Case Study 1: From the Strange Situation to Instagram Reels — How Attachment Theory Was Transformed

The Original Research

Mary Ainsworth's Strange Situation has been replicated hundreds of times across multiple cultures. It is one of the best-established paradigms in developmental psychology. The core findings — that infants form attachment patterns based on caregiver responsiveness, and that these patterns predict some later developmental outcomes — are robust.

Key meta-analytic findings (Groh et al., 2017; Fearon et al., 2010):

• Secure attachment in infancy predicts better peer relationships in childhood (effect size: small to moderate, r ≈ 0.20)
• Insecure attachment predicts higher levels of externalizing behavior problems in childhood (effect size: small, r ≈ 0.15)
• The predictive power is modest — attachment explains some variance in later outcomes but is far from deterministic

The original researchers (Bowlby, Ainsworth, Main) emphasized that attachment patterns are influenced by ongoing experiences, not just early childhood. Bowlby himself wrote about the potential for change through new relationships.

The Translation to Adult Romance

In 1987, Hazan and Shaver published "Romantic Love Conceptualized as an Attachment Process" in the Journal of Personality and Social Psychology. They proposed that adult romantic relationships function as attachment bonds, analogous to infant-caregiver bonds.

Their study: a newspaper survey asking adults to choose which of three descriptions best matched their romantic relationship experience (corresponding to secure, anxious, and avoidant). They found that the proportions roughly matched Ainsworth's infant distributions.

The study was creative but had significant limitations: - Newspaper survey (self-selected sample, no behavioral observation) - Forced-choice categorization (one of three descriptions, not dimensional measurement) - Cross-sectional (one time point, can't assess stability or change) - Retrospective self-report (people described their relationship experience, not observed behavior)

Later researchers developed better measures (the ECR, the AAI) and moved toward dimensional models (continuous scores on anxiety and avoidance rather than categories). But the popular version frozen in social media is closer to Hazan and Shaver's 1987 newspaper format than to the current research.

The Social Media Version

On Instagram and TikTok, attachment theory has been transformed into:

A fixed identity. "I'm anxious-attached" = "I AM this." The research says: "I currently score higher on attachment anxiety in the context of my current relationship." The difference is enormous.

A complete explanation for relationships. "My attachment style explains why I do everything I do in relationships." The research says: "Attachment dimensions predict some relationship behaviors with modest effect sizes, alongside many other factors."

A product. "Take this quiz to find your attachment style! Buy this course to heal your insecure attachment!" The quiz is not a validated instrument. The course may or may not be evidence-based.

A dating filter. "Only date securely attached people." The research says: security is associated with better relationships, but attachment changes and isn't the only factor that matters.

The Transformation Mapped

Discussion Questions

1. Hazan and Shaver's 1987 study used a newspaper survey with forced-choice categories. How might their results have been different with dimensional measurement and behavioral observation?

2. The infant attachment research is solid, but the extension to adult romance is more contested. Should these be presented as the same theory, or as related but distinct frameworks?

3. If someone finds genuine value in understanding their attachment patterns (even through an unvalidated quiz), does the inaccuracy of the measurement tool matter? When does accuracy matter and when doesn't it?

4. How would you redesign attachment content for social media to be both engaging and accurate? What elements would you keep and what would you change?