Personality & Individual Differences · Simplified

Same Average, Six Very Different Weeks

One number on a form, three lives it does not describe.

What this will not do. It asks you nothing about yourself, offers no questionnaire and produces no score for anybody. The three people are invented and the scale is arbitrary. Low or variable values here are characteristics that differ between people, not symptoms, and nothing on this page is a clinical statement, a screening instrument or a basis for interpreting anyone's wellbeing, your own included.

Step 1 of 2

First, a prediction

Three fictional people complete a self-esteem measure once, on the same afternoon. All three score exactly the same. What does that one score tell you about their next six weeks?

Step 2 of 2

Six weeks, one reading a day

Here are the same three people, measured once a day for six weeks. With nothing happening to any of them, their six-week averages are identical, and two of the three look much alike. Then deliver something.

Key terms

Level: Where somebody's self-esteem typically sits. It is what a one-off measure estimates, and here all three share it.
Variability: How much it moves from day to day with nothing in particular happening. It is a characteristic of the person and independent of their level.
Contingency: Which domains it moves in response to. Two people can be equally variable and contingent on quite different things.
Recovery: How quickly it returns after a knock. A third property, implied by neither of the other two.

One line per person, one point per day. An upright marker shows the day something was delivered, and the same thing is delivered to all three of them at once.


Person	Six-week average	Ordinary day-to-day movement	Largest drop after an event	Days until back within two points

What this shows

Four things, collapsed into one number

Key idea: A single self-esteem score estimates a level, and these three share it exactly. Everything else that differs between them is invisible to it, and there are at least three such things. How much somebody moves from day to day with nothing happening is one, and it is why Bo's six weeks look nothing like the others'. What moves them is a second, and it is quite separate: Cleo is steadier than Bo day to day and is the only one for whom a setback at work is a catastrophe, while a falling-out with a friend barely registers. How long it lasts is a third, and it is separate again: the identical work setback lands far harder on Cleo than on Ada, because work matters so much more to her, and her much slower recovery means Ada is back within days while Cleo is still down six weeks later. None of these is a fault or a symptom. They are characteristics, they differ between people, and a measure that reports one number per person has already decided not to see them.

Three invented people whose days are produced by a formula, with a fixed random seed so that the same six weeks appear every time. The model is deliberately tidy in ways life is not: events arrive when a button is pressed rather than in response to how somebody is already doing, recovery is a smooth decay towards exactly where the person started, and nothing anybody does changes their own contingencies. Real self-esteem responds to events that are partly consequences of earlier self-esteem, which this cannot represent at all. Two things to be careful of in the reading. Variability is a characteristic and not a symptom: this page is not a description of anybody and cannot be used to interpret anyone's wellbeing. And the domains here are two invented labels rather than a taxonomy; what a person's self-esteem is contingent on is a question for research and not something to be guessed from a chart.

The longer version adds a fourth person, lets you choose which day an event lands on and deliver several, distinguishes positive from negative events, and separates contingency across more domains. It is at Self-Esteem Stability Tracker in the main collection.