Attachment styles describe the predictable ways people respond when intimacy, threat, or separation puts a relationship under pressure. For exam candidates they are one of the few topic areas that bridges developmental psychology, adult personality pathology, and the therapeutic alliance — which is exactly why vignettes keep returning to them. In this article we translate Bowlby's theory and Ainsworth's observations into the four adult styles you must recognise, the instruments used to measure them, and the exam clues that separate one style from another.
By the end you will be able to name the four styles and their core expectations, describe how the Strange Situation and the Adult Attachment Interview differ, and explain — carefully — how disorganised attachment relates to borderline personality disorder without turning an association into a cause.
What attachment theory actually claims
John Bowlby's attachment theory (1969–1980) proposes that humans arrive with an evolved motivational system that drives proximity-seeking to a caregiver when the infant is distressed, and that the repeated quality of caregiving shapes an internal working model of relationships. The attachment system is activated by separation, fear, and illness; once soothed, it deactivates and exploration resumes. The balance between safe haven (returning to the caregiver under stress) and secure base (venturing out to explore) is the clinical heartbeat of the theory.
Mary Ainsworth operationalised Bowlby's ideas with the Strange Situation — a 20-minute laboratory procedure of eight episodes involving two separations and two reunions with the caregiver. Her Baltimore sample produced three patterns: secure, avoidant, and ambivalent/resistant. A fourth, disorganised pattern, was identified later by Main and Solomon (1986, 1990) in infants who showed contradictory, freeze, or disoriented behaviour on reunion.
The key mechanistic point examiners test: attachment is a dyadic regulatory system, not a personality label. The infant's behaviour is organised in relation to a particular caregiver, and the same infant can be securely attached to one parent and insecurely attached to another. Evidence for the framework comes from cross-cultural Strange Situation work (van IJzendoorn and Kroonenberg's 1988 meta-analysis showed security at roughly 62–75% across countries, with the avoidant/ambivalent balance varying culturally) and from longitudinal studies showing moderate, not perfect, stability into adulthood.
The four attachment styles explained
Adult attachment research converged on a two-dimensional model: attachment anxiety (fear of abandonment, hyperactivation of the attachment system) and attachment avoidance (discomfort with closeness, deactivation of the system). Four quadrants follow, and these are the four styles you will be asked to differentiate.
| Style | Behaviour in conflict | Core expectation of others | Approx. distribution in community samples |
|---|---|---|---|
| Secure (autonomous) | States the grievance directly, repairs quickly, tolerates the partner's distance | "Others are generally responsive; I am worthy of care" | ~56–64% |
| Anxious-preoccupied (ambivalent) | Escalates, pursues, keeps score, fears the relationship ending | "Others are unreliable; I must protest to be heard" | ~19–25% |
| Fearful-avoidant (disorganised) | Wants closeness then punishes it; approach–avoidance oscillation | "I want intimacy but expect harm from it" | ~16–20% |
| Dismissing-avoidant | Withdraws, minimises, contests the premise of the argument | "Reliance on others is unnecessary or humiliating" | ~10–20% |
Percentages are approximate and vary by sample and instrument (Hazan and Shaver's original 1987 vignette study used roughly 56/25/19% secure/anxious/avoidant; newer ECR-R samples re-distribute the avoidant group across dismissing and fearful quadrants). Treat ranges as estimates, never as published cutoffs.
Two pairs of distinctions carry most of the marks. First, dismissing-avoidant vs fearful-avoidant: both score low on reported distress, but dismissing individuals devalue closeness outright, while fearful-avoidant individuals desire closeness and simultaneously dread it — the approach–avoidance signature. Second, anxious vs fearful-avoidant: both protest, but the anxious person expects the relationship can be salvaged by louder signalling, whereas the fearful-avoidant person alternates between pursuit and withdrawal.
How attachment is measured: Strange Situation to ECR-R
Infancy: the Strange Situation remains the standard behavioural assay, valid for infants roughly 12–20 months old. Classifications rest on reunion behaviour — proximity-seeking and contact-comfort (secure), avoidance of contact despite distress (avoidant), contact-seeking followed with resistance (ambivalent), and disorganised sequences (fear, freezing, contradictory movements).
Adolescence and adulthood: the Adult Attachment Interview (AAI; George, Kaplan and Main) is a semi-structured 60–90 minute interview that scores coherence of narrative about childhood experiences, not the experiences themselves. Its categories are secure/autonomous, dismissing, preoccupied, and unresolved (the last coded for trauma or loss). The AAI's central, repeatedly tested finding is that the parent's state of mind — not the content of their memories — predicts the infant's Strange Situation classification at about 70–75% accuracy.
Self-report: the Experiences in Close Relationships questionnaire and its revised ECR-R dimension score attachment anxiety and avoidance across 36 items, and is the workhorse of experimental studies. The pitfall is a discrepancy effect: dismissing adults often report low distress while their AAI or physiological responses show otherwise. In an exam, if a stem mentions interview narratives, think AAI; if it mentions questionnaire subscale scores, think ECR-R.
Attachment, borderline personality disorder, and the alliance
Attachment research intersects with personality pathology in a way you must phrase precisely. Prospective and meta-analytic evidence (for example, the work collected in Winfield-Brown, Ogburn and colleagues' reviews) shows that fearful/disorganised attachment is over-represented in borderline personality disorder (BPD), and that it predicts identity disturbance, fear of abandonment, and emotion dysregulation — the very domains DSM-5-TR lists under Criterion 1 of BPD. Longitudinal data further suggest disorganisation in infancy predicts personality disorder symptoms in adulthood better than it predicts any single diagnosis.
Clinical pearl
Disorganised attachment is a risk marker associated with later BPD, not a cause and not a diagnosis. In essays and OSCE/CASC stations, write "associated with", "predicts risk", or "transdiagnostic vulnerability". Never write "disorganised attachment leads to BPD" — most examiners mark causal language about developmental risk as wrong.
The alliance application is the positive counterpart. Attachment-informed formulations hold that a patient's pattern in therapy mirrors their internal working model: the preoccupied patient seeks reassurance and reacts strongly to missed sessions; the dismissing patient cancels, intellectualises, and denies needing the therapist; the fearful-avoidant patient idealises then devalues. Meta-analysis of alliance-outcome research (Flückiger et al., 2018) consistently shows the working alliance accounts for a meaningful share of outcome variance across therapies — a fact worth one sentence in any "how would you engage this patient?" station.
Also worth knowing: attachment security shows moderate stability but is not destiny. Earned security — a dismissing or preoccupied adult who achieves a coherent AAI narrative through therapy, reflection, or a corrective relationship — is well documented, and is the evidence base for asking "can attachment styles change?" in exams.
Studying attachment for exams without rote-learning percentages
Convert the theory into discriminating features. When you read a vignette, flag three things: what happens at separation (does the child become inconsolable, indifferent, or disorganised?), how reunion unfolds (approach, avoidance, or contradiction?), and what the adult says about their childhood coherence. Then match to the style. Flashcards work best for the instrument–classification pairs (Strange Situation → four infant categories; AAI → four adult states of mind; ECR-R → two dimensions), while the vignette discrimination comes from practice, not memorisation.
That practice loop — concept note, spaced flashcards, then MCQs that force a choice between anxious and fearful-avoidant — is exactly what the 2,100+ concept library at Preppsy is built for, with each attachment concept linked to its exam-style questions.
Frequently asked questions
What are the four attachment styles and how do I tell them apart in a vignette?
The four styles are secure, anxious-preoccupied, dismissing-avoidant, and fearful-avoidant (disorganised). In a vignette, watch three moments: separation distress (secure children protest then settle; anxious children stay distressed; avoidant children appear unbothered), reunion behaviour (secure approaches and is soothed; anxious protests then is hard to soothe; dismissing avoids contact; fearful-avoidant oscillates), and adult narrative coherence if an interview is described. The single most discriminating question is whether closeness is welcomed, demanded, dismissed, or simultaneously wanted and feared.
Is attachment style the same as personality disorder?
No. Attachment style is a dimension of relational strategy measured by behaviour or interview narrative; a personality disorder is a clinical diagnosis requiring pervasive, inflexible patterns causing distress or impairment — in DSM-5-TR, a specific PD diagnosis requires impairments in at least two of four domains (identity, self-direction, empathy, intimacy) plus a pervasive pattern since adolescence. Attachment insecurity is a transdiagnostic risk factor associated with several conditions, BPD among them, but most people with insecure attachment do not have a personality disorder, and secure attachment does not exclude one.
Does the Adult Attachment Interview measure whether childhood was happy?
Not at all — this is the classic exam trap. The AAI codes the coherence, specificity, and reflectiveness of the narrative, not its content. A person who describes genuinely difficult childhoods can be classified secure/autonomous if their account is coherent and reflective, while someone recounting an apparently idealised childhood can be classified dismissing if the narrative is vague, contradictory, or resistant to detail. State of mind predicts the next generation's Strange Situation classification; the content of remembered experience does not.
Can attachment styles change in adulthood?
Yes, partially. Attachment shows moderate longitudinal stability, but movement — termed earned security — is documented when adults achieve coherence about previously unresolved experiences, often through therapy or a stable corrective relationship. Transitions are more common into security than out of it, and dimensional scores on instruments like the ECR-R can shift with sustained intervention. In exams, the defensible phrasing is "moderately stable but malleable", never "fixed in infancy".
Key takeaways
- Attachment is a dyadic regulatory system (Bowlby): safe haven under stress, secure base for exploration — not a fixed personality label.
- Four adult styles map onto two dimensions (anxiety × avoidance): secure, anxious-preoccupied, dismissing-avoidant, fearful-avoidant.
- Know your instruments: Strange Situation (infant behaviour), AAI (adult narrative coherence), ECR-R (self-report dimensions).
- The AAI scores coherence of narrative, not quality of childhood — the most reliably tested distinction in this area.
- Disorganised attachment is a risk marker associated with BPD; describe the link as association, never causation.
- Attachment is moderately stable but malleable: earned security is real and examinable.
Attachment theory rewards the candidate who can move from mechanism to discrimination under time pressure — practise that translation on MRCPsych-style questions and the four styles will stop being a memory list and start being a clinical tool.
Cover image: “Attachment Styles Explained (cover)” by Preppsy Editorial Team (Original artwork) via Preppsy.
