Why two EQ tests give two scores

Three sources of disagreement

First, the construct. Some instruments measure trait emotional intelligence, your typical self-described tendencies. Others measure ability, your performance on scored problems about emotion. The two kinds correlate surprisingly weakly, so a high score on one and a middling score on the other is not a contradiction; it is the finding.

Second, the norm sample. Any score expressed relative to a comparison group inherits that group. The same answers land differently against different samples, and no two tests share one.

Third, the scale. A 78 as a percentage of maximum, a 78 on a mean-100 scale and a 78th percentile are three unrelated claims that happen to share digits. The scales are catalogued here.

What to compare instead of totals

Shapes, not numbers. If two tests both say your self-awareness outruns your composure, that agreement across instruments is worth more than either total, because it survived two different sets of questions, samples and scales. Profile agreement is the closest thing self-report testing has to replication.

And compare each test’s honesty about itself: does it publish its method, its margin, and who is in its norm sample? A test that shows its working can be checked. A test that only shows a number cannot.

How we handle it

We publish every score as a percentage of its own maximum with its margin printed beside it, and the method is public. The test is free to take; the free result shows your classification and the shape of your profile, and the full scored report, every figure with its margin, is available as a one-off purchase or inside membership.

The rest of this family

Every page here does the same job for a different number: the full set is indexed on the hub.