There is no emotional quotient

The expression emotional intelligence appears in the scientific literature in 1990, from the pens of Peter Salovey and John Mayer. It then designates a set of four abilities: perceiving emotions, using them, understanding them, regulating them.

The abbreviation EQ, for its part, was born of the popularisation that followed the publication in 1995 of Daniel Goleman's book, and of the magazine cover that came with it. Its form deliberately imitates that of the intelligence quotient, and that imitation contains the promise of a measure of the same nature.

An intelligence quotient is a score standardised on a reference population, with a conventional mean and standard deviation, obtained through tasks with correct answers, and accompanied by a known margin of error. Each of these conditions took decades of psychometric work.

The best-selling emotional intelligence instruments do not meet these conditions. Most are self-report questionnaires, with no correct answer, normed on rarely described samples, and they aggregate dimensions that do not vary together.

Since the early 2000s, following Konstantinos Petrides and Adrian Furnham, the literature has distinguished emotional intelligence as ability and as trait. The first is measured by performance tasks, the second by self-report.

The correlations between the two families of instruments are weak, generally below 0.30. A person may judge themselves very at ease and obtain a mediocre score on the task, and vice versa. Two tests bearing the same name can therefore rank the same people in an almost different order.

The most concrete problem is that of the single score. Adding the capacity to perceive, the capacity to contain and the disposition to say what does not suit presupposes that these three things go in the same direction and that a high total is worth more than a low one.

Yet the data show nothing of the sort. A very perceptive and unassertive person, and an unperceptive and very composed person, would obtain similar totals while functioning in opposite ways. The number erases exactly what was interesting.

It remains possible, and useful, to describe positions on separate dimensions, saying where they come from, on what sample they are reported, and with what margin. That is more modest than a single figure, and it does not lend itself to being read as a level.

That is the choice made by this publisher's emotional profile: six dimensions, no total, no general ranking. An axis whose two ends describe ways of functioning has no top and no bottom, and it would be dishonest to pretend otherwise in order to produce a more saleable score.