SCHAFFER AND EMERSON 1964
Aims
b) identify and describe the ‘distinct stages’ by which attachment develops
a) assess whether there was a ‘pattern’ of attachment formation that was common to all infants
procedure
Babies were studied in their own homes, carers self-reported their observations of their infants’ behaviour using daily diaries and were interviewed about their babies’ responses
Evidence for the development of an attachment was recorded when the baby showed ‘separation anxiety’ after a carer left, and ‘stranger anxiety’ when the researcher approached.
They were studied at monthly intervals for the first year of their life, then again at 18 months.
conducted a longitudinal study on 60 new-born babies and their mothers from a working-class area of Glasgow
findings
most infants went to develop multiple attachments. at 18 months, 87% had at least two attachments
attachments to different people were of a similar nature
strongly attached infants had mother who responded to their needs quickly and gave more opportunities for interaction
39% of infants prime attachment was not to the main carer ie. the person who fed and bathed them
infants showed separation protest when parted from their attachment figure at between 6-8 months, stranger anxiety about one month later
conclusion
there is a pattern common to all infants, suggest the process is biologically controlled
attachments are more easily made with those who display sensitive reponsiveness
multiple attachments are the norm and of similar quality
strength
conducted under everyday conditions, meaning conclusions can been seen as having high ecological validity
most observations were made by the parents during ordinary activities, means that babies behaved naturally while being observed
high mundane realism
weakness
the data was biased as it was based on mothers reports of their infants
may have been less likely to report their less responsive behaviour, want to be seen in a positive light
unreliable data
social desirability bias
undermines the validity of the data
biased sample
lacks population validity
lacks temporal validity
working class background, may not apply to other social groups
child rearing practices have changed
more women go out to work
therefore, findings might be different if a similar study was conducted today