Schaffer and Emersons stages of attachment

• When was the study conducted?

  1964
• What type of study was it, and what does this mean?

  longitudinal, looking at variables over a long period of time
• what were the stages?

  • Asocial (birth-3 months)
  • Indiscriminate attachment (3-7 months)
  • Specific attachment (7/8 months +)
  • Multiple attachments (9/10 months +)
• What happens during the asocial stage?


  • behaviour towards humans and inanimate objects is fairly similar
  • they respond to other people more though
• what happens during the indiscriminate attachment stage?

  • generally no separation or stranger anxiety
  • develops more response to human company as they can tell people apart
• what happens during the specific attachment stage?

  • strong separation and stranger anxiety
  • primary caregiver attachment is formed
• what happens during the multiple attachments stage?

  • attachment to primary caregiver grows
  • multiple attachments are formed with different people
• who were the pps?

  60 babies from glasgow
• what was the procedure?

  • observed behaviour and kept diary records, interviewed mothers, visited monthly until 12 months old
  • measured attachment by amount of separation and stranger anxiety
• what were the findings?

  • babies who had attentive and responsive caregivers were more likely to have formed an attachment
  • 65% attached to mother only
  • 30% attached to mother + another (often father)
  • 3% attached to father only
• strengths?


  • good external validity - observations mainly made by parents so less chance of different behaviour around psychologist
  • real life application - can be used in child care to better understand infant behaviour
  • good ecological validity - done in naturalistic environment not in a laboratory, so reduced chance of behaviour changing
• limitations?


  • lacks temporal validity - changes since 1960s
  • asocial stage is hard to prove - cannot determine the reasons behind infant behaviour as they are basically immobile
  • lacks population validity - small sample size of 60 babies in glasgow cannot be generalised