Attachment style test (ECR-R)
See what an attachment style test actually measures, how it is scored, and how a result is read against the population that actually fits you.
The Experiences in Close Relationships - Revised is the research-standard self-report measure of adult attachment, the version most psychologists reach for when they study attachment style. Developed by Chris Fraley, Niels Waller and Kelly Brennan in 2000, it does not sort you into one of four boxes; it places you on two continuous dimensions, and the familiar styles emerge from where you land on them.
The model
What it measures
The ECR-R measures adult attachment on two near-independent dimensions, shown below: attachment anxiety and attachment avoidance. Each is the average of 18 statements about how you generally feel in close relationships, rated from strongly disagree to strongly agree. They are scored and interpreted separately - there is no single combined score.
The popular four "attachment styles" are regions of this two-dimensional space rather than separate scales. Low on both dimensions is secure; high anxiety with low avoidance is anxious-preoccupied; low anxiety with high avoidance is dismissing-avoidant; high on both is fearful-avoidant. Because the styles come from combining the two dimensions, a dimensional score is more precise than a single label. The aspects listed under each dimension are illustrative facets of how it shows up, not separately scored items.
- AXAttachment Anxiety
Worry about a partner's availability and responsiveness - fear of rejection or abandonment, and a strong need for reassurance.
Facets: Fear of abandonment, Worry the partner does not care enough, Need for reassurance, Sensitivity to a partner's moods.
- AVAttachment Avoidance
Discomfort with closeness, dependence and self-disclosure - a preference for independence and emotional distance.
Facets: Discomfort with closeness, Reluctance to depend on a partner, Difficulty opening up, Valuing self-reliance over support.
The evidence
Science and validity
The ECR-R was built by item response theory from a large item pool to measure each dimension precisely, especially near the secure end where older measures were weak. Internal consistency is high (Cronbach's alpha around .89 to .92 for anxiety and .91 to .95 for avoidance), and the two-factor structure replicates across many samples and translations. The dimensions predict relationship satisfaction, commitment, jealousy, conflict behaviour and how people seek and give support, and they map onto the Big Five in expected ways (anxiety to neuroticism, avoidance to lower agreeableness and extraversion) while still adding their own information.
You rate 36 statements about how you generally feel in close relationships on a 7-point agree-disagree scale. Fourteen items are reverse-keyed, then anxiety is the average of its 18 items and avoidance the average of the other 18, each on a 1-7 scale where the midpoint is 4. There are no clinical cut-offs and no pass/fail; a score is read against the midpoint and a comparison group, and the four styles are formed by crossing the two dimensions rather than by a separate test.
Where you stand
How a score becomes a percentile
A score only means something against a comparison group. The two dimensions run from 1 to 7, and in community samples both averages tend to fall around or a little below the midpoint of 4, with wide variation. For example, an attachment-anxiety average of 4.4 sits modestly above a typical comparison mean near 3.6, placing it around the 70th percentile - more relationship anxiety than most people in that group, while a low avoidance score keeps the overall pattern closer to the secure region. Drag the sliders to see how a score on each dimension maps to a percentile and to the four-style space; your real result is matched to the population that fits you when you take the test.
The reference data
Benchmarked against the population that fits you
We benchmark your result against the population that actually resembles you, across 20 reference groups.
Each reference group is used as its own benchmark, not to rank one country against another.
How it works
What the questions feel like
Illustrative statements showing the style of the items. These are examples, not the official scored items.
I often worry that a partner does not really care about me as much as I care about them.
Illustrative example in the style of the scale, not an official scored item.
I need a lot of reassurance that I am loved by my partner.
Illustrative example, not an official scored item.
I prefer not to let a partner see how I feel deep down.
Illustrative example, not an official scored item.
I find it easy to lean on a partner when I need support.
Illustrative reverse-keyed example, not an official scored item.
Honest strengths and limitations
Strengths
- The research-standard self-report of adult attachment, with very high reliability and a two-dimensional model that the popular four styles are built from.
- Open and freely available for research, with items and scoring posted by the authors and validated across 20+ languages.
- Item-response-theory refined for precision, especially near the secure end, where it measures more sharply than older attachment questionnaires.
Limitations
- It measures two dimensions, not four boxes - the popular styles (secure, anxious, avoidant, fearful) are regions formed by combining anxiety and avoidance, so forcing yourself into one category loses information.
- It asks about romantic relationships in general, so a result reflects a broad pattern rather than how you feel with one specific person, and attachment can shift with relationship events over time.
- The ECR-R has no single representative norm population; published means come from undergraduate and internet samples, so percentiles are a guide, not a verdict.
See your full profile
A complete report, matched to the population that fits you, with plain-language interpretation of every trait.
Frequently asked questions
What does the ECR-R attachment style test measure?
It measures adult attachment on two continuous dimensions: attachment anxiety (fear of abandonment and a need for reassurance) and attachment avoidance (discomfort with closeness and a preference for independence). The two are scored separately rather than as a single total.
How do the four attachment styles relate to the two dimensions?
The four styles are regions of the anxiety-avoidance space, not separate scales. Low on both is secure; high anxiety and low avoidance is anxious-preoccupied; low anxiety and high avoidance is dismissing-avoidant; high on both is fearful-avoidant. A dimensional score is more precise than a single style label.
How is the ECR-R scored?
You rate 36 statements on a 7-point agree-disagree scale. Fourteen items are reverse-keyed, then anxiety is the average of its 18 items and avoidance the average of the other 18, each on a 1-7 scale where 4 is the midpoint. There are no clinical cut-offs; scores are read against the midpoint and a comparison group.
Is the ECR-R free to use?
Yes. The items and scoring are openly available for research from the authors (Chris Fraley's measures site). On Psychology.me, the free Snapshot gives you a quick read on your attachment tendencies alongside other traits.
Related tests
- Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78(2), 350-365.
- Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult romantic attachment. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46-76). Guilford.
The Experiences in Close Relationships - Revised (ECR-R) is the work of R. Chris Fraley, Niels G. Waller and Kelly A. Brennan (2000), with items and scoring posted openly for research use; this independent informational page describes the instrument.