:) goodpredictivevalidity - research evidence supporting the idea that attachmenttypedoespredictlaterdevelopment - e.g. securelyattached babies tend to have greatersuccess and more lastingromantic relationships
:) goodinter-raterreliability - the controlledconditions and clearbehaviouralcategories lead to highlevels of agreement between researchers e.g. Bick et al - found 94% agreement in one team - we can be confident that attachment type selected doesn't depend upon who is observing the child
lacks IRLapplication in identifying quality of attachment - research suggests that some children don't fit into 3 attachment types e.g. main and solomon suggest some children display 'disorganisedattachment,' attachment types may be invalid
lacks generalisability - strange situation only measures attachment in USA/westerneurope - germanbabies are encouraged to be independent when they play and explore so are likely to appear as type A in the strange situation