What is your attachment style in relationships?

Attachment theory was developed by British psychiatrist John Bowlby in the 1950s and 1960s. He proposed that the bond between a child and their primary caregiver creates an internal working model that shapes how that person approaches relationships throughout life. Cindy Hazan and Phillip Shaver extended the theory to adult romantic relationships in 1987, demonstrating that the patterns of secure, anxious, and avoidant relating observed in children could be measured in adults. A fourth pattern, disorganised or fearful-avoidant, was later described by Mary Main and Erik Hesse. The theory does not claim that childhood determines destiny. It identifies default patterns that tend to activate under relationship stress, and these patterns can be understood and modified with awareness, new relationship experiences, and professional support. (Source: Hazan & Shaver, Journal of Personality and Social Psychology, 1987)

Yes. Mickelson et al. (1997) estimated disorganised or fearful attachment at roughly 5% of the adult population, making it the least common of the four patterns. Some researchers argue disorganised features are underreported in self-report measures because individuals with this style may lack consistent awareness of their own attachment behaviour. Clinical samples show higher rates of disorganised attachment, particularly in populations with trauma histories.

Yes, though research suggests anxious-anxious pairings can produce high relationship volatility. Both partners hyperactivating their attachment systems simultaneously can lead to escalating conflict cycles and difficulty with de-escalation. If both partners have awareness of their patterns and strong communication skills, the relationship can also be characterised by high intensity of positive connection. Couple therapy outcomes for anxious-anxious pairings are generally favourable when both partners engage.

Yes. Attachment is more accurately measured on two continuous dimensions, anxiety and avoidance, rather than as four discrete boxes. Bartholomew and Horowitz (1991) formalised this two-dimensional model, showing that most people fall at different points on each axis rather than cleanly into one category. Someone might score moderately high on anxiety and low on avoidance, placing them between secure and anxious-preoccupied. The four-category model is a useful simplification, but the reality is that most people recognise elements of more than one style in themselves. This is entirely normal and reflects the continuous nature of the underlying dimensions. Research using this model has found that people's scores shift meaningfully across different relationship contexts, for example, showing more avoidance in romantic partnerships than in friendships. (Source: Bartholomew & Horowitz, Journal of Personality and Social Psychology, 1991)

The anxious-avoidant pairing is one of the most commonly discussed attachment dynamics. The anxiously attached partner's pursuit of closeness triggers the avoidantly attached partner's need for distance, which in turn amplifies the anxious partner's fear of abandonment and escalates their pursuit behaviour. Research by Fraley and Shaver confirmed that this pursuer-withdrawer dynamic is predicted by the combination of anxious and avoidant attachment styles. The cycle is self-reinforcing and can persist without both partners developing awareness of their contribution to it; for data on how argument patterns unfold in this dynamic, see the argument frequency calculator.

Research consistently links insecure attachment, particularly anxious and disorganised patterns, to higher rates of anxiety, depression, and relationship distress. However, attachment style is a relational pattern, not a mental health condition. Secure attachment acts as a protective factor, buffering stress and supporting emotional regulation across the lifespan. Insecure attachment increases vulnerability but does not determine outcomes. Bakermans-Kranenburg and van IJzendoorn (2009), reviewing over 10,000 Adult Attachment Interviews, found that clinical populations show significantly higher rates of insecure and disorganised attachment compared to the general population, suggesting that attachment insecurity is a risk factor worth addressing in clinical settings. Therapy approaches like Emotionally Focused Therapy (EFT) specifically target attachment patterns and have strong evidence for improving both relationship satisfaction and individual wellbeing; secure attachment is also strongly associated with relationship longevity, as shown in the relationship longevity calculator. (Source: Bakermans-Kranenburg & van IJzendoorn, Attachment & Human Development, 2009)

Based on the Mickelson et al. (1997) nationally representative sample, yes: 56% of US adults are classified as securely attached. The proportion varies by study, population, and measurement approach. Some studies using dimensional rather than categorical classification find lower secure rates. What is consistent across studies is that secure attachment is the modal pattern in Western adult samples, meaning it is the most common single category even when insecure types are combined. The broader distribution of attachment styles in the adult population is covered on the relationship statistics page.

Attachment style develops in early childhood through repeated interactions with primary caregivers. Mary Ainsworth's Strange Situation experiments in the 1970s identified three infant attachment patterns based on how infants behaved when briefly separated from and reunited with their caregiver. Infants with consistently responsive caregivers tended to develop secure attachment. Those with inconsistently available caregivers often developed anxious-ambivalent patterns. Those with emotionally unavailable or rejecting caregivers developed avoidant patterns. Main and Hesse later described disorganised attachment in infants whose caregivers were frightening or frightened, creating an unresolvable dilemma for the child. These early patterns become internal working models that shape expectations in adult relationships. Attachment patterns can also affect parenting style, creating an intergenerational transmission effect: securely attached parents are more likely to raise securely attached children, while unresolved insecure patterns can be transmitted across generations until the cycle is interrupted through therapy or corrective relational experience. However, these patterns are not fixed: research shows that 30-40% of people show a different attachment classification when retested after significant life experiences. (Source: Ainsworth et al., Patterns of Attachment, 1978; Hazan & Shaver, 1987)

Sharing attachment knowledge with a partner can be helpful if done constructively. It works best when both partners take the quiz and discuss results together, framing the conversation around understanding rather than labelling. Saying "I tend toward anxious attachment, so I might need more reassurance during conflict" is useful. Saying "You are avoidant and that is why our relationship is struggling" is not. Attachment language is a tool for empathy and communication, not a diagnostic label. Research on couples who explicitly discuss their attachment patterns shows improved communication and greater relationship satisfaction when the conversation is collaborative rather than blaming. If the conversation feels charged, a couples therapist can facilitate the discussion in a way that keeps it productive. The evidence base for Emotionally Focused Therapy as a couples intervention is particularly strong for anxious-avoidant pairings. (Source: Hazan & Shaver, 1987; Johnson, 2004)

Several validated psychometric instruments are used in attachment research. The Experiences in Close Relationships-Revised (ECR-R), developed by Fraley, Waller, and Brennan (2000), is the most widely used self-report measure in academic research and measures attachment anxiety and avoidance on two continuous dimensions. The Relationship Scales Questionnaire (RSQ), developed by Griffin and Bartholomew (1994), assesses Bartholomew's four-category model. The Attachment Style Questionnaire (ASQ), developed by Feeney, Noller, and Hanrahan (1994), measures five dimensions of adult attachment. The Adult Attachment Interview (AAI), developed by George, Kaplan, and Main, is a semi-structured clinical interview that classifies attachment based on how people narrate their childhood experiences rather than what they report about current relationships; it is administered by trained clinicians and considered the gold standard in developmental research. FTN's quiz draws on the original Hazan and Shaver (1987) items, which categorise into three primary types (secure, anxious, avoidant) and closely follow the same theoretical framework as the ECR-R while being accessible for self-assessment.

Earned security describes the process by which someone who started out with an insecure attachment style develops secure functioning through corrective relational experiences. This can happen through a long-term relationship with a securely attached partner, through therapy (particularly attachment-focused or Emotionally Focused Therapy), or through sustained personal growth work. Longitudinal research has documented earned security as a genuine phenomenon: adults classified as insecure in early adulthood can show secure attachment representations in the Adult Attachment Interview later in life, even without a change in the events of their childhood. The key mechanism appears to be the development of a coherent, integrated narrative about one's early experiences, including difficult ones, rather than the elimination of all negative early memories. Earned security is associated with outcomes comparable to those of continuous security in relationships and parenting. (Source: Hesse, in Cassidy & Shaver, Handbook of Attachment, 2008; Bakermans-Kranenburg & van IJzendoorn, 2009)