Psychology World — home
Blog

The four attachment styles: what each one means, and where the names came from

Secure, anxious, avoidant, fearful-avoidant: what the four attachment styles look like in adults, the two scales behind them, and where the names came from.

Development13 min read
By Psychology World

Updated 21 November 2026

The four attachment styles in adult relationships are secure, anxious (also called preoccupied), avoidant (also called dismissing) and fearful-avoidant. Bartholomew and Horowitz set out this four-part model in 1991. They defined each style by two judgments: a positive or negative image of yourself, and a positive or negative image of other people.

Today most researchers measure the same ground with two scales: attachment-related anxiety and attachment-related avoidance. A secure adult is low on both. The style names are older than the adult research. They come from Mary Ainsworth's studies of one-year-olds, which found secure, avoidant and resistant patterns. Main and Solomon added a fourth infant pattern, disorganized, in 1990.

  • The infant patterns and the adult styles come from two different research traditions. That is why most styles have two or three names.
  • Adult attachment is a position on two continuous scales, not a box. The four styles are the four corners of that space.
  • Disorganized and fearful-avoidant are often used as one term. They are not the same thing: one is an infant observation code, the other an adult self-report category.
  • Interviews and questionnaires about adult attachment barely agree. A 2007 meta-analysis found a mean correlation of .09.
  • Attachment security is moderately stable over time, and it can change. A 2013 meta-analysis found no significant stability over intervals longer than 15 years.

The four attachment styles at a glance

Each adult style is a combination of a high or low score on the two scales. Attachment-related anxiety is worry about whether a partner is available and responsive. Attachment-related avoidance is a preference not to rely on other people or open up to them, according to R. Chris Fraley's overview of adult attachment research.

The four adult attachment styles, their other names, and where each sits on the two scales.
StyleOther namesAnxietyAvoidanceIn Hazan and Shaver's three styles (1987)
SecureSame name in every modelLowLowSecure
AnxiousPreoccupiedHighLowAnxious-ambivalent
AvoidantDismissing, dismissing-avoidantLowHighNo match: Bartholomew added it
Fearful-avoidantFearfulHighHighAvoidant

The table maps four labels onto two scales, and the figure shows the same map. Each corner is a region, not a point. Two people with the same label can still sit far apart inside it.

A square with attachment avoidance on the horizontal axis and attachment anxiety on the vertical axis, divided into four regions: secure, anxious, avoidant and fearful-avoidant.
The four attachment styles are regions on two scales: attachment anxiety and attachment avoidance.

What each attachment style looks like

The descriptions below follow the two scales. The example statements come from the ECR-R, a 36-item questionnaire that measures them. Each statement is rated for agreement, so a statement only shows the direction of a scale, not a verdict about a person.

Secure

A secure adult is low on both scales. Fraley describes people at the low end of anxiety as "more secure in the perceived responsiveness of their partners". People at the low end of avoidance are comfortable being close to others and depending on them. Secure is the most common style in the research. In Hazan and Shaver's first adult study, about 60 percent of people chose the secure description.

Anxious (preoccupied)

An anxious adult is high on anxiety and low on avoidance. The person wants closeness but worries that the partner is not available or does not care enough. Two ECR-R anxiety statements show the pattern. One is "I'm afraid that I will lose my partner's love". The other is "I often worry that my partner will not want to stay with me". In Bartholomew's model, this style combines a negative image of the self with a positive image of others.

Avoidant (dismissing)

An avoidant adult is low on anxiety and high on avoidance. The person prefers not to rely on others or open up to them. Two ECR-R avoidance statements show the pattern. One is "I prefer not to show a partner how I feel deep down". The other is "I find it difficult to allow myself to depend on romantic partners". Bartholomew took this kind of avoidance from a category of the Adult Attachment Interview, and paired a positive self-image with a negative image of others.

Fearful-avoidant

A fearful-avoidant adult is high on both scales. The person worries about whether a partner is available and responsive, and also prefers not to rely on others or open up to them. In Bartholomew's model, this style combines a negative image of the self with a negative image of others. This is the trap in the names. The "avoidant" style of Hazan and Shaver's 1987 study corresponds to Bartholomew's fearful style, not to her dismissing one. Shaver and Fraley's guide to the self-report measures gives this mapping.

Where the names came from: Bowlby and the Strange Situation

John Bowlby, a British psychoanalyst, started attachment theory. He wanted to understand the intense distress of infants who were separated from their parents. He proposed an attachment behavioral system, shaped by natural selection, that keeps an infant close to an attachment figure. In Fraley's summary, the system asks one question: is the attachment figure nearby, accessible and attentive? From repeated experience with a caregiver, the child builds expectations about relationships. Bowlby called them working models.

Mary Ainsworth turned the theory into a measurement, the Strange Situation. A one-year-old, the caregiver and a friendly stranger move through short episodes of about three minutes in a playroom, with two separations and two reunions. The classification depends on the infant's response to separation and especially to reunion, according to a 2023 meta-analysis of the procedure.

Ainsworth's 1978 book named three patterns, with letter codes still in use:

  • Secure (B): the infant is upset when the parent leaves, actively seeks the parent at reunion, and is easily comforted.
  • Avoidant (A): the infant does not appear very distressed by the separation and avoids contact at reunion, sometimes turning to the toys.
  • Resistant or ambivalent (C): the infant is very distressed by the separation, is hard to soothe at reunion, and shows conflicting behavior: it wants comfort and also seems to "punish" the parent.

The avoidant infant is not unbothered. Spangler and Grossmann (1993) found that avoidant infants showed less distress but heart-rate arousal similar to secure infants during separation. Cortisol also rose in avoidant and disorganized infants.

Some infants did not fit any of the three patterns. Main and Solomon reviewed 200 of these hard-to-classify procedures and, in 1990, proposed a fourth code: disorganized or disoriented (D). It applies when an infant shows conflicted, confused or apprehensive behavior toward the caregiver. Because these behaviors are often brief, a D infant also gets one of the three organized codes. The 2023 meta-analysis links the pattern to a caregiver who is "also, at times, a source of alarm".

How common each infant pattern is

Madigan and colleagues' 2023 meta-analysis in Psychological Bulletin pooled 285 studies with 20,720 parent-child pairs.

Global distribution of infant Strange Situation classifications, Madigan et al. (2023).
PatternCodeShare
SecureB51.6 percent
DisorganizedD23.5 percent
AvoidantA14.7 percent
ResistantC10.2 percent

The meta-analysis found no difference between mothers and fathers, and no difference by the child's age or sex. Disorganized attachment was more common in samples with maltreatment, parental mental illness, foster care or adoption from foster or institutional care. That is a rate for groups, not a label for a child. The older "standard" American distribution, from Ainsworth's 1978 data, was about 70 percent secure, 20 percent avoidant and 10 percent resistant.

From infants to adults: the love quiz and the four-category model

The adult styles start with Hazan and Shaver (1987). They took Ainsworth's three infant patterns and wrote three short paragraphs about how an adult with each pattern might feel in romantic relationships. Readers of a Denver newspaper took this "love quiz" and picked the paragraph that fitted them best. About 60 percent chose secure, about 20 percent avoidant and about 20 percent anxious-resistant, a distribution close to the infant one.

Bartholomew and Horowitz (1991) changed three categories into four. They defined each by a positive or negative self-image and a positive or negative image of others. The result was secure, preoccupied, dismissing and fearful. The terms "fearful-avoidant" and "dismissing-avoidant" come from this framework.

A third tradition runs beside these two. The Adult Attachment Interview, from George, Kaplan and Main (1985), asks adults about their childhood relationships with their parents. Its scoring looks at how coherently a person tells the story, not at whether the childhood was good or bad.

Is disorganized attachment the same as fearful-avoidant?

No, although many websites use the two terms for one style. They come from different traditions and are measured in different ways.

  • Disorganized is an infant code. Observers give it in the Strange Situation when an infant shows conflicted, confused or apprehensive behavior toward the caregiver. The infant also gets an organized code: A, B or C.
  • Fearful-avoidant is an adult self-report category. It means high anxiety and high avoidance on a questionnaire, or the matching paragraph on Bartholomew's measure.

Both labels describe insecure attachment, which is one reason people merge them. But adult research on disorganization needed its own tool. In 2015, Paetzold, Rholes and Kohn wrote that disorganization had not been studied from a social psychological view in adults, and they built a separate scale to measure it. A 2021 study by Paetzold and Rholes describes adult disorganized attachment as a construct that "can now be measured on a self-report scale".

Styles or scales? Why researchers use two dimensions

The four boxes are easy to remember, but the data do not show boxes. In 1998, Kelly Brennan and colleagues analyzed many attachment statements and found two main dimensions: anxiety and avoidance. In the same year, Fraley and Niels Waller found "no evidence for a true attachment typology". Shaver and Fraley's guide sums up the result: the styles are "regions in a two-dimensional space".

The most popular questionnaires today follow this result. The ECR has 36 items, with 18 for each scale. Fraley, Waller and Brennan (2000) revised it with item response theory, using data from 1,085 people, into the ECR-R. Fraley's notes on the ECR-R say there is no "natural" or "correct" way to turn the two scores into a style.

Why attachment style quizzes disagree

Two online quizzes can give you two different styles, and neither has to be broken. There are three reasons.

  1. 1.The styles are cut points on continuous scales. A person near a boundary can land in one style on one quiz and in the next style on another.
  2. 2.Interviews and questionnaires measure different things. The Adult Attachment Interview scores coherence of mind, while self-report scores conscious appraisals. Roisman and colleagues (2007) pooled the published studies, with 961 people in all, and found a mean correlation of .09 between interview security and questionnaire scores. By Cohen's criteria, that is trivial to small.
  3. 3.Attachment belongs to a relationship. Infant research classifies each relationship separately, and a child can be classified differently with each parent. But the two are not independent: a 1991 meta-analysis of 11 studies found that security with one parent depended on security with the other.

Can your attachment style change?

Yes. Attachment security is moderately stable, not fixed. Pinquart, Feussner and Ahnert (2013) pooled 127 papers with 21,072 attachments. They found an overall stability correlation of .39, and no significant stability over intervals longer than 15 years. Fraley's 2002 meta-analysis points the same way: in his overview, the stability of attachment to parents is a correlation of about .25 to .39.

Interview research has a name for one kind of change: earned security. In Roisman and colleagues' 2002 study, earned-secure adults describe negative childhood experiences, but describe them coherently. The same study found that adults classified this way from their own accounts were not more likely than other secure adults to have been insecure as infants. So a coherent story about a hard childhood is the marker, not proof of a change.

Relationships can also change attachment. Arriaga and colleagues (2018) proposed a model of how romantic relationships can build lasting security. Under their model, attachment anxiety should fall most in situations that give a person more confidence. Avoidance should fall most in situations of positive dependence on a partner. Fraley still notes that researchers do not have "a strong understanding of the precise factors" behind a change.

Childhood attachment also does not set adult relationships. Hazan and Shaver saw adult style as shaped "in part" by childhood relationships with parents. Fraley's overview concludes that child–parent attachment and adult romantic attachment are "only moderately related at best".

An attachment style is a research description, not a diagnosis. If patterns in your relationships cause you distress, talk to a qualified professional. Our guide to psychologists, psychotherapists and psychiatrists explains who does what.

What changes across cultures

Van IJzendoorn and Kroonenberg (1988) pooled 32 samples from eight countries, with 1,990 Strange Situation classifications. Avoidant classifications were more frequent in Western European samples, with a high share in Bielefeld, Germany. Resistant classifications were more frequent in Israel and Japan.

Their main finding was about the size of the differences. Variation between samples inside one country was nearly 1.5 times the variation between countries. So a national distribution does not describe a national character. The 2023 meta-analysis adds two points: avoidant attachment became less common over time, and North American samples differed from samples in Asia, Australia and New Zealand, and South America.

Learning the vocabulary

Attachment research gives one idea many names: anxious, preoccupied, ambivalent and resistant all appear above. Learn the terms against each other, in pairs and contrasts, not one at a time. Our guide on how to memorise psychology terms explains why. Psychology World, our dictionary app for iPhone and iPad, can help with this step. Look up the terms you want to keep, save them to one list, then practice them in the Flashcard, Question or Typing game.

Attachment words also travel into everyday speech, as therapy speak shows with "I'm avoidant". Two other theories look at early development from a different angle: Erikson's first stage, trust versus mistrust, and Baumrind's parenting styles, which describe the parent's approach, not the child's bond.

Frequently asked questions

How many attachment styles are there?

It depends on the model. Ainsworth's infant research and Hazan and Shaver's first adult study used three styles. Infant research uses four since Main and Solomon added disorganized in 1990, and Bartholomew and Horowitz's adult model also has four. Most researchers now measure adult attachment on two scales, anxiety and avoidance.

What is the most common attachment style?

Secure. In the 2023 meta-analysis of 285 infant studies, 51.6 percent of infants were classified secure. In Hazan and Shaver's first adult study, about 60 percent of people chose the secure description.

What are Ainsworth's attachment styles?

Ainsworth described three patterns in the Strange Situation: secure (B), avoidant (A) and resistant or ambivalent (C). The fourth infant pattern, disorganized (D), came from Main and Solomon in 1990, after Ainsworth's 1978 book.

What is the difference between Bowlby and Ainsworth?

Bowlby proposed the theory: an attachment system that keeps an infant close to a caregiver, and working models built from experience. Ainsworth made the theory measurable with the Strange Situation, and she named the secure, avoidant and resistant patterns.

Can you have more than one attachment style?

In infant research, yes: each relationship is classified separately, so a child can be secure with one parent and insecure with the other, although the two classifications are related. In adult research, a person has one position on two continuous scales, which may sit near the border between two styles.

Sources

  1. 1.Bowlby, J. (1969). Attachment and Loss, Volume 1: Attachment. Hogarth Press.
  2. 2.Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of Attachment: A Psychological Study of the Strange Situation. Erlbaum.
  3. 3.Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In Attachment in the Preschool Years. University of Chicago Press.
  4. 4.Madigan, S., Fearon, R. M. P., van IJzendoorn, M. H., et al. (2023). The first 20,000 strange situation procedures: A meta-analytic review. Psychological Bulletin.
  5. 5.Spangler, G., & Grossmann, K. E. (1993). Biobehavioral organization in securely and insecurely attached infants. Child Development.
  6. 6.Van IJzendoorn, M. H., & Kroonenberg, P. M. (1988). Cross-cultural patterns of attachment: A meta-analysis of the strange situation. Child Development.
  7. 7.Fox, N. A., Kimmerly, N. L., & Schafer, W. D. (1991). Attachment to mother/attachment to father: A meta-analysis. Child Development.
  8. 8.Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology.
  9. 9.Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology.
  10. 10.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.
  11. 11.Roisman, G. I., Holland, A., Fortuna, K., Fraley, R. C., Clausell, E., & Clarke, A. (2007). The Adult Attachment Interview and self-reports of attachment style: An empirical rapprochement. Journal of Personality and Social Psychology.
  12. 12.Roisman, G. I., Padrón, E., Sroufe, L. A., & Egeland, B. (2002). Earned-secure attachment status in retrospect and prospect. Child Development.
  13. 13.Fraley, R. C. (2002). Attachment stability from infancy to adulthood: Meta-analysis and dynamic modeling of developmental mechanisms. Personality and Social Psychology Review, 6(2), 123–151.
  14. 14.Pinquart, M., Feussner, C., & Ahnert, L. (2013). Meta-analytic evidence for stability in attachments from infancy to early adulthood. Attachment & Human Development.
  15. 15.Arriaga, X. B., Kumashiro, M., Simpson, J. A., & Overall, N. C. (2018). Revising working models across time: Relationship situations that enhance attachment security. Personality and Social Psychology Review.
  16. 16.Paetzold, R. L., Rholes, W. S., & Kohn, J. L. (2015). Disorganized attachment in adulthood: Theory, measurement, and implications for romantic relationships. Review of General Psychology.
  17. 17.Paetzold, R. L., & Rholes, W. S. (2021). The link from child abuse to dissociation: The roles of adult disorganized attachment, self-concept clarity, and reflective functioning. Journal of Trauma & Dissociation.
  18. 18.Fraley, R. C. (2018). A brief overview of adult attachment theory and research.
  19. 19.Shaver, P. R., & Fraley, R. C. Self-report measures of adult attachment.
  20. 20.Fraley, R. C. Information on the Experiences in Close Relationships-Revised (ECR-R) questionnaire.

Look the terms up yourself

18,000+ psychology and psychotherapy terms in English and German, on iPhone, iPad, Mac and Apple Vision.

See how Psychology World works
All posts