Why Your “Attachment Style” Quiz Might Be Telling You the Wrong Story
If you’ve spent any time on social media lately, chances are you’ve encountered a quiz, a reel, or a thread promising to reveal your “attachment style.” Anxious. Avoidant. Secure. Disorganised. These labels have become part of everyday language — the way people explain their relationship patterns to friends, justify their reactions, or understand why the same painful dynamics seem to follow them from one relationship to the next.
I understand why this has appeal. Attachment theory offers a genuinely profound framework for understanding ourselves. The problem is that what’s being popularised in these quizzes bears only a surface resemblance to the original theory — and the difference matters enormously, particularly for anyone doing serious work on their relational history or their mental health.
Let me explain why.
Two research traditions that speak very different languages
Attachment theory, as most clinicians understand it, originates with John Bowlby’s work in the 1960s and 1970s, and was built upon by Mary Ainsworth’s landmark Strange Situation research with infants. It was later extended to adults through Mary Main’s development of the Adult Attachment Interview (AAI) at UC Berkeley. This body of work — what researchers call the developmental perspective — is grounded in decades of observational research on how the child-caregiver relationship organises behaviour, mental representation, and ultimately the nervous system itself.
At its core, the developmental model understands attachment as a protection system. From an evolutionary perspective, the attachment behavioural system exists to keep us safe. When a child is frightened or threatened, they seek proximity to a caregiver. The quality of that caregiving — whether it is available, responsive, and consistent — shapes what Bowlby called internal working models: deeply held, largely unconscious templates for how relationships work and whether we are worthy of protection and care.
What most people don’t realise is that there is an entirely separate research tradition — the social personality perspective — that emerged in the 1980s, when researchers Hazan and Shaver applied attachment terminology to adult romantic love. They created a self-report questionnaire that placed adults into three groups — secure, avoidant, and ambivalent — based on their answers to questions about closeness, intimacy, and fear of abandonment. This approach, while generating a large body of research, measures something quite different from what Bowlby and Ainsworth were describing.
The popular online quizzes and self-help content about “attachment styles” are almost entirely derived from this second tradition — not the first.
The critical difference: what each approach actually measures
Here is where it gets technically important, and clinically significant.
The developmental model, embodied in the AAI, assesses unconscious mental processes — specifically, the coherence with which a person can reflect on and make sense of their early relational experiences, including loss and trauma. It doesn’t ask you to describe your relationship patterns. It activates the attachment system through specific questioning and observes how you think and speak about your childhood experiences — the coherence, the gaps, the contradictions, the unresolved grief or fear that emerges in the way someone narrates their story rather than in what they consciously report.
The romantic attachment questionnaires, by contrast, assess conscious self-evaluation. They ask you what you believe to be true about yourself and others in intimate relationships. Your answers are based on your current perception of yourself — which, as any clinician knows, is shaped by all manner of things that have nothing to do with your original attachment history: your current relationship, your mood, your self-serving tendencies, your wish to see yourself in a particular light.
These are genuinely different phenomena. Researchers have found little correspondence between how someone scores on a self-report romantic attachment measure and how they are classified on the AAI. They appear to be measuring different constructs — and this isn’t surprising, because Bowlby’s original model was specifically about protection and safety, not about intimacy and closeness, which are also shaped by sexual and affiliative systems that operate quite separately.
The clinical stakes: what gets missed
From a clinical perspective, the most significant problem with the social personality model is that it largely fails to capture attachment disorganisation — which is, according to extensive research, the form of insecure attachment most strongly linked to serious mental health difficulties.
Disorganised attachment, identified by Main and Solomon in infants and assessed as “Unresolved” status on the AAI in adults, arises under the most severe relational conditions: when the attachment figure is simultaneously the source of fear and the person the child must turn to for safety. The child faces what researchers describe as an irresolvable dilemma — approach and be harmed, or withdraw and lose the only source of protection available. The result is a collapse of coherent attachment strategy.
This disorganisation does not resolve neatly into an “anxious” or “avoidant” style. It often looks organised in everyday functioning but remains disorganised at the level of mental representation — meaning that under sufficient relational stress, it can break through in the form of flooding, dissociation, behavioural dysregulation, or emotional collapse. Research has linked unresolved disorganised attachment to borderline personality features, dissociative symptoms, partner violence, suicidal behaviour in adolescents, and a range of clinical presentations that are poorly served by thinking of oneself as simply “anxious-attached.”
A self-report questionnaire cannot access this. The AAI was specifically designed to activate the attachment system through the interview process, because, as the developmental model insists, internal working models can only be accurately assessed when the system they organise is actually running.
Why this matters for therapy and for personal growth
I am not suggesting that self-report attachment measures are without value. They reliably tell us something real about how a person consciously experiences closeness, what they expect from partners, how they feel about intimacy and dependency. That is genuinely useful information — particularly for understanding current relational patterns and symptoms.
What they cannot do is reveal the unconscious organising principles that drive relational behaviour from beneath conscious awareness. They cannot identify unresolved trauma or grief in the way the AAI can. They cannot illuminate the processes that most directly predict risk for serious mental and relational difficulties.
The people most likely to be harmed by over-reliance on the popular “attachment style” framework are often those with the most complex histories — people who may understand themselves to be “anxiously attached” or “avoidant” while carrying unresolved loss or trauma that requires a very different clinical approach. The label provides a sense of self-understanding without necessarily pointing toward what needs to change or how.
This matters for the therapists working with these clients too. If we accept a client’s self-identification as simply “avoidant” without exploring the possibility of disorganisation beneath that apparent strategy, we may inadvertently collude with a defensive structure rather than helping to gently dismantle it.
A more useful way to think about it
Rather than thinking of the two approaches as competing, it can help to see them as occupying different levels of the same picture. Self-report attachment measures can offer a window into conscious relational models and current interpersonal patterns — what a person believes about themselves in relationship, how they experience closeness and threat in the present. The developmental model, and formal assessment through the AAI, accesses something deeper: the unconscious processes related to the original caregiving history, and particularly whether loss or trauma remains unresolved at a representational level.
For most people reading a blog or doing personal reflection, the popular attachment framework is a reasonable starting place for beginning to name patterns. But it is only a starting place. If you recognise yourself in those patterns and find that understanding them hasn’t changed them — if the same relational dynamics persist despite years of self-awareness — that is often a signal that something unconscious needs a different kind of attention than a quiz can provide.
True attachment-informed work requires a relationship. It requires time, consistency, and the kind of therapeutic attunement that can help reorganise the very system that was shaped by early relational experience. That process cannot be reduced to four categories and a paragraph of description. It rarely can be.
With love,
Andi
If you are interested to explore further you can book in for a session and I also offer the AAI.
