Are online attachment style tests reliable?
This guide does not tell you which style you have. It explains what a result can mean, and what it cannot, based on three studies whose abstracts we read.
1What does an attachment style test measure?
It measures how you describe your close relationships on the day you answer. It does not observe your behaviour, and it is not an examination. The most used research questionnaires, including the ECR-R, rate two dimensions: attachment anxiety (the “model of self”) and avoidance (the “model of others”), in the words of Sibley and colleagues in 2005.
A result is therefore a position on two scales, not the name of a category. The four labels you see everywhere (secure, anxious, avoidant, disorganised) are a convenient reading of those two scales. Our guide to attachment styles explains what each label does and does not describe.
2Are attachment styles boxes or sliders?
Closer to sliders. In 2015, Fraley and colleagues tested the question on two samples, using statistical methods built to tell real categories from real dimensions. Their analyses fit a dimensional model better than a categorical one. This held for attachment in general, and for attachment in specific relationships, such as with parents or with friends.
| If they were boxes | What the study observes | For you |
|---|---|---|
| You would fall into one style or another | Differences between people are continuous | Your result is a position, with degrees |
| Two people in the same box would be close | Differences between people are spread along a scale | The style name says less than the two scores |
| A cut-off would separate two profiles | These analyses do not show distinct categories | A result cut-off is a display choice |
The middle column is our reading of the abstract's conclusion, not a quotation. The study does not say that “secure” or “avoidant” are wrong words. It says dimensions describe the differences between people better.
3Will you get the same result next time?
With a well-built questionnaire, yes, in the short run. Sibley and colleagues found that the ECR-R gave highly stable indicators of underlying attachment over a three-week period, with 85% shared variance between the two sittings. A very different score three weeks later would be surprising with this kind of questionnaire.
That does not mean your attachment never moves. Fraley, Vicary, Brumbaugh and Roisman followed adults daily for 30 days (203 people) and weekly for a year (388 people). Their results fit best with a stable core underneath temporary swings. Today's answer therefore reflects the core, plus temporary swings.
- Reliable means the test gives roughly the same answer when you retake it. That is a quality of the questionnaire.
- Valid means it measures what it claims to measure. That is a different quality, and it is shown in a different way.
- A test can be very reliable and still say little about your real life.
4Does one result hold for all your relationships?
No, and this is the limit most often forgotten. In the 2005 study, the ECR-R explained between 30% and 40% of the differences between people in diary ratings of attachment-related emotions with a romantic partner. For interactions with family and friends it explained only 5% to 15%.
| Relationship described in the diary | Share explained |
|---|---|
| Romantic partner | 30% to 40% |
| Family and friends | 5% to 15% |
The questionnaire was designed for romantic relationships. A result says little about how you live your friendships or family ties. Fraley and colleagues, for their part, studied attachment in general and also in specific relationships (parents, friends): it can be described at both levels.
5How can you read your result without being boxed in?
- Read the two dimensions separately before reading a style name.
- Treat a gap of a few points as noise: on another day, the result could have moved a little.
- Ask which relationship you had in mind, and whether your answers would be the same for another one.
- Retake the questionnaire in a few weeks, at a calmer moment, and compare.
- If some situations keep coming back and weigh on your relationships, talking to a psychologist helps more than another test.
What this guide does not say
This guide relies on the abstracts of three research articles, not their full text, and reports only what those abstracts say. It covers adult attachment in close relationships, not childhood, and does not discuss links between attachment and disorders. The figures in these studies belong to their own samples and questionnaires, and do not carry over to a questionnaire of a different origin. The Brain Quiz questionnaire is written from the literature without reproducing any published scale, and its weights and result bands were not calibrated on a population: its methodology page says so.
Frequently asked questions
- Is an online attachment test reliable?
- It depends what reliable means. A well-built questionnaire such as the ECR-R gave very similar answers three weeks apart (85% shared variance in the 2005 study). That does not show it describes your real life well, especially outside romantic relationships.
- Can you get a different result when you retake the test?
- Yes, a little. The close follow-ups published in 2011 show a stable core under temporary swings. A gap of a few points is normal. A large gap can come from context, such as tiredness, a recent argument or a different relationship in mind, as much as from a change at the core.
- Do attachment styles really exist?
- The labels exist and describe recognisable tendencies. But a 2015 study on two samples found that differences between people fit continuous dimensions better than separate categories. A style is a handy way to summarise a position, not a box you step into.
- Does an attachment test apply to my friendships and family?
- Only in part. In the 2005 study, the ECR-R explained 30% to 40% of the differences between people for emotions felt with a romantic partner, against 5% to 15% with family and friends. It was designed for couple relationships, and that is where it is worth the most.
- What should I do with a result that bothers me?
- Reread it without the style name, looking at the two dimensions and at the relationship you had in mind. Then let it rest and retake it later. If situations keep coming back and weigh on your relationships, a psychologist can help you make sense of them, which a test cannot do.
Sources
- Fraley, R. C., Hudson, N. W., Heffernan, M. E., & Segal, N. (2015). Are adult attachment styles categorical or dimensional? A taxometric analysis of general and relationship-specific attachment orientations.. Journal of Personality and Social Psychology, 109(2), 354-368. DOI 10.1037/pspp0000027 (opens in a new tab)Deux échantillons : les différences entre personnes se décrivent mieux par des dimensions que par des catégories. Seul le résumé a pu être lu. Consulte le 2026-10-30.
- Sibley, C. G., Fischer, R., & Liu, J. H. (2005). Reliability and Validity of the Revised Experiences in Close Relationships (ECR-R) Self-Report Measure of Adult Romantic Attachment. Personality and Social Psychology Bulletin, 31(11), 1524-1536. DOI 10.1177/0146167205276865 (opens in a new tab)Trois études sur la stabilité sur trois semaines et sur la portée du questionnaire selon la relation. Seul le résumé a pu être lu. Consulte le 2026-10-30.
- Fraley, R. C., Vicary, A. M., Brumbaugh, C. C., & Roisman, G. I. (2011). Patterns of stability in adult attachment: An empirical test of two models of continuity and change.. Journal of Personality and Social Psychology, 101(5), 974-992. DOI 10.1037/a0024150 (opens in a new tab)Deux suivis rapprochés d'adultes (30 jours, un an) : un fond stable et des variations temporaires. Seul le résumé a pu être lu. Consulte le 2026-10-30.
Related questionnaires
These questionnaires cover the subject of this guide. Each one states where its statements come from and how it is scored.
Attachment style test
Free online attachment style test, no sign-up: 24 items, two dimensions computed separately, anxiety and avoidance, and an immediate result.