Van Ijzendoorn and Kroonenberg (1988) pooled 32 Strange Situation studies from eight countries to ask whether infant attachment is universal or shaped by culture.
Secure attachment was the most common type everywhere. But insecure attachment patterns varied systematically by culture, and differences within a single country were even larger than differences between countries.
Mary Ainsworth (1970) conducted most of her research in the USA, but child-rearing practices vary considerably from place to place regarding environment, traditions, and beliefs about children.
To systematically examine the extent of intracultural versus cross-cultural variation, a broader analysis aggregating data from multiple samples in different countries was needed.
Two Main Types of Cultures
- Individualist cultures value independence, with each person working toward their own goals, e.g., USA and Europe (Western Cultures).
- Collectivist cultures value cooperation with each person working towards the family or group goals e.g., Japan and Israel (Eastern Cultures).
Van Ijzendoorn, M. H., & Kroonenberg, P. M. (1988). Cross-cultural patterns of attachment: A meta-analysis of the strange situation. Child Development, 59, 147-156. https://doi.org/10.2307/1130396
Aims
- How does attachment work in different cultures?
- To investigate attachment types across cultures and see how the three main attachment styles applied.
- To investigate whether attachment styles (secure and insecure) are universal, or culturally specific and shaped by local traditions and beliefs.
Procedure
Van Ijzendoorn & Kroonenberg (1988) collected Strange Situation attachment classification data from 32 samples in 8 different countries (N = 1,990).
This was a meta-analysis of existing Strange Situation studies rather than an original experiment. A meta-analysis pools data from many separate studies to reach one overall, more powerful conclusion.
The Strange Situation itself is Ainsworth’s standardized laboratory procedure. In it, an infant experiences a stranger’s entry, two brief separations from the caregiver, and two reunions. Coders use the infant’s reactions to classify attachment as secure, avoidant, or resistant.
The procedure was to aggregate and statistically analyze the published attachment classification frequencies from qualifying samples.
Studies were included if they:
- Used the strange situation.
- Observed only mother-infant pairs (i.e., not the father- or other carer-pairs).
- Classified infants into one of the attachment types: A (Avoidant), B (Secure), or C (Ambivalent/Resistant).
Samples smaller than 35 were excluded, as were special populations like Down syndrome or twins and children older than 2 years.
The total sample (N = 1,990) included 32 samples from 8 countries. By country: USA (18 samples), Netherlands (4), Germany (3), Japan (2), Israel (2), Sweden (1), UK (1), and China (1, with Chinese-American infants).
Sample sizes ranged from 35 to 115 (mean = 62). Infants’ age ranged from 11 to 24 months.
Using a meta-analysis (a statistical technique), they calculated the average percentage for each country’s attachment styles (e.g., secure, avoidant, resistant).
Statistical measures: The authors computed standardized residuals for each sample’s A, B, and C frequencies to assess deviations from the global distribution.
They used Bonferroni-corrected chi-square tests to evaluate the significance of cross-cultural and intracultural differences. Correspondence analysis graphically depicted similarities and differences among the sample distributions.
Findings
Secure attachment was the most common type in every culture. The lowest percentage of secure attachments was shown in China, and the highest in Great Britain.
Results showed that individualistic countries valuing independence, such as Germany, had high levels of insecure-avoidant attachment. Collectivist countries, such as Japan, had high levels of insecure-resistant attachment instead.
Conclusion
The overall consistency in secure attachment types leads to the conclusion that there may be universal, innate characteristics underpinning infant-caregiver interactions. They do not need to be taught to bond securely.
However, the significant variation in insecure attachment types shows that universality is limited. Child-rearing practices and cultural values shape how attachment is expressed and classified.
Belsky (1999) pushed this further. If ancestral environments varied, he argued, evolution may have favoured a repertoire of attachment patterns, each suited to a different rearing ecology, rather than one universal ideal.
Germany: Independence, Not Insecurity
Aim: To classify North German (Bielefeld) infants’ attachment behaviour in the Strange Situation, and relate the high rate of avoidance to German cultural values around early independence.
Method: A longitudinal study observed 49 mother-infant pairs in Bielefeld, assessing the Strange Situation at 12 months and, with fathers, at 18 months.
Results: Around half of the infants were classified insecure-avoidant, well above comparable US samples. The secure rate was correspondingly lower.
Conclusion: Grossmann et al. (1981) interpreted this as reflecting a North German cultural ideal of interpersonal distance and self-reliance, not insensitive parenting. As they put it, German parents seek ‘independent, non-clingy infants, who do not make demands on parents, but obey their commands.’
Rather than a maladaptive bond, the avoidant classification may partly capture culturally valued independence training, though it risks the Strange Situation over-pathologising German child-rearing norms.
Japan: Distress and the Limits of the Procedure
Aim: To examine whether the Strange Situation validly applies to Japanese infants, who are reared in near-constant proximity to their mothers.
Method: Takahashi (1986) observed 60 middle-class Japanese mother-infant pairs in Sapporo in the standard Strange Situation at 12 months, classifying infants and noting their distress across episodes.
Results: No infants were classified insecure-avoidant. The sample split between secure and an elevated rate of insecure-resistant attachment. Japanese infants became extremely distressed during the separation episodes, so unfamiliar that the procedure sometimes had to be stopped early.
Conclusion: Takahashi (1990) argued that the key assumptions of the Strange Situation are not universal. The procedure assumes separation causes comparable distress everywhere. But in Japan the separation was so unfamiliar that the resulting behaviour may reflect the method, not the child’s actual attachment.
Israel: The Kibbutz and the Unfamiliar Stranger
Aim: To examine attachment security among kibbutz-reared Israeli infants, whose upbringing is collective, comparing their attachment to mother, father, and the metapelet, the community’s professional infant caregiver.
Method: Strange Situation assessments classified the infants’ attachment. Researchers measured this separately to the mother, the father, and the metapelet.
Results: Kibbutz infants showed very low avoidance and an elevated rate of insecure-resistant attachment. A stranger’s entry drove much of this: infants rarely met unfamiliar adults in the closed kibbutz community, so the stranger provoked marked distress that inflated apparent resistance. Secure attachment nonetheless remained common.
Conclusion: The collectivist, stranger-averse ecology of the kibbutz raises apparent resistance. This again suggests a ‘cultural’ difference in classification can reflect unfamiliarity with the procedure itself, not the quality of the underlying bond.
Longitudinal cross-cultural research is still needed to track how globalizing trends in child-rearing and family life may reshape these attachment distributions over time.
Strengths
- Sample Size: Aggregating data from 32 samples in 8 countries for a more robust cross-cultural comparison.
- Statistical Methods: Using appropriate categorical data analysis methods, including standardized residuals, chi-square tests, and correspondence analysis.
- Variance Partitioning: Quantifying the relative degrees of intracultural and cross-cultural variation.
- Visualisation: Graphically depicting the similarities and differences among samples.
- Cautious Conclusions: Appropriately cautious conclusions emphasizing the need for more data from non-Western cultures.
Limitations
Ethnocentric Bias
The Strange Situation test assumes that behavior has the same meaning in all cultures when, in fact, cultural perception and understanding of behavior differ greatly.
The Strange Situation was created and tested in the USA. This means it may be culturally biased, or ethnocentric, reflecting the norms and values of American culture.
For example, the belief that attachment is related to anxiety on separation. This may not be the case in other cultures, e.g., Japan.
The concern is not just historical. A direct test compared infant crying during the Strange Situation across Japanese, Korean, US, and Czech samples. East-Asian infants cried substantially more during the separation and stranger episodes, suggesting the procedure does not create equally distressing conditions everywhere (Umemura et al., 2025).
Biased Samples
Many studies have biased samples that cannot be claimed to be representative of each culture, e.g., 36 infants in the Chinese study. Also, most of the studies analyzed were from Western cultures.
Lack of data from South America, Africa, and other less developed regions restricts the global generalizability of the findings.
More recent, larger syntheses have partly addressed this gap. Mesman, van IJzendoorn, and Sagi-Schwartz (2016) extended the evidence base to non-Western, small-scale societies, including the Dogon of Mali and the Efe foragers. Secure attachment remained the most common pattern, and sensitive caregiving still predicted security across these wider samples.
Distributions can vary widely within a country. Caution is needed in assuming an individual sample represents a (sub)culture.
This finding matters more than it first appears. Differences within cultures were about one-and-a-half times larger than differences between cultures. A single national sample cannot safely stand for a whole culture, because cultures are not internally homogeneous. That is a striking result.
Researchers can no longer assume an individual sample represents an entire culture. Differences from the U.S. “standard” do not by themselves show the Strange Situation is culturally invalid.
Trends do exist. Western European samples tend to show more avoidant infants, while Israeli and Japanese samples tend to show more resistant infants. Even so, within-country differences are at least as large as differences between countries. No single sample should stand for a whole culture.
The size of this within-country spread itself varies. It is large in Germany and the USA, for example, but much smaller in the Netherlands and Japan.
Cultures are not one thing. Every culture contains sub-cultures with their own child-rearing styles, shaped by ethnicity, race, or social class. In the UK, for instance, middle-class and working-class families have shown different child-rearing techniques.
Demographic differences between samples, such as socioeconomic status, can also blur comparisons within and between cultures.
These examples show what that spread looks like:
- One of the German samples was as different from another German sample as it was different from an Israeli kibbutzim or from a US sample.
- One of the Japanese samples was more similar to two of the US samples than to the other Japanese sample.
- The Israeli city sample was more like the US than it was to the Israeli kibbutzim sample.
Key Takeaways
- Secure Is Universal: Secure attachment was the most common type in all eight countries studied, supporting the idea of an innate caregiving system.
- Germany: German infants showed unusually high avoidant attachment, likely reflecting a cultural value for early independence rather than poor parenting.
- Japan: Japanese infants showed high resistant attachment because the separation episode was so unfamiliar it caused extreme, possibly unrepresentative, distress.
- Israel: Kibbutz-raised infants showed low avoidance but high resistance, driven by their limited experience of strangers rather than insecure bonds.
- Within > Between: Differences within a single country were about 1.5 times larger than differences between countries, so no one sample can represent a whole culture.
- Modern Evidence: Later research spanning more diverse, non-Western societies confirms secure attachment as the global norm, while still questioning whether a Western procedure measures it fairly everywhere.
References
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. https://doi.org/10.2307/1127388
Grossmann, K., Grossmann, K. E., Spangler, G., Suess, G., & Unzner, L. (1985). Maternal sensitivity and newborns’ orientation responses as related to quality of attachment in northern Germany. Monographs of the Society for Research in Child Development, 50(1–2), 233–256. https://doi.org/10.2307/3333836
Grossmann, K. E., Grossmann, K., Huber, F., & Wartner, U. (1981). German children’s behavior towards their mothers at 12 months and their fathers at 18 months in Ainsworth’s Strange Situation. International Journal of Behavioral Development, 4(2), 157–181. https://doi.org/10.1177/016502548100400202
Mesman, J., van IJzendoorn, M. H., & Sagi-Schwartz, A. (2016). Cross-cultural patterns of attachment: Universal and contextual dimensions. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (3rd ed., pp. 852–877). Guilford Press.
Sagi, A., Lamb, M. E., Lewkowicz, K. S., Shoham, R., Dvir, R., & Estes, D. (1985). Security of infant-mother, -father, and -metapelet attachments among kibbutz-reared Israeli children. Monographs of the Society for Research in Child Development, 50(1–2), 257–275. https://doi.org/10.2307/3333837
Takahashi, K. (1986). Examining the strange-situation procedure with Japanese mothers and 12-month-old infants. Developmental Psychology, 22(2), 265–270. https://doi.org/10.1037/0012-1649.22.2.265
Takahashi, K. (1990). Are the key assumptions of the “Strange Situation” procedure universal? A view from Japanese research. Human Development, 33(1), 23–30. https://doi.org/10.1159/000276500
Umemura, T., Jin, M. K., Kondo-Ikemura, K., Lacinová, L., Handa, K., Xu, Y., & Yoshikawa, K. (2025). Crying in the strange situation procedure: Comparisons between East-Asian and Western infants. International Journal of Behavioral Development, 50(2), 177–185. https://doi.org/10.1177/01650254251325777
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. https://doi.org/10.2307/1130396
