The question
Attachment theory grew out of John Bowlby’s work, which treated a baby’s bond with the person who cares for it as something with a biological, evolutionary function. Mary Ainsworth wanted to see that bond at work. Her idea was simple: attachment and exploration pull against each other. A baby who feels safe uses the parent as a base and goes off to explore; a baby who feels threatened stops exploring and looks for the parent.
So what happens if you gently raise the stakes? Put a one-year-old in an unfamiliar room, add a stranger, and have the parent leave for a few minutes. The most revealing moment, Ainsworth found, is not the separation but the reunion.
How it was done
Ainsworth and her team at Johns Hopkins University in Baltimore had already been visiting families at home since the babies’ birth, in visits of about four hours roughly every three weeks. Near the first birthday, each mother and baby came to the laboratory for a session of eight short episodes:
- Mother and baby are shown into a room with toys (about 30 seconds).
- Mother and baby are alone; the baby is free to explore (3 minutes).
- A stranger joins mother and baby (3 minutes).
- The mother leaves; the stranger stays with the baby (first separation, 3 minutes or less).
- The mother returns and the stranger leaves (first reunion).
- The mother leaves again; the baby is alone (second separation, 3 minutes or less).
- The stranger comes back in (3 minutes or less).
- The mother returns (second reunion).
Separation episodes were cut short if the baby became very upset. Researchers recorded exploration, crying, and especially what the baby did when the mother came back: did they seek her out, hold on, push away, or ignore her?
The first full report, by Ainsworth and Silvia Bell in 1970, described 56 white, middle-class infants aged 49 to 51 weeks. An earlier preprint shows how that total was made up: 23 babies from the home-observation study and 33 tested only in the laboratory. The book-length account, Patterns of Attachment by Ainsworth, Mary Blehar, Everett Waters and Sally Wall (1978), brought together the full longitudinal study of 26 Baltimore families, observed at home for more than 70 hours each over the first year.
What came out
Babies explored more while their mother was present and less once she left; crying and searching rose during separations. On reunion most babies sought closeness, but not all, and the exceptions became the heart of the study. Ainsworth sorted the babies into three groups, each with subgroups:
- Group B (secure): the baby actively greets or seeks the mother on her return, gets comfort if upset, and goes back to playing. This was the largest group.
- Group A (avoidant): the baby shows little interest in closeness on reunion and tends to look away from, move away from, or ignore the mother.
- Group C (resistant, or ambivalent): the baby wants contact but also resists it, often angrily, or is too passive to use the mother as a base; such babies are hard to settle.
In US samples the rough proportions came to be cited as about 70% secure, 20% avoidant and 10% resistant.
The home visits gave the categories their meaning. Mothers of B babies had been rated as the most sensitive to their babies’ signals during the year; mothers of A and C babies were rated less sensitive, with differences between them in being interfering, rejecting or hard to reach. In the Baltimore data, the correlation between maternal sensitivity and infant security was very high, about .78.
A fourth category. Over the next decade, researchers kept meeting babies who did not fit A, B or C. Mary Main, a former graduate student of Ainsworth’s who was by then at the University of California, Berkeley, found that many supposedly “unclassifiable” infants showed unusual behaviour in the parent’s presence. With Judith Solomon she described a disorganized/disoriented (D) pattern (1986, with coding procedures in 1990). It is defined by usually brief behaviours that seem to lack a goal or contradict each other: approaching with the head turned away, freezing, stilling, mistimed movements, or open signs of apprehension towards the parent. Coders give a D classification alongside a best-fitting A, B or C pattern, and only when such behaviour is intense enough; mild, fleeting versions are common.
What really happened (replications and critiques)
The patterns travel, but their mix changes. In 1988, Marinus van IJzendoorn and P. M. Kroonenberg pooled nearly 2,000 Strange Situation classifications from 32 samples in eight countries. Secure attachment was the most common pattern everywhere. There was a trend towards more avoidant classifications in Western European samples and more resistant ones in Israel and Japan. But the striking finding was that differences within countries were about one and a half times larger than differences between them: two samples from the same country could differ more than samples from different continents.
Is the stress the same for every baby? An early Japanese sample reported 37% resistant infants and no avoidant ones. The attachment researcher L. A. Sroufe suggested that Japanese babies of the time were rarely separated from their mothers, so the procedure was more frightening for them, especially when separations ran their full length. In Israeli kibbutz samples, many sessions had to be shortened or stopped because babies were inconsolable. In 1990 the Japanese researcher K. Takahashi questioned whether the procedure’s core assumptions hold in Japan. In 2000, Rothbaum and colleagues went further, arguing that attachment theory’s ideas of sensitivity, competence and the “secure base” reflect Western values of independence, while in Japan closeness and dependence (amae) are prized; behaviour that looks “ambivalent” in Baltimore may be ordinary there. Defenders, among them van IJzendoorn and Sagi (1999), concluded that the cross-cultural findings are broadly consistent with the theory.
Temperament or relationship? The developmental psychologist J. Kagan argued that the categories may partly reflect inborn differences in how easily babies become distressed: a baby who gets very upset when the mother leaves is more likely to rush to her on return and be scored secure. Sroufe answered in 1985 with several points: some babies are classified differently with mother and with father, which is hard to explain by a fixed trait; middle-class classifications were highly stable over six months; and crying alone does not separate the groups, since some secure babies cry as little as avoidant ones and others as much as resistant ones. Sroufe allowed that temperament might shape how a pattern is expressed. How much temperament contributes is still discussed.
A smaller effect than the original. The Baltimore link between sensitivity and security was later found to be much weaker. A 1997 meta-analysis of 66 studies (4,176 mother–infant pairs) found a correlation of about .24 (after correction for measurement error) in the strictest set of studies, against .78 in Baltimore, a gap the authors linked partly to the small, exploratory original sample. Sensitivity matters, but it is one influence among several.
How common is D? A 1999 meta-analysis of nearly 80 studies found that about 15% of infants in middle-class, non-clinical families were classified as disorganized, with rates two to three times higher in some high-risk groups, for example after maltreatment. In 2017, a large group of attachment researchers, including Solomon and Sroufe, warned that D had been misused: it is a research classification, not a diagnosis, it cannot screen for abuse, and removing a child “can never be justified solely” on that basis.
Ethics
The Strange Situation does not rely on deception: the mothers were part of the procedure, leaving and returning on cue. It does, by design, cause mild stress. The original rule of cutting separations short when a baby is very upset is part of the procedure. The reports we read do not describe how consent was obtained in the 1960s.
The main ethical question raised since is cultural: stress that is mild for some babies may be much stronger for babies who are almost never apart from their mothers, as the Japanese and kibbutz samples suggested.
What it means today
The Strange Situation gave developmental psychology one of its most influential tools and a shared language: secure, avoidant, resistant, disorganized. Researchers stress points that often get lost when it becomes a popular label:
- The categories describe a relationship at a particular moment, not a fixed quality of a child. A baby can be classified differently with each parent.
- They were developed for babies of about 12 to 20 months, in a standardised room, scored by trained coders. They are not something to judge at home or in a nursery.
- A disorganized classification does not inevitably lead to later problems; at group level it is only a small-to-moderate predictor.
Ainsworth herself, in her last interview, wanted the procedure used alongside careful observation of families in their everyday lives, not on its own.
Going beyond what the study tested: the research suggests that reunions carry a lot of information, and that babies vary widely in how they show attachment. This page is not parenting or clinical advice; any concern about a particular child is a question for a qualified professional who knows the family.
Would you have passed?
Coming soon. Walk through the eight episodes in a drawn room, then read short descriptions of fictional reunions and guess which pattern a trained coder would assign. You will see how often experts’ categories hinge on a few seconds of behaviour. One honest caveat: this is a guessing game about the research, not a test of you or of any real child.
Questions people ask
What is the Strange Situation?
It is a laboratory procedure created by Mary Ainsworth and colleagues for infants of about one year. In eight short episodes, the baby plays with the parent in an unfamiliar room, meets a stranger, is left twice for a few minutes and is reunited twice. Researchers pay most attention to how the baby responds when the parent comes back.
What are the attachment types A, B, C and D?
B (secure) babies seek or welcome the parent on return and then settle. A (avoidant) babies tend to ignore or turn away from the parent at reunion. C (resistant or ambivalent) babies seek contact but also resist it, often angrily, and are hard to settle. D (disorganized), added by Mary Main and Judith Solomon, describes brief contradictory, frozen or fearful behaviour that does not fit an organized strategy. These are research classifications of a relationship, not diagnoses of a child.
Are attachment patterns the same in every culture?
Secure attachment was the most common pattern in every country in a 1988 meta-analysis of nearly 2,000 classifications from eight countries. But the mix of insecure patterns varied, with more avoidant classifications in Western European samples and more resistant ones in Israel and Japan, and differences within countries were about 1.5 times larger than differences between them. Critics argue the procedure and its categories carry Western assumptions about independence.
Does a disorganized classification mean a child was abused?
No. A 2017 review by many leading attachment researchers states that it is not a clinical diagnosis, cannot be used to screen for maltreatment, and is not by itself a reason to remove a child from a family. It is more common in maltreated and high-risk groups, but it has other pathways and many maltreated infants are not classified as disorganized. This site does not offer advice on individual children; concerns belong with a qualified professional.
Sources
- Ainsworth, M. D. S., & Bell, S. M. (1970). Attachment, exploration, and separation: Illustrated by the behavior of one-year-olds in a strange situation. Child Development, 41(1), 49–67. doi:10.2307/1127388
- Ainsworth, M. D. S., Bell, S. M. V., & Stayton, D. J. (1969). Individual differences in strange-situation behavior of one-year-olds. Johns Hopkins University (ERIC ED 056742). Later published in H. R. Schaffer (Ed.), The origins of human social relations (1971, pp. 17–58). Academic Press.
- Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Lawrence Erlbaum Associates. Reissued 2015 by Psychology Press.
- Main, M., & Solomon, J. (1986). Discovery of a new, insecure-disorganized/disoriented attachment pattern. In M. Yogman & T. B. Brazelton (Eds.), Affective development in infancy (pp. 95–124). Ablex.
- Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years (pp. 121–160). University of Chicago Press.
- van IJzendoorn, M. H., & Kroonenberg, P. M. (1988). Cross-cultural patterns of attachment: A meta-analysis of the Strange Situation. Child Development, 59(1), 147–156. doi:10.1111/j.1467-8624.1988.tb03202.x
- Sagi, A., van IJzendoorn, M. H., & Koren-Karie, N. (1991). Primary appraisal of the Strange Situation: A cross-cultural analysis of preseparation episodes. Developmental Psychology, 27(4), 587–596. scholarlypublications.universiteitleiden.nl/access/item%3A2901347/view
- De Wolff, M. S., & van IJzendoorn, M. H. (1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68(4), 571–591. www.marinusvanijzendoorn.nl/wp-content/uploads/2012/07/De-Wolff-Van-IJzendoorn-1997.-Child-Development.-Sensitivity-Attachment-meta-analysis-parental-antecedents.pdf
- van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999). Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants, and sequelae. Development and Psychopathology, 11(2), 225–250. doi:10.1017/S0954579499002035
- Granqvist, P., Sroufe, L. A., Dozier, M., Hesse, E., Steele, M., van IJzendoorn, M. H., Solomon, J., Schuengel, C., Fearon, P., Bakermans-Kranenburg, M. J., et al. (2017). Disorganized attachment in infancy: A review of the phenomenon and its implications for clinicians and policy-makers. Attachment & Human Development, 19(6), 534–558. doi:10.1080/14616734.2017.1354040
- Duschinsky, R. (2018). Disorganization, fear and attachment: Working towards clarification. Infant Mental Health Journal, 39(1), 17–29. doi:10.1002/imhj.21689
- Sroufe, L. A. (1985). Attachment classification from the perspective of infant–caregiver relationships and infant temperament. Child Development, 56, 1–14. psychology.psy.sunysb.edu/attachment/online/Sroufe%20articles/Sroufe_1985CD%20Attachment%20and%20temperament.pdf
- Kagan, J. (1982). Psychological research on the human infant: An evaluative summary. W. T. Grant Foundation.
- Waters, E. (1978). The stability of individual differences in infant–mother attachment. Child Development, 49, 483–494.
- Miyake, K., Chen, S., & Campos, J. (1985). Infant temperament, mother's mode of interaction, and attachment in Japan: An interim report. Monographs of the Society for Research in Child Development, 50(1–2, Serial No. 209), 276–297.
- Takahashi, K. (1990). Are the key assumptions of the 'strange situation' procedure universal? A view from Japanese research. Human Development, 33(1), 23–30.
- Rothbaum, F., Weisz, J., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security in the United States and Japan. American Psychologist, 55(10), 1093–1104. doi:10.1037/0003-066X.55.10.1093
- Spies, R., & Duschinsky, R. (2021). Inheriting Mary Ainsworth and the Strange Situation: Questions of legacy, authority, and methodology for contemporary developmental attachment researchers. SAGE Open, 11(3). doi:10.1177/21582440211047577
- California Magazine (Cal Alumni Association, UC Berkeley), Summer 2012: feature on attachment research and Mary Main's work. alumni.berkeley.edu/california-magazine/summer-2012-north-south/stuck-you/