Sixteen statements, two minutes, no email. Scored on the two dimensions the research actually uses rather than sorted into a personality type — and with the limits of the result stated up front instead of buried.
Scored in your browser. Nothing you answer is sent anywhere or stored.
The four patterns this test can return are Secure, Anxious (preoccupied), Avoidant (dismissing), Fearful (disorganised).
Attachment research started with infants and their caregivers, and moved into adult romantic relationships with Hazan and Shaver in 1987. Their first measure asked people to pick one of three paragraphs. It worked well enough to open the field, and it had an obvious problem: people do not fall neatly into three paragraphs.
Brennan, Clark and Shaver settled it in 1998 by taking a pool of 482 items covering essentially every attachment concept in the literature and factor-analysing them. Two dimensions came out, and they have held up ever since:
The four familiar names are what you get when you cut both dimensions at the middle: low-low is secure, high anxiety with low avoidance is preoccupied, low anxiety with high avoidance is dismissing, and high on both is fearful. Bartholomew and Horowitz described that four-way map in 1991. It is a useful shorthand and it is still a shorthand.
Three failure modes, and you can spot all of them in about ten seconds:
The reason to care is that people make decisions with these results — about whether to stay, whether to explain themselves, whether to try. A label handed over with false confidence is worse than no label.
The useful move is not to adopt the label. It is to notice the trigger. Anxious patterns fire on a specific cue — the delayed reply, the changed tone — and avoidant ones fire on a different one, usually a bid for closeness arriving faster than expected. Once you can name your own cue, the gap between the cue and what you do about it is where the choice lives.
Two pieces on this site go deeper: telling love apart from anxious attachment and why intelligent people keep choosing the same wrong person. If you scored high on both dimensions, take the fearful description seriously — that is the pattern where professional help does the most work.
Secure, anxious (preoccupied), avoidant (dismissing) and fearful (disorganised). They are not four separate types — they are the four quadrants you get from crossing two continuous dimensions, attachment anxiety and attachment avoidance, as set out by Bartholomew and Horowitz in 1991.
No. The ECR by Brennan, Clark and Shaver has 36 validated items; this is a 16-item educational adaptation that follows the same two-dimension structure. It will point you in the right direction and it is not a clinical instrument. Anyone presenting a short quiz as the validated scale is overselling it.
Yes. Attachment patterns are relatively stable but not fixed. They shift with the relationship you are in, with major life events and with therapy. Research following people over years finds meaningful movement, which is why a result is a snapshot rather than a permanent category.
No. Every question is scored in your browser and nothing is transmitted or saved. There is no account, no email field and no results page with a unique URL. Close the tab and the answers are gone.
Then the label is close to arbitrary and you should read both neighbouring descriptions. The midpoint is 4 out of 7, so somebody at 3.9 and somebody at 4.1 get different names while being essentially the same. This is the cost of turning two continuous scales into four boxes.