one limitation is that caregiver-infant interactions are not found in all cultures.
Le Vine et al reported that Kenyan mothers have little physical interactions and contact with their infants, yet the infants have a high proportion of secure attachment.
Therefore the research may be ethnocentric and ignores how attachments may be formed in different cultures and that the infant doesn't have to imitate the caregiver to form an attachment.
This reduces the validity of the research as all cultures should be considered in order to generalise as much as possible.