Cross-Culture meta

• How was the meta-analysis set up?

  • Van Ijzendoorn and Kroonenberg (1988) compared results of the SS from different cultures using a meta-analysis.
  • They only used studies which used the strange situation. Studies were excluded if they used Downs’ syndrome or used less than 35 infants.
  • They looked at 2000 strange situations from 32 studies across 8 countries.
• What was the findings of the meta analysis?

  • They found striking similarities between cultures.
  • Secure was the most common type with 65% on average (very close to the 66% average for Ainsworths’ study itself).
  • Avoidant was the 2nd most common (in Germany it was 35%) except Japan and Israel with high levels of resistant attachment.
  • There was 1.5x more variation within a culture than variation between cultures.
• What was the conclusion of the cross-cultural study?

  • Global pattern shows that the US being secure is the norm
  • This shows idea of healthy development being associated with secure attachment.
  • Mass media may have an impact on parenting views.
• What is a negative evaluation of cross-cultural validity?

  • American only culture
  • The meta-analysis only includes SS studies.
  • A problem with this is that the SS was designed for American infants specifically and might not be valid in other cultures.
  • This is known as imposed etic.
  • Different cultures have different norms and the SS may not translate across these cultures.
  • The USA is considered normal and other cultures may be considered abnormal as a result.
  • Therefore, the SS lacks cross-cultural validity.
• Why may SS lack cross-cultural validity using Israel as an example?

  • Israel was more resistant and less avoidant than the US.
  • They were raised on a kibbutz and saw few strangers but were used to separation.
  • Maybe they were not used to strangers, accounts for the large % of resistant
  • Where their anxiety was not shown when the mother left but when the stranger entered.
• Why may SS lack cross-cultural validity using Japan as an example?


  • Japanese found more resistant but low avoidant similar to Israel but for different reasons.
  • Japanese mothers rarely leave their infants so the SS may have been stressful when the mother left.
  • Japanese children’s resistant behaviour was more likely due to mother leaving than the stranger arriving.
• Why may SS lack cross-cultural validity using Germany as an example?

  • German infants showed large numbers of avoidant, they are encouraged to keep interpersonal distance between caregiver and children.
  • Children are also more encouraged to be independent than in other countries.
  • This reflects being independent rather than insecure.
• What is one reason for internal validity regarding the temperament hypothesis?

  • This may be due to the temperament hypothesis.
  • Some infants may be born more independent or more friendly and their behaviour is due to individual differences.
  • Some infants may get more easily stressed than others so they are more vulnerable to stress and respond differently.
• What is one reason for internal validity regarding the past experience?

  • It also reflects on past experience with separation with the mother and whether the infant is used to strangers.
  • It is supported by Japanese babies who appear resistant as they are with their mothers most of the time and rarely experience separation
  • They get upset and angry when she leaves as a result.
  • This reflects previous experience and not attachment type.
• What is one reason for internal validity regarding the demand characteristics?


  • It may reflect on demand characteristics of the mother.
  • The mother is very well aware of at least part of the aim of the study and observed it with her child.
  • The mother may then interact with the baby more than she usually does as a result.
  • For example, this could also be due to social desirability bias as she wants to be seen as a good mother.
• What is a methodological problem for the meta-analysis?

  • Methodological problems - Half of the studies were done in America.
  • The overall averages are likely to be biased and not representative of all cultures.
  • Sample sizes were too small in some countries like China with 36 infants but having a very big population themselves.
  • The study may lack population validity.