Is the Enneagram scientifically valid?

The evidence is mixed. Some parts of the Enneagram have partial research support. The central claim — that people sort into exactly nine types — has not been confirmed. Two of its secondary ideas, wings and arrows, have little research behind them.

That puts it between two popular positions. Research has not proved the Enneagram, and it has not shown it to be meaningless. It is a model with some measurable signal, a debated structure, and large untested areas.

The most complete summary is a systematic review by Joshua Hook, Todd Hall, Don Davis, Daryl Van Tongeren and Mackenzie Conner: "The Enneagram: A systematic review of the literature and directions for future research," Journal of Clinical Psychology, 77(4), 865–883 (2021). DOI: 10.1002/jclp.23097.

The authors wrote it for clinicians, because many clients use the Enneagram to understand themselves. They reviewed 104 independent samples and summarised the result in one phrase: mixed evidence of reliability and validity.

Reliability asks whether a measure gives consistent results. Validity asks whether it measures what it claims to. "Mixed" means some studies found support on each, and others did not.

Partial structure. Hook and colleagues report that some factor-analytic studies — statistical analyses of how test answers group together — show partial alignment with Enneagram theory.

Links to established measures. Enneagram subscales relate to other constructs, including the Big Five personality traits, in ways the theory predicts.

Usefulness for growth. Several studies in the review found the Enneagram helpful for personal and spiritual growth.

One growth study: Daniels, Saracino, Fraley, Christian and Pardo (2018), Journal of Adult Development, 25(4), 229–241, DOI 10.1007/s10804-018-9289-x. It measured ego development in 122 people enrolled in Narrative Tradition trainings, 18 months apart, and found gains among some participants with at least 40–50 hours of training. Its participants were already enrolled in Enneagram courses. Both groups studied the Enneagram; there was no comparison group that didn't, so the gains can't be separated from 18 months of ordinary adult development. David Daniels co-founded the Narrative Tradition.

An early peer-reviewed study is Wagner and Walker (1983), Journal of Clinical Psychology, 39(5), 712–717, DOI record. Its 390 participants typed themselves, and their typing was stable. On two other personality inventories, the nine groups showed significantly different profiles that the authors describe as congruent with the types. The sample was people who already knew the system well enough to self-type. Jerome Wagner later published a commercial Enneagram inventory (*WEPSS: Wagner Enneagram Personality Style Scales*, Western Psychological Services, 1999).

Fewer than nine factors. When researchers factor-analyse Enneagram test data, they have typically found fewer than nine factors (Hook et al., 2021). Answers don't separate cleanly into nine dimensions.

No study has derived the types from data (as of the 2021 review). Hook and colleagues found no study that used clustering methods — statistics that look for natural groups of people — to see whether nine types emerge on their own. So far the tests assume nine types; the data has not been used to find them independently.

Wings. The review found little research supporting wings. An earlier study tested one prediction drawn from wing theory: if each type shares traits with its neighbours, people reading brief descriptions of the nine types should judge adjacent types the most similar. That hypothesis was not supported (Edwards, 1991, Social Behavior and Personality, 19(1), 11–20, DOI 10.2224/sbp.1991.19.1.11). The study measured how people perceived written descriptions, not the traits of people of each type.

Arrows. The review also found little support for intertype movement — the stress and growth lines.

The library's wings and arrows pages say the same: both are traditional lenses for reflection, with thin research backing.

Subtypes and instincts. The instinct and subtype framework is central to modern Enneagram teaching. This library has not found a peer-reviewed test of it to cite.

Typing methods. Tests, panels, interviews and self-typing all produce a type. None of them has an independent answer key to be checked against, because a person's type is inferred from what they report, never measured directly. Consistency can still be studied, such as whether a method gives the same result twice or whether two typers agree, but this library has not found studies comparing methods that way. The test or interview page compares the routes.

Who has been studied. Two of the studies above drew on people who had already chosen to learn the Enneagram: self-typers who knew the system, and people enrolled in trainings. Results from enthusiasts may not carry over to everyone else.

Hook and colleagues close their review with a list of directions for future research.

A type result from any method is a hypothesis about you. Treat it that way.

• Read the top two or three. Because test data doesn't split cleanly into nine, the gap between your first and second score may be small. Read the pages for each.
• Check it against remembered moments. Pick a real situation where you acted in character. Ask why you did it. A type fits when its motivation explains the moment (see look-alikes).
• Hold wings and arrows lightly. They are the least supported parts of the model. Use them as prompts for reflection.
• Watch for flattering fits. Self-typing is biased toward self-image (see how typing is done). Ask whether you're choosing the description you'd like to be.
• Expect to revise. Settling on a type can take time. Changing your mind after more reflection is normal.

Use the Enneagram as a reflective tool: a vocabulary for asking why you do what you do. The research supports using it for reflection. It does not support treating a type as settled fact.

It is not a clinical or diagnostic instrument, and it should not be used for hiring or selection (see what it is not). For the history and the rest of the sources this library draws on, see sources and lineage.