What an attachment test actually is
Modern adult attachment tests are self-report questionnaires that measure two underlying dimensions: attachment anxiety (how worried you are about abandonment, rejection, or insufficient closeness) and attachment avoidance (how uncomfortable you are with intimacy, dependence, and emotional exposure). Your scores on those two dimensions map onto the familiar four-style picture: secure (low on both), anxious (high on anxiety, low on avoidance), avoidant (low on anxiety, high on avoidance), and disorganized or fearful-avoidant (high on both).
The important thing to notice: most people are not pure types. The dimensions are continuous. You can be moderately anxious and slightly avoidant. You can shift between styles depending on the relationship. The label is a shorthand for a position on the map, not a fixed identity.
The ECR-R
The Experiences in Close Relationships-Revised (ECR-R) is the most widely used research measure of adult attachment. Developed by Fraley, Waller, and Brennan, it’s a 36-item self-report — 18 anxiety items, 18 avoidance items — rated on a 7-point agreement scale. Research labs have used it in hundreds of studies, and it has strong reliability and well-documented validity.
R. Chris Fraley’s lab at the University of Illinois maintains a free online version at yourpersonality.net. It takes about 10–15 minutes and produces dimensional scores you can use to read where you sit on both axes. This is the one to take if you want a measure with real academic grounding.
Earlier variants like the ECR and the AAS (Adult Attachment Scale) are still in circulation; they measure the same underlying constructs with slightly different wording. Any of them gives a useful picture.
Honest limits of self-report
- You see what you’re willing to see. Avoidantly attached people, in particular, often under-report avoidance on self-report measures — the very defense that keeps distance feeling okay also keeps it invisible.
- Context matters. Your attachment can show up differently with a partner versus a parent versus a boss. Most tests ask about close relationships broadly, which averages out some real variation.
- Self-report vs. interview. Clinical attachment measurement uses the Adult Attachment Interview — a structured narrative interview rated by trained coders — which often produces different classifications than self-report. Both are useful; they’re not interchangeable.
- State, not only trait. A recent breakup, a stressful period, or a new relationship can push your scores around. Take the test a couple of times, months apart, for a more honest picture.
How to take it honestly
Answer for how you feel in close relationships generally, not for a current fight or a recent specific partner. If a question makes you uncomfortable, that’s useful information — pick the answer that’s true rather than the one that sounds better.
When you look at your scores, resist the urge to pick the nearest label and stop there. The most useful read: what do my scores tell me about the specific kinds of situations that activate my system? A high anxiety score means reassurance-seeking dynamics; a high avoidance score means closeness-escape dynamics. Look at your actual relationships with that lens on.
What the research supports
Attachment measures have predicted, across many studies, relationship satisfaction, breakup likelihood, parenting patterns, response to stress, and even physical health markers. The effect sizes are usually moderate — real, but not huge. Attachment is one lens on relationships, not the whole story. Taken alongside personality measures (Big Five), your picture becomes more honest.
After the test
Scores turn into understanding when you read the styles themselves: secure, anxious, avoidant, disorganized, and the hopeful earned-security pattern. And — no label defines you. Return anytime to the scientific overview.