Several studies had used the ** procedure to assess similiarities and differences in patterns of attachment types between cultures BUT these studies had comprised small sample sizes and hadn't considered intra-cultural against inter-cultural differences (whether differences within cultures are greater that those between cultures).
Solution was to perform a meta-analysis of ** studies in different cultures.
AIMS:
To assess whether within separate samples there was a pattern in the distribution of different attachment types.
To asses the extent of of inter (between) and intra (within) cultural differences in attachment types in separate samples.
To assess similarities and differences i the amount of Type A, B & C attachment types in separate samples.
1 of 5
Method & Results:
32 studies from 8 countries that used the ** procedure to assess mother-child attachments (not any other caregiver-infant attachments were analysed in this study) and which classified attachments as either Type A, B or C.
1,990 separate ** classifications were used.
Countries were UK, USA, Sweden, Japan, China, Holland, Germany & Israel.
All studies comprised at least 35 mother-child paris with infants below 2 years of age.
RESULTS:
Overall attachment was:
Type A - 21%
Type B - 67%
Type C - 12%
In samples from 7 of the 8 countries the modal attachment type was Type B, apart from in Germany where it was Type A.
2 of 5
Results Cont. & Conclusions:
Intra-cultural differences (differeces between samples from 1 country) were often greater than inter-cultural differences (differences between countries). E.g. in 1 USA sample there was a 94% Type A attachment rate, whilst in another USA sample there was a 47% Type A attachment rate.
Type A attachment was found more in Western cultures, while Type C attachment was found more in Israel, China and Japan.
CONCLUSIONS:
Data suggests a difference in the pattern of cross-cultural attachment types across cultures.
Intra-cultural differences are often greater than inter-cultural ones.
Overall patterns of attachment types were similar to those Ainsworth found.
An important cross-cultural similarity is the predominance of Type B attachments in all cultures.
3 of 5
Evaluation:
As intra-cultural differences were often found in different samples from the same researcher it suggests such differences were NOT down to methodological differences.
The number of participants in each study varied from country to country, making a convincing comparison difficult.
Not all countries are represented in meta-analysis e.g. Africa and South America, so global data would be required before universal conclusions could be drawn.
Nor are there any "less developed" countries included in this meta-analysis, where lifestyles, opportunities and communities are very different. The 8 selected countries all represent "wealthy" societies.
Some intra-cultural differences may be due to socio-economic differences e.g. some USA samples were of middle-class pairings,whilst others were from poorer socio-economic backgrounds. So even within 1 specific country, attachments did not necessarily follow the usual pattern. The researcher's said this was most likely due to economic factors or stress levels.
Cross-cultural studies, such as this one, can suffer from an imposed etic where researchers analyse findings in a biased manner in terms of their own cultural beliefs, wrongly imposing cultural-specific beliefs onto other cultures. E.g. Ainsworth, an American, assumed that separation anxiety was an indiciation of secure attachment but it may represent something else in other countries and cultures.
4 of 5
Evaluation Cont.
The sample is limited. Are we truly justified in saying that children throughout the world have similar attachments based on 8 countries and 2000 participants? In addition 16 of the 32 studies were American, presenting a strong bias towards western ideals. On this basis we would expect to see mostly secure attachments.
The researcher's believed that the fact that most of the studies they examined showed secure attachments is linked to the effects of mass media spreading a universal concpet of parenting. Although this may be true in part, we know that social learning theory is very influential in children's development, it is less convincing then perhaps political and health interventions in affecting adult choices.
Comments
No comments have yet been made