The Strange Situation: How Ainsworth Studied Infant Attachment

Mary Ainsworth wanted to see attachment, not just theorize about it. After years of watching mothers and babies at home, first in Uganda and then in Baltimore, she and her colleagues designed a short laboratory procedure that puts mild, everyday stress on a young child’s attachment system and records what happens.^[1][6] They called it the Strange Situation because everything in it is new to the baby: the room, the toys and the adult who walks in.

What did Ainsworth observe before the laboratory?

The Strange Situation grew out of home visits. In Uganda, Ainsworth watched mothers and babies in their everyday surroundings and saw infants use their mothers as a base from which to explore. Back in the United States she repeated the work in Baltimore, where observers visited families at home again and again during the baby’s first year.^[6]

The home notes let Ainsworth rate how each mother responded to her baby’s signals: whether she noticed crying or reaching, read it correctly and answered promptly and appropriately. She called this sensitivity. In her Baltimore sample, babies of more sensitive mothers were more likely to be classified as secure in the Strange Situation at the end of the first year.^[1]

Later studies checked whether this link holds up. A meta-analysis by Marianne De Wolff and Marinus van IJzendoorn pooled 66 studies and concluded that sensitivity is an important but not an exclusive condition for attachment security. The link was clearest in studies built like Ainsworth’s, and other aspects of parenting played a role as well.^[8]

What is the Strange Situation?

The Strange Situation is a standardized observation, usually with infants around twelve to eighteen months old. A caregiver and the infant enter an unfamiliar playroom. Over a sequence of short episodes a stranger joins them, the caregiver leaves twice and returns twice, and for one episode the infant is alone.^[1]

The logic follows Bowlby’s theory. In a safe setting, a securely attached infant should explore and use the caregiver as a secure base. When stress rises through the unfamiliar adult and the separations, the infant’s attachment system should activate and the infant should seek the caregiver as a safe haven. The procedure is built to observe both.

The eight episodes

Episode	Present	What happens
1	Caregiver, infant, observer	Observer brings them into the room and leaves
2	Caregiver, infant	Infant is free to explore the toys
3	Caregiver, infant, stranger	Stranger enters, sits, talks with the caregiver, then approaches the infant
4	Infant, stranger	First separation: caregiver leaves
5	Caregiver, infant	First reunion: caregiver returns, stranger leaves
6	Infant alone	Second separation: caregiver leaves again
7	Infant, stranger	Stranger returns
8	Caregiver, infant	Second reunion: caregiver returns

Most episodes last about three minutes. Separation episodes are shortened if the infant becomes very distressed.^[1][7] The whole procedure takes roughly twenty minutes.

What researchers watch for

The first episodes matter as a baseline. Before any separation, researchers note whether the infant explores the toys and how often they check back with the caregiver. A baby who plays freely while glancing at the caregiver now and then is showing secure base behavior in action.

Crying when the caregiver leaves is not the main signal, since babies differ a lot in how much they protest. The key moments are the reunions. Ainsworth’s coding looked at how much the infant sought closeness, how well contact comforted them, and whether the infant showed resistance or avoidance toward the returning caregiver.^[1]

The infant attachment categories

From these observations Ainsworth and her colleagues described three organized patterns, later joined by a fourth.^[1][2]

• Secure (B). The infant explores, may be upset by separation, and on reunion seeks the caregiver, is comforted fairly quickly and returns to play.
• Avoidant (A). The infant shows little obvious distress at separation and on reunion ignores or turns away from the caregiver.
• Resistant or ambivalent (C). The infant is very upset by separation and on reunion seeks contact but also resists it, for example by pushing away or staying hard to soothe.
• Disorganized (D). Mary Main and Judith Solomon noticed that some infants did not fit any of the three. Their behavior lacked a coherent strategy: approaching with the head turned away, freezing, or showing contradictory movements in the caregiver’s presence. They developed criteria for this pattern, now known as disorganized attachment.^[2]

A, B and C are called organized because each reflects a consistent way of dealing with the caregiver, even the insecure ones. D describes a breakdown of such a strategy under stress.

Limits of the Strange Situation

The Strange Situation is one of the most influential procedures in developmental psychology, and it has clear limits.

It measures a relationship, not a child. The classification describes an infant’s behavior with one particular caregiver. The same infant can show a different pattern with another caregiver.^[7]

Culture matters. A meta-analysis by Marinus van IJzendoorn and Pieter Kroonenberg compared Strange Situation studies across several countries and found that the distribution of categories varied between and within cultures.^[3] Practices around separations and strangers differ, which can change what the procedure means for the child. Fred Rothbaum and colleagues went further. Comparing the United States and Japan, they argued that the ideas of sensitivity, competence and secure base used in attachment research reflect Western values of autonomy and exploration, and they called for approaches rooted in each culture.^[9]

Categories may be a simplification. When Chris Fraley and Susan Spieker tested whether infant Strange Situation behavior falls into natural types, the data were better described by continuous dimensions.^[4] This mirrors the move toward attachment anxiety and avoidance as dimensions in adult research.

It requires expertise. Coding is done by trained observers. It is not a home test for parents.

It is a snapshot. The procedure captures about twenty minutes on one day. A tired, ill or unsettled infant may behave differently than usual, which is one reason researchers treat a single classification with care rather than as a permanent verdict about the child.

What do the infant categories predict?

The categories were never meant only as labels for twenty minutes of behavior. Researchers wanted to know whether they say something about later development. One answer comes from a meta-analysis by Pasco Fearon and colleagues on behavior problems such as aggression and rule breaking. Across 69 samples, insecure children showed more of these problems on average than secure children. The link was clearest for disorganized attachment and weaker for the avoidant and resistant patterns.^[10]

What the categories do not do is predict an individual child’s future. The findings describe modest average differences between groups, which leave plenty of room for individual children to develop in very different ways. Attachment security is also only moderately stable across the years, so later relationships and circumstances keep shaping the picture.^[5]

What the Strange Situation does not tell you about adults

Adult attachment styles borrowed their names from the infant categories, but the link is looser than it sounds.

The adult style names echo the infant letters, but they are measured differently and are not the same thing. Adult secure attachment shares its name with pattern B, anxious attachment (also called preoccupied) is related to the resistant C pattern, avoidant attachment is related to pattern A, and fearful-avoidant attachment is often linked with disorganized D.

Adults are never assessed with the Strange Situation. Adult self-report research asks people how they feel in close relationships now; it does not reconstruct how they behaved at one year old. A fearful-avoidant adult was not necessarily a disorganized infant, and a secure adult may have been classified as insecure as a baby. The four-category model article shows how the frameworks relate.

The Adult Attachment Interview

The closest adult counterpart to the Strange Situation is not a questionnaire but an interview. In the mid 1980s Mary Main and her colleagues moved attachment research “to the level of representation”: instead of watching behavior, they looked at how children and adults think and talk about attachment.^[11] The Adult Attachment Interview (AAI), developed by Carol George, Nancy Kaplan and Mary Main, asks adults about their childhood relationships and is coded for how they talk about those experiences, not only for what happened.^[12]

The AAI links directly back to the Strange Situation. A meta-analysis by van IJzendoorn found that parents’ interview classifications predicted whether their infants were secure or insecure in the Strange Situation, and also predicted how responsive the parents were to their infants. How that transmission works is still only partly understood, which he called the transmission gap.^[12] Self-report questionnaires such as the one behind our quiz belong to a different research tradition.

Research following people from infancy onward suggests that attachment security is only moderately stable over the years, which leaves plenty of room for later relationships and experiences to matter.^[5] So an infant classification is not a life sentence, and your current attachment style is not proof of what happened in your first year. The attachment myths article takes up this point.

If you are curious about your own patterns as an adult, our attachment style quiz is a starting point for self-reflection, not a clinical test. How the quiz works explains its limits. The broader story of Bowlby and Ainsworth is told in attachment theory.