Van Ijzendoorn and Kroonenberg's study used a very large sample of 1990 which increases the internal validity of the findings as it reduces the impact of anomalousresults.
This also makes it easier to drawconclusions and patterns regarding attachmenttypes within countries and betweendifferentcountries as well.
However, one problem is that many studies used in this meta-analysis were not entirelyrepresentative of eachculture. For example, only36 infants were used in the Chinese study, which is a smallsamplesize for suchapopulatedcountry.
Furthermore, the study was notgloballyrepresentative as African, SouthAmerican and Eastern European countries were notrepresented.
This means that the apparentconsistencybetweencultures might not genuinelyreflecthow much attachment types vary between cultures.
This study claimed to study culturalvariation, however, the comparisons were betweencountriesnotcultures. This is a problem because there are manydifferentcultures with differentchild-rearingpracticeswithinacountry and so cultural characteristics can affect caregiving styles and also patterns of attachment.
For example, an analysis by Van Ijzendoorn & Sagi (2001) found that distributions of attachmenttype in Tokyo (an urban setting) were similar to the Westernstudies, whereas a moreruralsample had an over-representation of insecure-resistantindividuals.
This means that studies conducted in differentcountries may not tell us anything about cross-culturalpatterns of attachments.
The study uses the Strange Situation and Ainsworth's classificationofattachmentstyles which is an example of imposedetic as this classification reflects the Westernunderstanding of the concept of attachment.
For example, in Germany, insecure attachment behaviour might be seen more as independence rather than avoidance and therefore it is not a sign of insecuritywithin the culturalcontext.
This means that the behavioursmeasured by the Strange Situation maynot have the samemeaning in different culturalcontexts and comparing them acrosscultures is meaningless.