Caregiver-infant interactions

From an early age, babies have meaningful and soial interactions with carers. It's believed that these interactions have important functions for the child's social development, in particular for the development of the caregiver-infant attachment. 


  • A decsription of how two people interact. Mother-infant interaction is reciprocal in both infant and mother respond to each other's signal and elicits a response from the other. 
  • From around 3 months this interaction tends to be increasingly frequent and involves close attention to each other's verbal signals and facial expressions (Feldman 2007). A key element of this interaction is reciprocity.
  • Traditional views of childhood have seen the baby in a passive role, receiving care from an adult. However, it seems that the baby takes an active role. Both mother and child can initiate interactions and they appear to take turns in doing so. Brazelton et al (1975) described this interaction as a 'dance' because it is just like a couple's dance where each partner responds to each other's moves.
1 of 66

Caregiver-infant interactions

Interactional synchrony

  • Mother and infant reflect both the actions and emotions of the other and do this in a co-ordinated (synchronised) way.
  • Meltzoff and Moore (1977) observed the beginnings of interactional synchrony in infants as young as two weeks old. An adult displayed one of tree facial expressions or one of three distinctive gestures. The child's response was filmed and identified by independent observers. An association was found between the expression or gesture the adult had displayed and the actions of the babies. 
  • It is believed that interactional synchrony is important for the development of mother-infant attachment. Isabella et al (1989) observed 30 mothers and infants together and assessed the degree of synchrony. The researchers also assessed the quality of mother-infant attachment. They found that high levels of synchrony were associated with better quality mother-infant attachment.
2 of 66

Caregiver-infant interactions - Evaluation

It is hard to know what is happening when observing infants

Many studies involving observation of interactions between mothers and infants have shown the same patterns of interaction (Gratier 2003). However, what is being observed is merely hand movements or changes in expression. It is exteremely difficult to be certain, based on these observations, what is taking place from the infant's perspective is. We cannot know for sure that behaviours seen in mother-infant interaction have a special meaning.

Controlled observations capture fine detail

Observations of mother-infant interactions are generally well-controlled procedures, with both mother and infant being filmed, often fom multiple angles. This ensures that very fine details of behvaviour can be recorded and later analysed. Futhermore babies don't know or care that they are being observed so their behaviour does not change in response - which is generally a problem for observational research.

This is a strength of this line of research because it means the research has good validity.

3 of 66

Caregiver-infant interactions - Evaluation

Observations don't tell us the purpose of synchrony and reciprocity

Feldman (2012) points out that synchrony simply describe behaviours that occur at the same time. These are robust phenomena in the sense that they can be reliably observed, but this may not be particularly useful as it does not tell us their purpose. 

However, there is some evidence that reciprocal interacion and synchrony are helpful in the development of mother-infant attachment, as well as helpful in stress responses, empathy, language and moral development. 

4 of 66

What is an attachment

An attachment can be deifned as a close two-way emotional bond between two individuals in which each individual sees the other as essential for their own emotional security. Attachment in humans takes a few months to develop. We can recognise an attachment when people display the following behaviours:

  • Proximity - people try to stay physically close to those to whom they are attached.
  • Separation distress - people are distressed when an attachment figure leaves their presence.
  • Secure-base behaviour - even when we are independent of our attachment figures we tend to make regular contact with them. Infants display secure-based behaviour when they regularly return to their attachment figure while playing.
5 of 66

Attachment figures

Parent-infant attachments

Schaffer and Emerson (1964) found that the majority of babies did become attached to their mother first and within a few weeks or moths formed secondary attachments to other family members, including the father. In 75% of the infants studied, an attachment was formed with the father by the age of 18 months. This was determined by the fact that the infants protested when their father walked away - a sign of attachment.

The role of the father

Grossman (2002) carried out a longitudinal study looking at both parent's behaviour and its relationship to the quality of children's attachments into their teens. Quality of infant attachment with mothers but not fathers was related to children's attachments in adolescence suggesting that father attachment was less important.

However, the quality of fathers' play with infants was related to the quality of adolescence attachments. This suggests that fathers have a different role in attachment - one that is more to do with play and stimulation, and less to o with nurturing. 

6 of 66

Attachment figures

Fathers as primary carers

There is some evidence to suggest that whn fathers do take on the role of being the main caregiver they adopt behaviours more typical of mothers. Tiffany Field (1978) filmed 4-month old babies in face-to-face interaction with primary caregiver mothers, secondary caregive fathers and primary caregiver fathers.

Primary caregiver fathers, like mothers, spent more time smiling, imitating and holding infants than the secondary caregiver fathers. This behaviour appears to be important in building an attachment with the infant. So it seems that fathers can be the more nurturing attachment figure. They key to the attachment relationship is the lev el of responsiveness not the gender of the parent.

7 of 66

Attachment figures - Evaluation

Inconsistent findings on fathers 

Research into the role of fathers in attachment is confusing because different researchers are interested in different research questions. On one hand, some psychologists are interested in understanding the role fathers have as secondary attachment figures whereas others are more interested in the father as the primary attachment figure. The former have tended to see fathers behaving differently from mothers and having a distict role. The latter have tended to find that fathers can take on a maternal role.

If fathers have a distinct role why aren't children without fathers different?

The study by Grossman found that fathers as secondary attachment figures had an important role in their children's development. However, other studies have found that children growing up in single or same-sex parent families do not develop any differently for tose in two-parent heterosexual families. Suggests that father's role as secondary attachment figure is not important.

8 of 66

Attachment figures - Evaluation

Why don't fathers generally become primary attachments?

The fact that fathers tend not to become the primary attachment figure could simply be the result of traditional gender roles, in which women are expected to be more caring and nurturing than men. Therefore fathers simply don't feel they should act like that.

On the other hand, it could be that female hormones create higher levels of nurturing and therefore women are biologically pre-disposed to be the primary attachment figure.

Working mothers 

Research into mother-infant interaction is socially sensitive because it suggests that children ay be disadvantaged by particular child-rearing practices. In particular, mothers who retur to work shortly after a child is born restrict opportunities for achieving interactional synchrony, which Isabelle et al showed to be important in the developing infant-caregiver attachment. This suggests that mothes should not return to work so soon and has socially sensitive implications.

9 of 66

Key study: Schaffer and Emerson (1964)

Aimed to investigate the formation of early attachments; in particular the age at which they developed, their emotional intensity and to whom they were directed.


  • 60 babies - 31 male, 29 female. All from Glasgow & working class.
  • Babies & mothers visited at home every month for the first year and again at 18 months later.
  • Researchers asked the mothers quetions about the kind of protest their babies showed in 7 everyday seperations. 
  • Designed to measure the infant's attachment. They also assessed stranger anxiety.


  • Between 25-32 weeks of age about 50% of the babies showed signs of seperation anxiety towards an adult, usually the mother (specific attachment).
  • Attachment tended to be to the caregiver who was most interactive and sensitive to infant signals and facial expressions (reciprocity). Not the person who they spend the most time with.
  • By 40 weeks 80% of the babies had a specific attachment and 30% had multiple As.
10 of 66

Scaffer and Emerson - Evaluation

Good external validity

Study carried out in families' own homes and most of the observation was actually done by parents during ordinary activities and reported to researchers later. This means that the behaviour of the babies was unlikely to be affected by the presence of observers. 

Longitudinal design

The same children were followed up and observed regularly. The quicker alternative would have been to observe different children at each age. This is called a cross-sectional design. However, the longitudinal designs have better internal validity than cross-sectional designs because they do not have the confounding variable of individual differences between participants.

Limited sample characteristics

The fact that all the families involved were from the same district and same social class in the same city and at a time over 50 years ago is a limitation. The results are not generalisable.

11 of 66

Stages of attachment

Scaffer and Emerson proposed that attachments develop in 4 stages.

Stage 1: Asocial (first few weeks)

  • Baby recognises and forms bonds with its carers. However, the baby's behaviour towards non-human objects and humans is quite similar. Babies show some preference for familiar adults in that those individuals find it easier to calm them. Babies are also happier when in the presence of other humans. 

Stage 2 (Indiscriminate attachment)

  • From 2-7 months babies display more observable social behaviour. Preference for people rather than indiscriminate objects, and recognise & prefer familiar adults. At this stage babies usually accept cuddles & comfort from any adult, and they do not usually show seperaton anxiety or stranger anxiety. Their attachment behaviour is therefore said to be indiscriminate because it is not different towards any one person.
12 of 66

Stages of attachment

Stage 3 (Specific attachment)

  • From around 7 months the majority of babies start to display anxiety towards strangers and to become anxious when seperated from one particular adult. Baby has formed a specific attachment. This adult is termed the primary attachment figure. 

Stage 4 (Multiple attachments)

  • Shortly after babies start to show attachment behaviour towards one adult they usually extend this attachment behaviour to multiple attachments with other adults with whom they regularly spend time. These relationships are called secondary attachments. In Schaffer and Emerson's study, 29% of the children had secondary attachments within a month of forming a primary attachment. By the age of 1 year the majority of infants had developed multiple attachments.
13 of 66

Stages of attachment - Evaluation

Problem studying the asocial stage

The problem is that babies that are young have poor co-ordination and are generally pretty much immobile. It is therefore very diffucult to make any judgements about them based on observations of their behaviour. There isn't much actual observable behaviour. This doesn't mean that the child's feelings and cognitons are not highly social but the evidence cannot be relied on.

Conflicting evidence on multiple attachments

Some research seems to indicate that most if not all babies form attachments to a single main carer before they become capable of developing multiple attachments (Bowlby 1969).

Other psychologists, believe babies form multiple attachments from the outset (Van Ijzendoorn et al 1993). Such cultures are called collectivist because families work together in child rearing.

14 of 66

Stages of attachment - Evaluation

Measuring multiple attachment

Just because a baby gets distressed when an individual leaves the room does not necessarily mean that the individual is a 'true' attachment figure. Bowlby (1969) pointed out that children have playmates as well as attachment figures and may get distressed when a playmate leaves the room but this does not signify attachment.

This is a problem for Schaffer and Emerson's stages because their observation does not leave us a way to distinguish between behaviour shown towards secondary attachment figures and shown towards playmates.

Scaffer and Emerson used limited behavioural measures of attachment

Scaffer and Emerson were able to carry out a scientific study of attachment development because they used simple behaviours - stranger anxiety and seperation anxiety - to define attachment. Some critics believe these are too crude as measures of attachment.

15 of 66

Animal studies of attachment

Animal studies in psychology are studies carried out on non-human animal species rather than on humans, either for ethical or practical reasons - practical because animals breed faster and researchers are interested in seeing results across more than one generation of animals. 

16 of 66

Lorenz's research


Lorenz first observed the phenominon of imprinting when he was a child and a neighbour gave him a newly hatched duckling that then followed him around.

Procedure as an adult researcher Lorenz set up a classic experiment in which he randomly divided a clutch of goose eggs. Half the eggs were hatched with the mother goose in the natural enviroment. The other half hatched in an incubator where the first moving object they saw was Lorenz.

Findings The incubator group followed Lorenz everywhere whereas the control group hatched in the presence of their mother, followed her. When the two groups were mixed up the control group continued to follow the mother and the experimental group followed Lorenz.

This phenomenon is called imprinting - whereby bird species that are mobile from birth attach to and follow the first moving object they see. Lorenz identified a critical period in which imprinting needs to take place. Depending on the species this can be as brief as a few hours after hatching. If imprinting does not occur within that time Lorenz found that chicks did not attach themselves to a mother figure.

17 of 66

Sexual imprinting

Lorenz also investigated the relationship between imprinting and adult mate preferences. He observed that birds that imprinted on a human would often later display courtship behaviour towards humans. In a case study Lorenz (1952) described a Peacock that had been reared in the reptile house of a zoo where the first moving objects the peacock saw after a hatching were giant tortoises. As an adult this bird would only direct courtship behaviour towards giant tortoises. Lorenz concluded that this meant he had undergone sexual imprinting. 

18 of 66

Lorenz - Evaluation

Generalisability to humans

Lorenz was interested in imprinting in birds. Although some of his findings have influenced our understanding of human development, there is a problem in generalising from findings on birds to humans. It seems that the mammalian attachment system is quite different from that in birds. This means that it is not appropriate to try to generalise any of Lorenz's ideas to humans.

Some of Lorens's observations have been questioned

Later researchers have questioned some of Lorenz's conslusions. Guiton et al (1966) found that chickens that imprinted on yellow washing up gloves would try to mate with them as adults, but that with experience they eventually learned to prefer mating with other chickens.

This suggests that the impact of imprinting on mating behaviour is not as permanent as Lorenz believed. 

19 of 66

Harlow's research

The importance of contact comfort

Harlow observed that newborns kept alone in a bare age usually died but that they actually survived if something soft to cuddle.

Procedure: Harlow (1958) tested the idea that a soft object serves some of the functions of a mother. In one experiment he reared 16 baby monkeys with two wire model 'mothers'. In one condition milk was dispensed by the plain wire mother whereas in second condition was milk ws dispensed by the cloth-covered mother. 

Findings: It was found that the baby monkeys cuddles the soft object in preference to the wire one and sought comfort form the cloth one when frightened regardless of which dipensed milk. This showed that 'contact confort' was one of more importance to the monkeys than food when it came to attachment behaviour. 

20 of 66

Harlow's research

Maternally deprived monkeys as adults

Harlow and colleagues also followed the monkeys who had been deprived of a 'real' mother into adulthood to see if this early maternal deprivation had a permanent effect. The researchers found severe consequences. The monkeys reared with wire mothers only were the most dusfunctional; however, even those reared with a soft toy as a substitute did not develop normal social behaviour. They were more aggressive and less sociable than other monkeys and they bred less often than is typical for monkeys, being unskilled at mating. As mothers some of the deproved monkeys neglected their young and others attacked their children, even killing them in some cases.

The critical period for normal development

Like Lorenz, Harlow concluded that there was a critical period for this behaviour - a mother figure had to be introduced to an infant monkey within 90 days for an attachment to form. After this time attachment was impossible and the damage done by early deprivation became irreversible.

21 of 66

Harlow - Evaluation

Theoretical value

Harlow's findings have had a profound effect on psychologists' understanding of human mother-infant attachment. Most importantly Harlow showed that attachment doesn't develop as the result of being fed by a mother figure but as a result of contact comfort. Harlow also showed us the importance of the quality of early relationships for later social development including the ability to hold down adult relationships for later social development including the ability to hold adult relationships and successfully rear children. 

Practical value 

The insight into attachment from Harlow's research has had important application in a range of practical contexts; eg social workers understand risk factors in child neglect and abuse and so intervene to prevent it, the findings can also be applied in schools. 

Ethical isssues

The monkeys suffered greatly as a result of Harlow's procedures. The species is similar enough to humans to generalise which is why they had such a human like response. 

22 of 66

Learning theory and attachment

A set of theories from the behaviourist approach to psychology, that emphasise the role of learning in the acquisition of behaviout. Expanations for learning of behaviour include classical and operant conditioning.

Learning theorists Dollard and Miller (1950) proposed that caregiver-infant attachment can be explained by learning theory. Their approach is sometimes called 'cupboard love' approach because it emphasises the importance of the caregiver as a provider of food. Simply, they proposed that children love whoever feeds them.

This includes: 

  • classical conditioning
  • operant conditionign
  • attachment as a secondary drive.
23 of 66

Classical conditioning

Classical conditioning involes learning to accosiate two stimuli together so that we begin to respond to one in the same way as we already respond to the other. In the case of attachment, food serves as an unconditioned stimulus. Being fed gives us pleasure - we don't have to learn it.

A caregiver starts as a neutral stimulus, i.e. a thing that produces a neural response. When the same person provides the food over time they become associated with 'food' - when the baby sees this person there is an immediate expectation of food. The neutral stimulus has become a conditioned stimulus. Once conditioning has taken place the sight of the caregiver produces a conditioned response of pleasure. 

24 of 66

Operant conditioning

Operant conditioning involevs learning to repeat behaviour, not not, depending on its consequences. If a behaviour produces a pleasant consequence, that behaviour is likely to be repeated again. The behaviour has been reinforced. If a behaviour produces an unpleasant consequence it is less likely to be repeated

Operant conditioning can explain why babies cry for comfort - an important behaviour in building attachment. Crying leads to a response from the caregiver for example feeding. As long as the caregiver provides the correct response, crying is reinforced. The baby then directs crying for comfort towards the caregiver who responds with comforting 'social surpressor' behaviour. 

This reinforcement is a two-way process. At the same time as the baby is reinforced for crying, the caregiver recieves negative reinforcement becaused the crying stops - escaping from something unpleasant is reinforcing. The interplay of mutual reinforcement strengthens an attachment. 

25 of 66

Attachment as a secondary drive

As well as conditioning, learning theory draws on the concept of drive reduction. Hunger can be thought of as a primary drive - it's an innate, biological motivator. We are motivatedd to eat in order to reduce the hunger drive.

Sears et al (1957) suggested that, as caregivers provide food, the primary drive of hunger becomes generalised to them. Attachment is thus a secondary drive learned by an association between the caregiver and the satisfaction of a primary drive.

26 of 66

Learning theory and attachment - Evaluation

Counter-evidence from animal research

  • Young animals do not necessarily attach to those who feed them. Lorenz's geesde imprinted before they were fed and maintained these attachment figures regardless of who fed them. Harlow's monkeys attached to a soft surrogate in preference to a wire one that dispensed milk.

Counter-evidence from human research

  • Research with human infants also shows that feeding does not appear to be an important factor in humans. For example, in Schaffer and Emerson's study many of the babies developed a primary attachment to their biological mother even though other carers did most of the feeding. 

Learning theory ignores other factors associated with forming attachments

  • Research into early infant-caregiver interaction suggests that the quality of attachment is associated with factors like developing reciprocity and good levels of interactional synchrony. It's hard to reconcile these findings with the idea of cupboard love. It attachment developed primarily due to feeding, there would be no purpose for these complex interactions.
27 of 66

Learning theory and attachment - Evaluation

Some elements of conditioning could still be involved

We do believe that many aspects of human development are affected by conditioning. The problem with learning theory as an explanation for attachment is mostly the idea that feeding provides the unconditioned stimulus, reinforcement or primary drive. It is still credible that association (clasical conditioning) between the primary caregiver and the provision of comfort and social interaction is part of what builds the attachment. 

A newer learning theory explanation

Hay and Vespo (1998) have proposed a newer explanation for infant-caregiver attachment based on SLT. SLT is based on the idea that social behaviour is acquired largely as a result of modelling and imitation of behaviour. Hay and Vespo suggest that parents teach children to love them by modelling attachment behaviour eg. hugging other family members, and instructing and rewarding them with approval when they display attachment behaviour of their own.

28 of 66

Bowlby's monotropic theory

John Bowlby rejected learning theory as an explanation for attachment because. Instead he looked at the work of Lorenz and Harlow for ideas and proposed an evolutionary explanation: that attachment was an innate system that gave a survival advantage. Imprinting and attachment evolved because they ensure that young animals stay close to their caregivers and this protects them from hazards. Millions of years ago this might have been wild animals, today it is traffic and electricity. 

29 of 66


Monotropic - A term sometimes used to describe Bowlby's theory. The mono means 'one' and indicates that one particular attachment is different from all others and of central importance to the child's development.

Bowlby believed that the more time a baby spent with their mother-figure - or primary attachment figure as we usually call them - the better. He put forward two principles to clarify this:

  • The law of continuity stated that the more constant and predictable a child's  care, the better the quality of their attachment.
  • The law of acumulated seperation stated that the effects of every separation from the mother add up 'and the safest dose is therefore a zero dose'.
30 of 66

Social releasers and the critical period

Bowlby suggested that babies are born with a set of innate 'cute'  behaviours like smiling, cooling and gripping that encourage attention from adults. He called these social releasers because their purpose is to activate the adult attachment system, i.e. make an adult love them. Bowlby recognised that attachment was a reciprocal process. Both mother and baby have an inate predisposition to become attached and social releasers trigger that response in caregivers. 

The interplay between infant and adult attachment systems gradually builds the relationship between infant and caregiver, beginning in the early weeks of life. Bowlby proposed that there is a critical period around two years when the infant attachment system is active. In fact Bowlby viewed this as more of a sensitive period. A child is maximally sensitive at the age of two but, if an attachment is not formed in this time, a child will find it much harder to form one later. 

Critical period - this refers to the time within an attachment must form if it is to form at all. Lorenx and Harlow noted that an attachment in birds and monkeys had critical periods. Bowlby extended the idea to humans, proposing that human infants have a sensitive period after which it will be much more difficult to form an attachment

31 of 66

Internal working model

Bowlby proposed that a child forms a mental representation of their relationship with their primary caregiver. This is called an internal working model because it serves as a model for what representations are like. It therefore has a powerful effect on the nature of the child's relationships. A child whose first experience is of a loving relationship with a reliable caregiver will tend to form an expectations that all relationships are as loving and reliable, and they will bring these qualities to future relationships. However, a child whose first relationship involves poor treatment will tend to form further poor relationships in which they expect such treatment from others in that way. 

Most importantly the internal working model affects the child's later ability to be a parent themselves. People tend to base their parenting behaviour on their own experiences of being parented. This explains why children from functional families tend to have similar families themselves.

32 of 66

Bowlby's monotonic theory - Evaluation

Mixed evidence for monotropy

Bowlby believed that babies generally formed one attachment to their primary caregiver, and that this attachment was special, in some way different from later attachments. Only after this attachment was established could a child form multiple attachments. This is not supported by Schaffer and Emerson (1964). As they found most babied did attach to one person first. However, they also found that a significant minority appeared able to form multiple attachments at the same time.

It is also unclear whetehr there is something unique about the first attachment. Studies of attachment to mother and father tend to show that attachment to the mother is more important in predicting later behaviour. However, this could simply mean that attachment to the primary attachment figure is just stronger than other attachments, not necessarily that is different in quality.

33 of 66

Bowlby's monotonic theory - Evaluation

Mixed evidence for monotropy

Bowlby believed that babies generally formed one attachment to their primary caregiver, and that this attachment was special, in some way different from later attachments. Only after this attachment was established could a child form multiple attachments. This is not supported by Schaffer and Emerson (1964). As they found most babied did attach to one person first. However, they also found that a significant minority appeared able to form multiple attachments at the same time.

It is also unclear whetehr there is something unique about the first attachment. Studies of attachment to mother and father tend to show that attachment to the mother is more important in predicting later behaviour. However, this could simply mean that attachment to the primary attachment figure is just stronger than other attachments, not necessarily that is different in quality.

34 of 66

Bowlby's monotonic theory - Evaluation

Support for social releasers

There is clear evidence to show that cute infant behaviours are intended to initiate social interaction and that doing so is important to the baby,. Brazelton et al (1975) observed mothers and babies during their interactions, reporting the existance of interactional synchrony. They then extended the study from an observation to an experiment. Primary attachment figures were instructed to ignore their babies' signals - in Bowlby's terms, to ignore their social releasers. The babies initially showed some distress but, when the attachment figures continued to ignore the baby, some responded by curling up and lying motionless.

The fact that children responsed so strongly supports Bowlby's ideas about the significance of infant social behaviour in elicitibg caregiver. 

35 of 66

Bowlby's monotonic theory - Evaluation

Support for internal working models 

The idea of internal working models is testable because it predicts that patterns of attachment will be passed on from one generation to the next. Bailey et al (2007) tested this idea. They assessed 29 mothers with one-year-old babies on the quality of their attachment to their own mothers using a standard interview procedure. The researchers also assessed the attachment of the babies to the mothers by observation. It was found that the mothers who reported poor attachments to their own parents in the interviews were much more likely to have children classified as poor according to the observation.

This supports the idea that, as Bowlby said, an internal working model of attachment was being passed through the families. 

36 of 66

Bowlby's monotonic theory - Evaluation

Monotropy is a socially sensitive idea

Monotropy is a controversial idea because it has major implications for the lifestyle choices mothers make when their children are young. The law of accumulated separation staters that having substantial time apart from a primary attachment figure risks a poor quality attachment that will disadvantage the chid in a range of ways later. Feminists like Burman (1994) have pointed out that this places a terrible burden of responsibility on mothers, setting them up to take the blame for anything that goes wrong in the rest of the child's life. It also pushes mothers into particular lifestyle choices like not returning to work when a child is born. 

Temperament may be as important as attachment

Bowlby's approach emphasises the role of attachment in the child's developing social behaviour. However, a different tradition of child development emphasised the role of termperament in the developnent of social behaviour. Temperament is the child's genetically influenced personality. Temperament researchers often accuse Bowlby of over-emphasising the importance of a child's early experiences and the quality of their attachment. 

37 of 66

The strange situation - Procedure

The strange situation was developed by Ainsworth (1969). The aim was to be able to observe key attachment behaviours as a means of assessing the quality of a child's attachment to a caregiver. 

Strange situation - A controlled observation designed to test attachment security. Infants are assessed on their response to playing in an unfamiliar room, being left alone, left with a stranger and being reunited with a caregiver. It takes place with a two-way mirror through which psychologists can observe the infant's behaviour.

The behaviours used to judge attachment included:

  • Proximity seeking - an infant with a good attachment will stay fairly close to the caregiver. 
  • Exploration and secure-base behaviour - good attachment enables a child to feel confident to explore, using their caregiver as a secure base, i.e. a point of contact that will make them feel safe. 
  • Stranger anxiety - one of the signs of becoming closely attached is a display of anxiety when a stranger approaches.
  • Separation anxiety - another sign of becomng attached is to protest at separation from the caregiver. 
  • Response to reunion - with the caregiver after a short period of separation.
38 of 66

The strange situation - Procedure

The procedure has seven episodes, each of which last three minutes.

  • Beginning: child and caregiver enter an unfamiliar playroom.
  • Child encouraged to explore - tests exploration and secure base. 
  • A stranger come in and tries to interact with the child - tests stranger anxiety.
  • The caregiver leaves the child and stranger together - tests separation and stranger anxiety.
  • The caregiver returns and the stranger leaves - tests reunion behaviour and exploration/ secure base. 
  • The caregiver leaves the child alone - tests separation anxiety
  • The stranger returns - tets stranger anxiety
  • The caregiver returns and is reunited with the child - tests reunion behaviour.
39 of 66

The strange situation - Findings

Ainsworth et al (1978) found that there were distinct patterns in the way that infant's behaved. She identified three main types of attachment:

  • Secure attachment (Type B) - These children explore happily but regularly go back to their caregiver (proximity seeking and secure base behaviour). They usually show moderate separation distress and moderate stranger anxiety. Securely attached children require and accept comfort from the caregiver in the reunion stage. About 60-75% of British toddlers are classified as secure. 
  • Insecure-avoidant attachment (Type A) - These children explore freely but do not seek proximity or show secure base behaviour. They show little or no reaction when their caregiver leaves and they make little effort to contact when the caregiver returns. They also show little stranger anxiety. They do not require comfort at the reunion stage. About 20-25% of toddlers are classified as insecure-avoidant.
  • Insecure-resistant attachment (Type C) - These children seek greater proximity than others and so explore less. They show huge stranger and separation distress but they resist comfort when reunited with their carer. Around 3% of British toddlers are classified as insecure-resistant. 
40 of 66

Strange situation - evaluation

Support for validity - Attachment type as defned by the Strange Situation is strongly predictive of later development. Babies assessed as secure typically go on to have better outcomes in may areas, ranging from success at school to romantic relationships and friendships in adulthood. Insecure-resistant attachment is associated with the worst outcomes including bullying in later childhood (Kokkinos 2007) and adult mental health problems (Ward et al 2006).

This is evidence for the validity of the concept because it can explain sunsequent outcomes. 

Good reliability - The strange situation shows good inter-rater reliability. This may be because the strange situation takes place under controlled conditions and because the behavioural categories are easy to observe. In a recent study, Bick et al (2012) looked at inter-rater reliability in a team of trained strange situation observers and found agreement on attachment type for 94& of tested babies. 

This means we can be confident that the attachment type of an infant identified in the strange situation does not just depend on who is observing them.

41 of 66

Strange situation - evaluation

The test may be culture-bound - There is some doubt whether the strange situation is a culture-bount test. This is for two reasons. First, cultural differences in childhood experiences are likely to mean that children respond differently to the strange situation. Second, caregivers from different cultures behave differently in the strange situation. 

For example Takahashi (1990) has noted that the test does not really work in Japan becase Japanese mothers are so rarely separated from their babies that, as we would expect, there are very high levels of separation anxiety. Also in the reunion stage Japanese mothers rushed to the baby and scooped them up, meaning the child's response was hard to believe. 

What does the strange situation measure?

The strange situation measures a child's responses to the anxiety produced by being in an unfamiliar environment. This is not in doubt. However, what is ore controversial is whether the main influence on the anxiety is attachment, as Ainsworth assumed. Kagan (1982) has suggested that in fact temperment, the genetically influenced personality of the child, is a more important influece on behaviour in the strange situation than attachment. It means that temperment may be a confounding variable. 

42 of 66

Strange situation - evaluation

There is at least one more attachment type

Ainsworth conceived of three attachment types: insecure-avoidant, secure and insecure-resistant. However Main and Soloman (1886) pointed out that a minority of chidren display atypical attachments that do not fall within types A, B or C behaviour. 

This atypical attachment is commonly known as disorganised attachment. Disorganised children display an odd mix of resistant and avoidant behaviours. 

43 of 66

Cultural variation in attachment

Cultural variations - 'Culture' refers to the norms and values that exist within any group of people. Cultural variations then are the differences in norms and values that exist between people in different groups. In attachment research we are concerned with the differences in the proportion of children of different attachment types. 

Van Ijzendoorn - conducted a study to look at the proportions of secure, insecure-avoidant and insecure-resistant attachments across a range of countries. They also looked ar the differences within the same countries to get an idea of variations within a culture. 

Procedure: Meta analysis of the strange situation used in 32 studies across 8 countries, 15 in the USA. Overall the 32 studies had the results of 1990 children.

Findings: Wide variation between the proportions of attachment types in different studies. In all countries secure attachment was the most common classification. The proportion varied from 75% in Britain to 50% in China. Insecure-resistant was overall the least common type although the proportions ranged from 3% in Britain to around 30% in Israel. Insecure-avoidant attachments were observed most commonly in Germany and least commonly in Japan. 

An interesting finding is that variations between results of studies within the same country were 150% greater than those between countries. 

44 of 66

Other studies of cultural variation

An Italian Study - Simonella et al (2014) conducted a study in Italy to see whether the proportions of babies of different attachment types still matches those found in previous studies. The researchers assessed 76 12-month olds using the strange situation. They found 50% were secure, with 36% insecure-avoidant. This is a lower rate of secure attachment than has been found in many studies. The researchers suggest this is because increasing numbers of mothers of very young children work long hours and use professional childcare.  These findings suggest that cultural changes can make a dramatic difference to patterns of secure and insecure attachment. 

A Korean study - Jin et al (2012) conducted a study to compare the proportions of attachment types in Korea to other studies. The strange situation was used to assess 87 children. The overall proportions of insecure and secure babies were similar to those in most countries, with most infants being secure. However, more of those classified as insecurely attached were resistant and only one child was avoidant. This distribution was similar to the distribution of attachment types found in Japan. Since Japan and Korea quite similar child-rearing styles this similarly might be explained in terms of child-rearing style.

Conclusions - Secure attachment seems to be the norm in a wide range of culture, supporting Bowlby's idea that attachmet is innate and universal and this type is the universal norm. 

45 of 66

Cultural variation in attachment - evaluation

Large samples - A strength of combining the results of attachment studies carried out in different countries is that you can end up with a very large sample size. For example, in Van Ijzendoorn meta-analysis there was a total of nearly 2000 babies and their primary attachment figures. Even studies like those of Simonells and Jin et al had large comparison sizes from previous research, although their own samples were smaller.

The overall sample size is a strength because large samples increase internal validity by reducing the impact of anomalous results caused by mad methodology or very unusual participants.

Samples tend to be unrepresentative of culture - The meta-analysis by Van Ijendoorn and Kroonenberg claimed to study cultural variation, whereas, in fact, the comparisons were between countries not cultures. Within any country there are many different cultures each with different child-rearing practices. One sample might, over-represent people living in poverty, the stress of which might affect caregiving and hence patterns of attachment. An analysis by Van Ijzendoorn and Sagi (2001) found that distributions of attachment type in Tokyo (an urban setting) were similar to the Western studies, whereas a more rural sample had an over-representation of insecure-resistant individuals. 

46 of 66

Cultural variation in attachment - evaluation

Method of assessment is biased - Cross-cultural psychology includes the ideas of etic and emic. Etic means cultural universals whilst emic means cultural uniqueness.

The strange situation was designed by an American researcher based on a British theory (Bowlby). There is a question over whether Anglo-American thoeries and assessmenrs can be applied to other cultures. Trying to apply a theory or technique designed for one culture to another culture is known as imposed etic. 

An example of imposed etis may be the ides that a lack of separation anxiety and lack of pleasure on reunion indicate an insecure attachment in the strange situation. In Germany this behaviour might be seen more as independence than avoidance and hence not a sign of insecurity within that cultural context (Grossmann & Grossmann 1990).

Alternative explanation for cultural similarity - Bowlby's explanation for cultural similarities is that they are due to the fact that attachment is innate and universal and this produces the same kind of behaviours all ovet the world. Van Ijzendoorn and Kroonenberg proposed an alternative possibility. They suggest that small cross-cultural differences may reflect the effects of the mass media, in which a large number of books and TV programmes 'that advocate similar notions of parenting are disseminated across countries'.

47 of 66

Bowlby's theory of maternal deprivation

Bowlby proposed a theory of attachment but prior to this he developed the thoery of maternal deprivation (1951). 

Maternal deprivation - The emotional and intellectual consequences of separation between a child and his/ her mother or mother substitute. Bowlby proposed that continuous care from a mother is essential for normal psychological development, and that prolonged separation from this adult causes serious damage to emotional and intellectual development. 

Separation v deprivation 

These is an important distinction to be made between separation and deprivation. Separation simply means the child not being in the presence of the primary attachment figure. This only becomes an issue for development if the child is deprived, i.e. they lose an element of her care. Brief separations, particularly where the child is with a substitute caregiver, are not significant for development but extended separations can lead to deprivation, which by definition causes harm.

48 of 66

Bowlby's theory of maternal deprivation

The critical period 

Bowlby saw the first 30 months of life as a critical period for psychological development. If a child is separated from their mother in the absence of suitable substitute care and so deprived of her wmotional care for an extended period during this critical period then psychological damage was inevitable. 

Effects on development 

Intellectual development - Bowlby believed that if children were deprived of maternal care for too long during the critical period they would suffer delayed intellectual development, characterised by abnormally low IQ. Goldfarb (1947) found lower IQ in children who had remained in institutions opposed to those who were fostered and thus had a higher standarf of emotional care.

Emotional development 

Bowlby identified affectionless psychopathy as the inability to experience guilt or strong emotion for others. This prevents the person developing normal relationships and is associated with criminality. Affectionless psychopaths cannot appreciare the feelings of their victims.

49 of 66

Bowlby's 44 thieves study

Examined the link between affectionless psychopathy and maternal deprivation.

Procedure: The study consisted of 44 criminal teenagers accused of stealing. All 'thieves' were interviewed for signs of affectionless psychopathy: characterised by lack of affection, lack of guilt about their actions and lack of empathy for their victims. Their families were also interviewed in order to establish whether the 'theives' had prolonged early separations from their mothers. A control group of non-criminal but emotionally disturbed young peopke was set up to see how often natural separation/deprivation occurred in the chidren who were not theives. 

Findings: Bowlby (1944) found that 14 of the 44 thieves could be described as affectionless psychopaths. Of this 14, 12 had experienced prolonged separation from their mothers in the first two years of their ives. In contrast only 5 of the remaining 30 'thieves' had experienced separations. It was concluded that prolonged early separation/deprivation caused affectionless psychopathy. 

50 of 66

Maternal deprivation - evaluation

The evidence may be poor.  Bowlby drew on a number of sources of evidence for maternal deprivation including studies of children orphaned during WW2, those growing up in poor quality orphanages, and of course his 44 thieves study.

However, these are all flawed as evidence. War-orphans were traumatised and often had poor after-care, therefore these factors might have been the causes of later development difficulties rather than separation. Similarly, children growing up fro birth in poor quality institution were deprived of many aspects of care, not just maternal care. 

Furthermore, the 44 thieved study had some major design flaws, most importantly bias; Bowlby himself carried out the assessments for affectionless psychopathy and the family interviews, knowing what he hoped to find.

Counter evidence. Not all research has supported Bowby's findings. For example, Lewis (1954) partically replicated the 44 thieves study on a larger scale, looking at 500 young people. In her sample a history of arly prolonged separation from the mother did not predict criminality or difficulty forming close relationships. Suggests that other factors may affect the outcome of early maternal deprivation. 

51 of 66

Maternal deprivation - evaluation

The critical period is actually more of a sensitive period. Bowlby used the term 'critical period' because he believed that prolonged separation inevitably caused damage if it took place within that period. However, later research has shown that damage is not inevitable. Some cases of very severe deprivation have had good outcomes provided the child has some social interaction and good aftercare. 

For example, Koluchova (1976) reported the case of twin boys from Czechoslovakia who were isolated from the afe of 18 months until they were 7 years old. Subsequently they were looked after by two loving adults and appeared to fully recover. Cases like this show that the period identified by Bowlby may be a 'sensitive' one but it cannot be critical. 

Animal studies show effects of maternal deprivation. Although most psychologists are very critical of the theory of maternal deprivation, an interesting line of research has provided some support for the idea that maternal deprivation can have long-term effects. Levy et al (2003) showed that separating baby rats from their mother for as little as a day had a permanent effect on their social development though not other aspects of development. 

52 of 66

Maternal deprivation - evaluation

Failure to distinguish between deprivation and privation. Rutter (1981) claimed that, when Bowlby talked of deprivation he was muddling two concepts together. Rutter drew a distinction betwee deprivation, which really means the loss of the primary attachment figure after attachment has developed whereas privation is the failure to form any attachment in the first place. 

Rutter claimed that the severe long-ter damage Bowlby associated with deprivation is actually more likely to be the result of privation. 

53 of 66

Effects of institutionalisation

The theory of maternal deprviation predicted long-term negative effects redult from early deprivation. This can be studied in the context of institutional care. Much of our modern understanding of this has come from studies carried out in the last 25 years in Romania because historical events left a large number of children there in poor quality institutions. 

Institutionalisation - A term for the effects of living in an institutional setting. The term 'institution' refers to a place like a hospital or an orphanage where children live for long, continuous periods of time. In such places there is often very little emotional care provided. In attachment research we are interested in the effects of institutional care on children's attachment and subsequent development. 

Orphan studies - These concern children placed in care because their parents cannot look after them. An orphan is a child whose parents have either died or abandoned them. 

54 of 66

Romanian orphan studies

Rutter's ERA (English and Romanian Adoptee) study

Procedure: Rutter and colleagues (2011) have followed a group of 165 Romanian orphans adopted in Britain to test to what extent good care could make up for poor early experiences in instirutions. Physical, cognitive and emotional development has been assessed at aged 4, 6, 11 and 15 years. 52 British children adopted around the same time have served as a control group. 

Findings: When they first arrived in the UK half the adoptees showed signs of delayed intellectual development and the majority were severely undernourished. At age 11 the adopted children showed differential rates of recovery that were related to their age of adoption. The mean IQ of those adopted before the age of 6 months was 102, compared with 86 for those adopted between 6 months and 2 years and 77 for those adopted after 2 years. Differences remained until 16. 

In terms of attachment, there appeared to be a difference in outcome related to whether adoption took place before or after 6 months. Those children adopted after they were 6 months showed signs of a particular attachment style called disinhibited attachment. Symptoms include attention seeking, clinginess and social behaviour directed indiscriminately towarda all adults, both familiar and unfamiliar. For those adopted before 6 months they rarely showed disinhibited attachment. 

55 of 66

The Bucharest early intervention project

Procedure: Zeanah at al (2005) assessed attachment in 95 children aged 12-31 months who had spent most of their lives in institutional care (90% on average). They were compared to a control group of 50 children who had never lived in an institution. Their attachment type was measured using the strange situation. In addition carers were asked about unusual social behaviour including clingy, attention-seeking behaviour directed inappropriately at all adults.

Findings: They found that 74% of the control group came out as securely attached in the strange situation. However, only 19% of the institutional group were securely attached, with 65% being classified with disorganised attachment. The description of disinhibited attachment applied to 44% of institutionalised children as opposed to less than 20% of the controls. 

56 of 66

Effects of institutionalisation

Disinhibited attachment is a typical effect of spending time in an institution. They are equally friendly and affectionate towards people they know well or who are strangers that they have just met. This is highly unusual behaviour; remember that most children in their second year show stranger anxiety. 

Rutter (2006) has explained disinhibited attachment as an adaption to living with multiple caregivers during the sensitive period for attachment formation. In poor quality institutions like those in Romania a child might have 50 carers none of whom they see enough to form a secure attachment. 

Mental retardation in Rutter's study most children showed signs of retardation when they arrived in Britain. However, most of those adopted before they were 6 months old caught up with the control group by age 4.

It appears that, like emotional development, damage to intellectual development as a result of institutionalisation can be recovered provided adoption takes place before the age of 6 months - the age at which attachments form. 

57 of 66

Effects of institutionalisation - evaluation

Real-life application. Studying the Romanian orphans has enhanced our understanding of the effects of institutionalisation. Such results have led to improvements in the way children are cared for in institutions (Langton 2006). For example, orphanages and children's homes now avoid having large numbers of caregivers for each child and instead ensure that a much smaller number of people, perhaps only one or two people, play a central role for the child. This person is called the key worker. Having a key worker means that children have the chance to develop normal attachments and helps avoid disinhabited attachment. 

Fewer extraneous variables than other orphan studies. There were many studies before the Romanian orphans became available to study but often these studies involve children who had experienced loss or trauma befrore they were institutionalised. For example, they may have experienced neglect, abuse or berevement. These children were often traumatised by their experiences and suffered berevement. It was very hard to observe the effects of institutionalisation in isolation becase the children were dealing with multiple factors which functioned as confoundign participant variables. 

In the case of Romanian orphans is has been possible to study institutionalisation without these confounding variables, which means the findings have increased internal validity. 

58 of 66

Effects of institutionalisation - evaluation

The romanian orphanages were not typical. Although much useful data about institutionalisation has come out of Romanian orphan studies, it's possible that conditions were so bad that results cannot be applied to understanding the impact of better quality institutional care or indeed any situation where children experience deprivation. For example, Romanian orphanages has particularly poor standards of care, especially when it came to forming any relationship with the children, and extremely low levels of intellectual stimulation. 

There is a limitation of the Romanian orphan studies because the unusual situational variables mean the stdies may after all lack generalisability.

Ethical issues. One of the methodological issues for Rutter's ERA is that children were not randomly assigned to conditions. The researchers did not interfere with the adoption process, which means that those children adopted early may have been the or sociable ones, the confounding variable. To control for such variables, another major investigation of fostering versus institutional care, did use random allocation. In the Bucharest Early Intervention project, Romanian orphans were randomly allocated to institutional care or fostering. This is methodologically better because it removes the confounding variable of which children are chosen by parents but it raises ethical issues. 

59 of 66

Attachment and later relationships

Internal working model 

Bowlby (1969) suggested that a chid having their first relationship with their primary attachment figure forms a mental representation of this relationship. This internal working model acts as a template for future relationships. 

The quality of the child's first attachmentis crucial because this template will powerfully affect the nature of their future relationships. A child whose first experience is of a loving relationship with a reliable caregiver will tend to assume that this is how relationships are meant to be. They will then seek out functional relationships and behave functionally within them i.e without wither being too uninvolved or being too emotionally close (which would typify type A behaviour)  or being controlling and argumentative (type C).

A child with bad experiences of their first attachment will being these bad experiences to bear on later relationships. This may mean they struggle to form relationships in the first place or they may not behave appropriately when they have them, displaying type A or C behaviour towards friends and partners. 

60 of 66

Attachment and later relationships

Relationships in later childhood

Attachment type is associated with the quality of peer relationships in childhood. Securely attached infants tend to go on to form the best quality childhood friendships whereas insecurely attached infants later have friendship difficulties (Kerns 1994).

In particular, bullying behaviour van be predicted by attachment type. Wilson and Smith (1998) assessed attachment type and bullying involvement uing standard questionnaires in 196 children aged 7-11 from London. Secure children were very unlikely to be involved in bullying. Insecure-avoidant children were the most likely to be victims and insecure-resistant were most likely to be bullies. 

61 of 66

Attachment and later relationships

Relationships in adulthood with romantic partners

In a study of attachment and both romantic relationshis and friendships McCarthy (1999) studies 40 adult women who had been assessed when they were infants to establish their early attachment type. Those assessed as securely attached infants has the best adult friendships and romantic relationships. Adults classed as insecure-resistant as infants had particular problems maintaining friendships whilst those classed as insecure-avoidant struggled with intimacy and romantic relationships. 

Hazan & Shaver (1987) conducted a classic study of the association between attachment and adult relationships. 

Procedure: Analysed 620 replies to a love quiz printed in a local newspaper. Quiz has 3 sections. First assessed current or most important relationships. Second assessed general love experiences. Third assessed attachment type by asking them to choose which of 3 statements best described their feelings.

Findings: 56% of respondants were securely attached, 25% insecure-A, 19% insecure-R. Secure - most likely to have long and good romantic experiences. Avoidants tended to reveal jealousy and fear of intimacy. Attachment behaviour is reflected in romantic relationships. 

62 of 66

Attachment and later relationships

Relationships in adulthood as parent 

Internal working models also affect the child's ability to parent their own children. People tend to base their parenting style on their internal working model so attachment type tends to be passed on through generations of a family. Recall the study by Bailey (2007). They considered the attachments of 99 mothers to their babies and to their own mothers. Mother-baby attachment was assessed using the strange situation and mother-own mother attachment was assessed using an adult attachment interview. The majority of women had the same attachment classification both to their babies and their own mothers.

63 of 66

Attachment and later relationships - evaluation

Evidence on continuity of attachment type is mixed. Internal workind models predict continuity between the security of an infant's attachmet and that of its later relationships. Evidence for this continuity is mixed. Some studies, like McCarthy do appear to support continuity and so provide evidence to support internal working models. 

Not all studies, however, support internal working models. For example, Zimmerman (2000) assessed infant attachment type and adolescent attachment to parents. There was very little relationship between quality of infant and adolescent attachment.

This is a problem because it is not what we would expect if internal working models were important in development. 

Most studies have issues with validity. Most studies of attachment to the promary caregiver and other significant people do not make use of the strange situation but assess infant-parent attachment by means of interview or questionnaire, not in infancy but years later. Issues with self reports to assess the quality of relationships. 

A related probem concerns the retroactive nature of assessment of infant attachment. Looking back in adulthood at one's early relationship to a primary attachmet figure probably lacks validity because it relies on accurate recollections. 

64 of 66

Attachment and later relationships - evaluation

Association does not mean causality. In those studies where infant attachment type is associated with the quality of later relationships the implication is that infant attachment type causes the attachment. However, there are alternative explanations for the continuity that often exists between infant and later relationships. A third environmental factor such as parenting style might have a direct effect on both attachment and the child's ability to form relationships with others. Alternatively the child's temperament may influence both infant attachment and the quality of later relationships. 

This is a limitation because it is counter to Bowlby's view that the interal working model caused these later outcomes. 

The infuence of early attachment is probabilistic. It does not seem very likely that the quality of infant attachments is an influence on later relationships. However, some attachment researchers, including Bowlby, have probably exaggerated the significance of this influence. Clarke & Clarke (1998) describe the influence of infant attachment on later relationships as probabilistic. People are not doomed to always have bad relationships just because they had attachment problems. They just have a greater risk of problems. There is a further issue that by emphasising this risk we become too pressimistic about people's futures.

65 of 66

Attachment and later relationships - evaluation

Self-report is concious but internal working models are not. There is a theoretical problem with most research related to internal working models. Internal working models are unconcious; we are not directly aware of their influence on us. We woudld not really expect to get direct evidence about them by meanss of interviews or questionnaires because people can only self-report what they are aware of. 

When participants self-report on their relationships they are relying on their concious understandign of those relationships. At best the self-report gives us inirect evidence about internal working models. This is a potential limitation of most research involving the concept of internal working models. 

66 of 66


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Attachment resources »