Evaluations - A03

• Strength of strange situation - observations have high reliability. EG Ainsworth found almost perfect agreement when rating exploratory behaviour of infants during observation with a 0.94 correlation between observers. Demonstrates that observation had high inter-rater reliability and almost unanimous agreement among observers about behaviour of infants. Therefore strength as observations can be subjective and therefore having high inter-rater reliability crucial as we can be confident in reliability.
• Limitation of SS - not apply to non-western countries. Takahashi noted test does not work in Japan because Japanese mothers rarely separated from their children. A mother leaving her child with stranger not a realistic day to day scenario, issue as different childhood experiences mean children from different cultures react differently to strange situation or caregivers from these cultures behave differently in such situations. The SS reflect norms of American culture and not consider differences in child rearing practices in other cultures making it ethnocentric, reduces population validity
• Limitation of strange situation - later research shows fourth type of attachment. Main and Solomon analysed over 200 strange situation types and proposed type D - insecure disorganised. Characterised by lack of consistent social behaviour and attachment, these infants lacked coherent strategy of dealing with stress of separation, showed very strong attachment suddenly followed by avoidance or looking scared of caregiver. Since Ainsworth's research did not account for this type of attachment, research seen as lacking sufficient detail, thus reduces validity of findings.
• Limitation of strange situation - doesn't account for fact that infants may have different attachments with different caregivers. EG SS aimed to measure attachment type of child however researchers claim that it is too focused on one relationship between carer (usually mother) and infant being assessed. Main and Weston found children behaved differently based on which parent they were with. Therefore while believing she is measuring child's general attachment tendencies, might have been assessing quality of relationship between mother and child. Therefore reducing internal validity.
• Strength of cultural variations - combining results of attachment studies carried out in different countries is that one can have a large sample. For example Ijzendoorn and Kroonenberg's meta-analysis had a total of 1990 infants and their primary attachment figures from 32 studies, strength as large sample reduces impact of anomalous results from bad methodology on overall conclusions drawn. Therefore increases internal validity of study.
• Limitation of CV (1) - while sample for Ijzendoorn and Kroonenberg's study large enough to rule out effects of anomalous results, there is imbalance of studies used. EG still quite biased towards having heavily American population. Out of 32 studies in meta-analysis, 18 conducted on American participants and only one conducted in China with 25 infants. 5 out of 32 conducted in non-western countries with collectivistic culture and remaining 27 conducted in individualistic cultures.
• Limitation of CV (1) - Problematic as it questions accuracy of drawing conclusions about Chinese culture based on only 25 parent infant pairs, while there were hundreds of such pairs to draw balanced conclusions for a country which has a significantly smaller population than China. Therefore meta-analysis not fairly representative of all cultures, reducing validity of its findings.
• Limitation of CV (2) - using strange situation to investigate cultural variations in attachment is that method of assessment if culturally biased. For example since the strange situation has been designed by American researcher and is based on British theory (Bowlby) there is a wide debate on whether such western theories and assessments can be applied to other cultures. Issue as assumption that theory designed for one culture can be applied to another is imposed etic as well as ethnocentric and leads to biased results in other cultures.
• Limitation of CV (2) - For example lack of separation anxiety in German children (as reported by Grossman and Grossman) might be perceived more as desirable independence rather than a sign of insecurity. Therefore strange situation is culturally biased and not suitable method to assess attachment in all cultures reducing validity of research into cultural variations.