What an attachment style test actually measures

Almost all serious attachment tests for adults are based on the same idea. In the late 1990s, Kelly Brennan, Catherine Clark and Phillip Shaver pooled the questionnaires in use at the time, 323 questions in total, and showed that two basic factors lie underneath. One is attachment anxiety, the worry about being left or not being enough. The other is attachment avoidance, the discomfort with too much closeness and dependence. How the four familiar styles come out of this is explained in the article on attachment styles.

Keep readingThe four attachment styles, simply explained

So an attachment style test isn’t a personality test about everything under the sun. It gives you statements like “I often worry that the person I love will leave me someday” and asks how much they apply to you. Your answers produce a score on each of the two dimensions. That’s all it is, and that’s already quite a lot.

How well these questionnaires have been studied

Attachment questionnaires are among the most thoroughly researched measures in relationship psychology. Three findings matter for you.

The questions were tested one by one

In 2000, Chris Fraley, Niels Waller and Kelly Brennan examined the questions using a statistical method that shows how well each individual statement distinguishes between people. This led to a revised version, the “Experiences in Close Relationships – Revised” (ECR-R for short). Today it’s one of the most widely used measures in attachment research.

Scores are stable over a short time

A test that said something different every week would be worthless. Chris Sibley and James Liu had people fill out the ECR-R twice, six weeks apart. The scores for attachment anxiety and avoidance stayed very stable. So a good questionnaire doesn’t just measure the mood of the day.

There are no real types

In 2015, Chris Fraley and colleagues used a method designed for exactly this purpose to test whether people fall into clearly separate attachment types. The result: no. Attachment runs on a continuum, a bit like height. So a test that puts you in a box with no shades in between simplifies more than the research supports.

What a test can’t do

As well studied as the questionnaires are, they have limits. It’s worth knowing them before you take your result too seriously or too lightly.

It doesn’t make a diagnosis

A high score for attachment anxiety isn’t a disorder, and neither is a high score for avoidance. They’re common patterns. No attachment style test tells you whether you have an anxiety disorder, depression or trauma. There are professionals and other methods for that.

It measures what you know about yourself

A questionnaire can only capture what you notice about yourself and tick. In some people, though, self-image and experience diverge. Mary Dozier and Roger Kobak interviewed young adults about their childhood while measuring skin conductance, a sign of inner tension. Those who strongly downplayed their feelings in the conversation showed a clear rise at the sensitive questions. Someone who works this way may tick calmer answers on a questionnaire than their body would give.

Interview and questionnaire don’t measure the same thing

Besides questionnaires, research uses a detailed interview that goes back to Mary Main and colleagues. In it, adults talk about their childhood, and what’s assessed above all is how they tell it: coherently and in a balanced way, or rather patchy and entangled. Glenn Roisman and colleagues summarized the studies that used both methods with the same people. The overlap was small. Both methods predict something about relationships, but they partly capture different sides of attachment. So an online test shows you one side, not the whole picture.

It depends on the relationship you’re thinking of

Chris Fraley and colleagues developed a questionnaire that measures attachment separately for different people: mother, father, partner, best friend. The scores were related, but only moderately. Many people are fairly secure in one relationship and noticeably less so in another. Jennifer La Guardia and colleagues also found that the same people felt differently secure with different people, depending on how much that person supported their basic needs. If you take the test in the middle of a difficult relationship, your result will look different than after three calm years with a reliable person.

It’s a snapshot

Over six weeks the scores are stable; over years they can shift. In 2002, Chris Fraley summarized that early experiences form a basic pattern, and new experiences keep being layered on top of it. William Chopik and colleagues found in three long-term studies that attachment anxiety decreases on average over the course of adult life. So your result says where you are today, not where you have to stay.

How our test is built

Since we recommend honesty to others, it applies to us too. The free test on this website works like this:

  • 16 statements, 8 on attachment anxiety and 8 on avoidance. You answer each one on a scale from 1 to 5.
  • Some statements are reverse-worded, for example “I rarely worry about being left.” That prevents someone from simply clicking the same number everywhere.
  • The average is calculated for each dimension. From that, we show you what shares of the four styles are in your profile, instead of putting you into a single type. That fits with the finding that attachment runs on a continuum.
  • The statements are our own wording, based on the two-dimensional model by Brennan, Clark and Shaver and on the ECR-R by Fraley, Waller and Brennan.

What our test isn’t: it isn’t the ECR-R, it’s considerably shorter than research questionnaires, and it hasn’t been tested in a scientific study of its own. It’s an honest starting point that shows you a first direction. The stability figures from research can’t simply be transferred to it.

How to read your result sensibly

1. Look at both dimensions, not just the name

“Anxious” can be just above the middle or right at the top. And someone who scores high on attachment anxiety can score low or also high on avoidance. The two scores tell you more than the label. And even the lowest scores don’t mean “perfect”: what secure attachment feels like in everyday life is described on the page about the secure attachment style.

2. Check it against real moments

Think back over the last few weeks. What happened when a reply didn’t come? When someone wanted to get very close to you? After the last fight? If your experience and your result fit together well, you can trust the result more. If they don’t, that’s important information too.

3. Think of a specific relationship

Take the test a second time in your head, this time with a different person in mind, for example a close friend instead of your partner. Many people notice that they aren’t the same everywhere.

4. Repeat it after some time

After a few months, after a breakup or after a period in which you’ve worked a lot on yourself, a fresh look is worthwhile. Small shifts are normal and often encouraging.

5. Don’t use it as a label for others

“Classifying” a partner, an ex or a friend from a distance is tempting and usually unfair. A test describes how someone experiences themselves. If you want to know what it looks like inside another person, only one thing helps: ask.

When a test isn’t enough

A questionnaire can give you a language for your pattern. It can’t replace treatment. If your relationships weigh heavily on you, if anxiety or low mood shapes your everyday life, or if difficult experiences from your past keep coming up, it’s worth talking to a professional. In Switzerland, you can find psychotherapists through psychologie.ch, for example. Since 2022, basic health insurance has covered psychological psychotherapy when it’s prescribed by a doctor. If you’re in a very bad place right now: in Switzerland, Die Dargebotene Hand (143) is there around the clock. Elsewhere, contact your local crisis line or emergency number.