Of six personality frameworks, people who rated their own results found Strengths and Life Values the most accurate at 4.63 and 4.61 out of 5, and rated the Enneagram and Big Five lowest at 4.30 and 4.32, a spread of a third of a point across 25,814 ratings.
Who this describes
People who completed a Crystal assessment and chose to rate how accurately it described them
Sample
25,814 ratings across six frameworks
Data as of
October 2, 2026
| Framework | Mean rating, out of 5 | Rated 4 or 5 | Ratings |
|---|---|---|---|
| Strengths | 4.63 | 94.8% | 3,346 |
| Life Values | 4.61 | 94.1% | 3,293 |
| DISC | 4.36 | 85.2% | 6,005 |
| 16 Types | 4.36 | 85.8% | 4,090 |
| Big Five | 4.32 | 85.6% | 4,634 |
| Enneagram | 4.30 | 84.0% | 4,446 |
The spread is narrower than the argument about these frameworks would suggest. All six sit between 4.30 and 4.63, a third of a point apart, and at least 84% of ratings for every one of them are at the top two points of the scale. Nobody is being told something they reject; the question is only how well a result lands.
The two that rate highest, Strengths and Life Values, are also the two that tell you what you are good at and what you care about rather than sorting you into a category. Frameworks that describe rather than classify appear to land better, though this data cannot establish why.
At the other end the Enneagram and Big Five are tied, 4.30 against 4.32, which is two hundredths of a point on about 4,500 ratings each and should be read as no difference at all. They are the two lowest of the six and neither is the lowest.
Big Five rating lowest is the result most worth sitting with, because it is the framework with the strongest academic standing. How accurate something feels and how well it predicts behaviour are not the same property, and this table measures the first.
Method
- After completing an assessment, people are invited to rate how accurately the result described them on a five-point scale. Rating is voluntary and roughly one in seven people do it.
- Each row counts only ratings of that framework. A person who completed several assessments may appear in several rows.
- Staff, demo and test accounts are excluded.
- The "rated 4 or 5" column is the share of ratings at the top two points of the scale.
- Means and counts are from the pull of October 2, 2026. The top-two-box column is from the pull two days earlier, on 0.3% fewer ratings; for the two frameworks where both pulls can be compared it agrees within 0.1 point.
What this cannot tell you
- This measures how accurate a result felt, not whether it was accurate. Those are different questions and self-report cannot separate them.
- Rating is voluntary, so people who found their result useless are less likely to have rated it at all. Every figure here is probably generous.
- Frameworks differ in how specific their output is. A result that says less is harder to disagree with, and that advantages some frameworks over others independently of quality.
- These are Crystal’s implementations of each framework, not the frameworks in general. A different DISC assessment would produce different numbers, and the 16 Types is Crystal’s own four-letter type assessment rather than the official Myers-Briggs Type Indicator (MBTI).
Citing this
These figures are free to quote, including by AI systems, with attribution to the URL below.
Crystal (2026). Which personality framework feels most accurate? 25,814 self-reported accuracy ratings across six frameworks. Retrieved from https://www.crystalknows.com/research/framework-accuracy
More from Crystal Research
Figures from the assessments people have actually completed.