Understanding your assessment

Understanding your assessment · 01

Reliability, validity and accuracy: what do they actually mean?

Research language should help you judge an assessment. Understanding three commonly confused terms makes the evidence easier to read and the claims easier to question.

5-minute read

The key idea

Reliability concerns the consistency and precision of scores. Validity concerns the evidence supporting a particular interpretation or use. Typing accuracy requires a credible basis for judging whether an assigned type is correct. These concepts cannot be reduced to one interchangeable percentage, which is why the iEQ9 research reports them separately.

Reliability: how consistent are the scores?

An assessment combines responses to describe something about a person. Reliability asks how consistently those scores behave. Different reliability measures look at different kinds of consistency.

Internal consistency concerns how coherently the questions within a scale work together. Test–retest evidence concerns agreement between results from separate sittings under controlled conditions. A scale can show strong internal consistency while its results vary over time; the two findings answer different questions.

In the 2026 iEQ9 study, ordinal alpha ranged from .74 to .87 across the nine type scales in five languages, with a mean of .80. That result describes internal consistency. [1] The iEQ9 has not yet had a controlled retest study; the research page sets out what is known about people who took it again.

A worked example: the Type 8 scale

The Active Controller scale has an ordinal alpha of .85.

What this tells you: the questions in that scale work together strongly. A person who agrees with one tends to agree with the others, and the score is not being driven by one or two odd questions.

What it does not tell you: that 85% of Eights are typed correctly, or that a person's score will be the same next year. Those are different questions and they need different evidence.

Validity: what interpretation does the evidence support?

Validity concerns whether evidence supports the meaning given to scores and the way those scores are used. Useful questions include whether the scales have the intended structure, how they relate to other measures, and whether an interpretation suits the people and the purpose.

Evidence for one interpretation does not automatically validate every use. Research about a questionnaire's scale structure does not establish that it predicts job performance. The testing standards published by AERA, APA and NCME ask for evidence that matches each intended use. [2]

The iEQ9 is intended for coaching, development and self-awareness, and its research should be read within that scope.

Accuracy: correct according to what?

A claim about typing accuracy needs a comparison: a way of knowing a person's type that does not depend on the questionnaire being tested. No such gold standard exists for the Enneagram, or for most of personality psychology. Motivation cannot be read from a blood test.

That is why a single accuracy percentage should be treated with caution, and why a reliability figure should never be read as one. What a publisher can do is gather several kinds of evidence that each bear on the question. [1]

How often people take the assessment again
In the 2026 study year, about 2% of people took the iEQ9 a second time. This is not a measure of accuracy, but a questionnaire that regularly gave people a type they did not recognise would see far more. Available now.
What practitioners find in the debrief
Accredited practitioners discuss each result with the person, and can ask for the assessment to be redone when the type does not fit. In the 2026 research, the cases practitioners flagged tended to be those where the profile itself was less clear-cut. Available now.
Agreement when people take it again
Among people who retook the iEQ9 and whose first result was clear-cut, 63.4% received the same type. This group chose to retake and does not represent iEQ9 users in general. A controlled retest study is planned.
Comparison with an independent typing
A study comparing iEQ9 results with types established independently, for example through structured interviews, has not yet been done.

The terms on the research page, in plain language

TermWhat it means
Ordinal alpha (α)How consistently the questions in one scale work together, from 0 to 1. A value of .70 is the usual minimum; .80 and above is strong.
Omega (ω)A second measure of the same thing, calculated from a statistical model of the scale. Read it like alpha.
Factor modelA test of whether people's answers group into the patterns the theory predicts. For the iEQ9, that is nine.
CFIHow well the factor model fits the data. A value of .90 and above is acceptable; .95 and above is good.
RMSEA and SRMRHow far the model misses the data. Lower is better, and values under .05 are good.
ICCHow closely two sets of scores from the same people agree, from 0 to 1.
Kappa (κ)How often two results put a person in the same category, after allowing for agreement by chance.
Big FiveThe most widely researched model of personality traits, used as a reference point for other measures.

Four questions to ask when reading a claim

  1. What was measured? Look for the named statistic and what it represents.
  2. Who was studied? Check the sample, language version and assessment version.
  3. What use is supported? Coaching, clinical diagnosis and employment decisions require different evidence.
  4. What are the limitations? Review status, sampling and self-report limits belong alongside the findings.

The most informative research summaries make these distinctions visible. They give you a way to understand the evidence instead of asking you to rely on a single impressive number.

Sources & scope

  1. Integrative Enneagram Solutions Limited. iEQ9 Statistical Manual, version 2026.07, reliability, temporal stability and limitations sections. See the public research summary for the study findings and research status.
  2. AERA, APA & NCME (2014). Standards for Educational and Psychological Testing. General guidance on score interpretation, validity and reliability.
Dirk Cloete
Author of the iEQ9

Dirk Cloete

Dirk Cloete is the founder of Integrative Enneagram Solutions and the author of the iEQ9. He trained in person with Dr Claudio Naranjo for three years, and has also trained with Cynthia Merchant, Dr Beatrice Chestnut, Ginger Lapid-Bogda, Dr Jerry Wagner, Russ Hudson, Kathleen Stinnett, Dr Lee Kingma, Dr R. Karl Hebenstreit and Dr Robert Barnes.

He has served on the Global Board of Directors of the International Enneagram Association (IEA) and has been published in the IEA journal. Read Dirk's full profile.

Continue your exploration

Read the iEQ9 research with its sample definitions, separate measures and stated limitations.

Explore the evidence

Explore your type with iEQ9 · Practitioner accreditation · Teams and organisations