Cultural variations in attachment

Van Ljzendoorn and Kroonenberg's research

- Van Ljzendoorn and Kroonenberg (1988) conducted a study look at the proportions of secure, insecure-avoidant and insecure-resistant attachments across a range of countries to assess cultural variation.

- They also looked at the differences within the same countries to get an idea of variations within a culture

Procedure

- The researchers located 32 studies of attachment where the Strange Situation had been used to investigate the proportions of babies with different attachment types. These were conducted in eight countries- 15 were in the US.

- Overall, the studies yielded results for 1,990 children. The data for these 32 studies was meta-analysed, meaning the results of the studies were combined and analysed together, weighting each study for its sample size

Findings

- In all countries secure attachment was the most common classification. However the proportion varied from 75% in Britain to 50% in China. In individualist cultures rates of insecure-resistant attachment were similar to Ainsworth's original sample (all under 14%) but this was not true for the collectivist samples from China, Japan, and Israel where rates were above 25% (and where rates of insecure-avoidant attachment were reduced)

- Variations between results of studies within the same country were 150% greater than those between countries. In the US, one study found only 46% securely attached compared to one sample as high as 90%

Conclusion

- Secure attachment seems to be the norm in a wide range of cultures, supporting Bowlby's idea that attachment is innate and universal and this type is the universal norm. However, the research also clearly shows that cultural practices have an influence on attachment type

Evaluation- Strengths

- Most studies were conducted by psychologists who were from the same cultural background as the participants. For example, van lJzendoorn and Kroonenberg included research by a German team (Grossmann et al 1981) and Takahashi (1986) who is Japanese. This kind of research means that many of the potential problems in cross-cultural research can be avoided, such as researchers' misunderstanding of the language used by participants or having difficulty instructing them. Difficulties can also include bias because of stereotypes

- This means there is an excellent chance that researchers and participants communicated successfully- enhancing the validity of the data collected

Biassed?

Evaluation- Weaknesses (1)

- There is the impact of confounding variables on findings. Studies conducted in different countries are not usually matched for methodology when they are compared in reviews or meta-analyses. Sample characteristics such as poverty, social class, and urban/rural make-up can found results as can the age of participants studied in different countries. Environmental variables might also differ between studies and confound results. For example the size of the room and the availability of interesting toys there- babies might appear to explore more in studies conducted in small rooms with attractive toys. Less visible proximity-seeking because of room size might make a child more likely to be classified as avoidant

- This means that looking at attachment behaviour in different non-matched studies conducted in different countries may not tell us anything about cross-cultural patterns of attachment

Evaluation- Weaknesses (2)

- It is trying to impose a test designed for one cultural context to another context. This is imposed etic, which is when we assume an idea or technique that works in one cultural context will work in another. An example of this in attachment research is in the use of babies' responses to reunion with the caregiver in the Strange Situation. In Britain and the US, lack of affection on reunion may indicate an avoidant attachment. But in Germany such behaviour would be more likely interpreted as independence rather than insecurity.

- This means that the behaviours measured by the Strange Situation may not have the same meanings in different cultural contexts, and comparing them across cultures is meaningless

Weakness 3

Samples tend to be unrepresentative if cultures-The meta analysis conducted by Ijzendoorn & Kroonenberg claimed to study cultural variation but in reality comparisons were between countries not cultures.Each country contains many different cultures.This can lead to over or under representation of the proportion of attachment types.This means that comparisons between countries are invalid and have to be more specific with particular cultural characteristics.