The four names — secure, anxious, avoidant, fearful-avoidant — are the popular version of a measurement, and the measurement is more useful than the names. Underneath the categories are two independent dials, and almost everything people want to know about themselves is in where those two dials sit rather than in which of four boxes they land closest to.
So the honest answer to the question has two halves. There is a lot you can read off your own behaviour tonight, without any instrument at all, and there is a specific place where reading yourself stops working. Both halves are below.
The two dials underneath the four names
Attachment anxiety is how much distance costs you. High on this dial, a delay is not a delay; it is a change in the weather that has to be resolved before anything else can happen. You notice tone. You reread. You are quick to repair, quick to apologise, and quick to assume the problem is you.
Attachment avoidance is how much closeness costs you. High here, being depended on feels like pressure and being known feels like exposure. You handle things yourself, not out of pride but because handing something over has a price you can feel. You are fine — genuinely fine — and nobody can find the door.
The four names are just the quadrants. Low on both is secure, which is the majority of people and is less interesting than it sounds — it mostly means conflict is unpleasant rather than dangerous. High anxiety with low avoidance is anxious. Low anxiety with high avoidance is avoidant. High on both is fearful-avoidant, the smallest group and the one the category names describe worst, because a person who wants closeness badly and cannot tolerate it does not behave like a blend of the other two. They behave like someone with two feet on two pedals.
This is why "what is my attachment style" is slightly the wrong question. Two dials, not one box.
How to read yourself tonight
Five observations, each of which is worth more than an hour of introspection, because they are about what you do rather than what you believe about yourself.
- The twenty-four hours after a conflict. Do you pursue, or do you go level and quiet? Whichever one you do, you do it automatically, and that is the single most diagnostic thing about you.
- What you do with good news. Who do you tell, and how fast? People high on avoidance often notice they enjoyed something for a whole day before it occurred to them to tell anyone. That is not coldness; it is a habit of self-containment, and it shows up here more clearly than in an argument.
- The unanswered message. Sit with it for an hour and watch what your mind does. Does it build a case? Does it move on so completely that you forget you sent it? Both are answers.
- How you talk about your exes. This is the one the research instruments actually care about most. A flat, brief, generous-sounding summary with no detail in it, or a long account that is still hot years later, both say more than the content does.
- What happens when someone gets closer than usual. Not worse — closer. A weekend that went too well, a compliment that landed. Notice whether the next few days are lighter or heavier.
Where reading yourself stops working
Two specific failures, and they are worth knowing before you trust your own answer.
The first is that avoidance is quiet by construction. It does not produce a feeling you can catch, because the entire function of it is to keep feelings below the level where they would demand something. People high on avoidance frequently report themselves as secure, and they are not lying — from inside, low distress is indistinguishable from low distress.
The second is that your current partner is a confound. Attachment is a tendency, not a constant, and it is partly relationship-specific: the same person can be noticeably more secure with a steady partner and noticeably more anxious with an erratic one. If you are currently with someone whose attention is unpredictable, almost anyone would score high on anxiety, and that score is describing a situation rather than a trait. Answer for how you are across relationships, not for this month. How much of any reading belongs to the relationship rather than to the person has been measured directly, and the answer is larger than the four names suggest: why security shows up with one person and not another.
It is also worth knowing that the two main research traditions do not agree with each other as much as you would expect. The interview method scores how coherently an adult can narrate their own childhood; the questionnaire method scores the two dials directly from how you say you behave. The two barely correlate. Neither is wrong; they are measuring different layers, and only the second is what any online test — including this one — is giving you.
What it depends on
If your two dials are both low or clearly one-sided, you can probably name your style from the five observations above and skip everything else. The reason to measure is the case where it is not clean: high on both, or high on one and uncertain about the other, or a reading that changes depending on which relationship you think about while answering.
That case is more common than the neat quadrant diagrams suggest, and it is the case where the advice diverges most — the corrections for anxious and avoidant contradict each other, so getting the quadrant wrong means doing the opposite of what would help. Twelve questions, scored on both dials, is a better instrument for that than an honest afternoon of thinking about yourself. And once you have a reading, the next question is usually whether it is permanent, which has its own answer.