accuracy.

Which shape is most stable on a retake?

Ranges, because a small movement stays a small movement. In a type system the same small movement either changes nothing or changes the label entirely, depending on where the person sat relative to a boundary.