Abstract
This paper described current infant and early childhood attachment research and assessment procedures. First, the limitations of Strange Situation procedure, which has been used frequently over the past twenty years, were discussed in relation to several issues. These issues included the optimal assumption of the B-type attachment, causes of variation in attachment classification, predictive validity for later social competence, and a reconceptualization of attachment from a life-span perspective. Overall it seemed very doubtful that the Strange Situation is a valid method for assessing Japanese infant-mother attachment. As an alternative approach, Q-sorts present several advantages for assessment in accordance with current Japanese research. Thus the attachment Q-sort will be valuable in future research.
