The evidence
You should not have to become a psychometrician to buy an assessment.
The headline reliability and validity figures are below, along with how the instrument is built and tested. The documentation behind them exists in two forms, and both are available if you need them.
The numbers
Is DISC reliable? Here are the figures
Drawn from the Discflow 2.0 Technical Manual. The full tables, methodology and sample documentation sit in the manual itself.
Is it reliable? Internal consistency across the DISC scales runs from .73 to .87 (Cronbach's alpha, ten-item form), and Emotional Intelligence sub-domains from .76 to .86, with domain reliabilities of .91 for Self-Awareness and .92 for Awareness of Others.
Is it stable over time? Test-retest reliability across a four to six week interval runs from .72 to .86 across all scales, with composite scores at or above the .80 threshold for strong stability.
Is it valid? Confirmatory factor analysis supports the eight-style DISC structure (CFI .934, RMSEA .057) and the Emotional Intelligence framework (CFI .951, RMSEA .051). Against established measures, scores correlate with expected traits at r = .29 to .68 — related but distinct, which is the result you want.
Reliability, in full
Internal consistency across DISC scales ranges from .73 to .87 on the ten-item form. Emotional Intelligence sub-domains range from .76 to .86, with overall domain reliabilities of .91 for Self-Awareness and .92 for Awareness of Others.
Test-retest reliability across a four to six week interval (N = 218) ranges from .72 to .86 across all scales, with composite scores at or above the .80 threshold for strong stability.
Structural validity
Confirmatory factor analysis supports the eight-style DISC structure (CFI = .934, RMSEA = .057, SRMR = .063) and the hierarchical emotional intelligence framework (CFI = .951, RMSEA = .051, SRMR = .048).
Adjacent primary styles jointly account for 85% to 94% of variance in each blended style score, supporting the structural coherence of the blended-style framework.
Convergent validity
Compared with the Mini-IPIP6 and the TEIQue-SF on a sub-sample of 312 respondents, Discflow scores correlate with conceptually expected external traits at moderate magnitudes (r = .29 to .68).
That supports the position that the instrument measures related but distinct workplace-relevant constructs — which is the finding you want. Correlations near 1.0 would mean we had rebuilt an existing instrument.
- 994 — norming sample, international representation
- .73–.92 — internal consistency, Cronbach's α
- .72–.86 — test-retest stability, four to six weeks
What it is not
What it is not
Discflow is not a selection instrument and does not predict job performance. It describes how someone tends to operate and how readily they adapt, which is useful for development and useless as a forecast.
How it is built
Four things we do, every time
Items written and tested
Questions are drafted, reviewed and tested before they go anywhere near a live assessment. Items that do not behave are removed rather than explained away.
Reviewed by an expert panel
Content is reviewed independently by practitioners and subject experts, to check that each scale measures what it claims to and reads fairly across cultures.
Normed on a working population
Scores are interpreted against a sample of working adults rather than students, because that is who takes it.
Re-checked as the data grows
Reliability and structure are re-tested as the sample grows, and the Technical Manual is updated when the figures move.
What gets reported
Reliability, in the usual terms
Most people evaluating an assessment want one thing first: internal consistency, reported as Cronbach's alpha, for every scale.
We report it for every DISC scale and every Emotional Intelligence dimension, alongside the scale structure, the norming sample and the validity evidence behind each claim.
No instrument should ever be finished. We re-test as the sample grows, publish what is still open, and update the manual when the figures move. A manual that reads as though everything is settled is telling you something about the publisher rather than the instrument.
If you want more
Two levels of detail, and most people need the first
The evidence is documented twice: once for the people who buy and use assessments, and once for the people who check the workings.
Both are available to clients and to anyone seriously evaluating the instrument. Tell us what you are trying to establish and we will point you at whichever of the two answers it, rather than sending a manual to someone who needed a paragraph.
Start here
The plain-English guide
Written for people who use assessments rather than for psychologists. It covers what is measured, how the scoring works, what the reliability figures mean in ordinary terms, and what the instrument cannot tell you.
It is short, and it answers the questions that actually come up in a procurement conversation.
For almost everyone, it is the only document needed.
When it is genuinely needed
The full Technical Manual
The complete psychometric documentation: item development, scale structure, reliability coefficients for every scale, the norming sample and the validity evidence behind each claim.
It exists for in-house psychologists and assessment specialists who need to see the workings. If the guide has not settled the question, this will.
Next
Or judge it on the thing your people will actually read.
A sample report tells you more about whether this works for you than any coefficient will.