Attachment

Caregiver-infant interactions in humans: reciprocity and interactional synchrony

• Attachment is a two way enduring emotional bond that develops between the infant and the caregiver early in the infant’s life. It develops as a result of their two-way communication. 

• This process of responding to each other builds emotional bonds and ultimately results in the infant showing distress when separated from the caregiver.

Types of interactions

• Interactional synchrony- Infant and caregiver reflect each other's actions and emotions in a coordinated manner.

• A behaviour linked to this is reciprocity. Reciprocity- is interaction wherein the adult and infant continuously respond to each others actions and can initiate or respond to communication.

• Interaction can also occur via direct imitation where the infant mimics/copies the adult’s behaviour exactly (i.e: smiles elicit smiles).

• Sensitive responsiveness is when the adult pays close attention to the infant’s communication and responds in an appropriate manner (i.e: providing milk, changing).

• Caregiverese is where adults modulate their voice, slowing it down and raising the pitch to make it almost song-like. Most adults will do this when encountering an infant.

• Bodily contact (physical contact, often skin-to-skin) is seen as important in bonding, particularly in the first few hours of life.

Evaluation

• Melzoff and Moore (1977) had an experimenter display facial gestures such as sticking their tongue out, opening mouth in shock, and manual gestures such as opening and closing their hand to infants between 12 and 21 days old. Infants responses were recorded and rated by people blind to the aims of the experiment. The ratings showed that these infants imitated the experimenter. This suggests that the ability to observe and imitate develops very early on in infants, potentially as a way to develop an attachment bond with their caregiver.

• Papusek et al (1991) Showed that the tendency to produce caregiverese (high pitched baby talk) is common across American, Chinese, and German mothers. This suggests that some aspects of caregiver-infant interactions in the development of attachment are not culturally biased but are universal to all or many cultures and thus may be innate within infants and caregivers.

• Modern techniques of studying attachment tend to use multiple observers, providing inter-rater reliability and reducing potential for bias. 

• They also often have a system of video cameras to document and slow down micro-sequences of interactions between caregivers and infants that may not be observable in real time. This also allows other researchers to more easily review the evidence at a later date.

• An issue with studies on infant-caregiver interaction is that infants are unable to communicate their thoughts and emotions, so findings depend on inferences about internal mental states based on observations of infant behaviour. This is unscientific as it is subjective and so some studies could suffer from observer bias, that is an interpretation that matches the observer’s preconceptions rather than what they were actually viewing. 

• Also researcher are unable to claim intentionality (that imitation behaviour is deliberate) as it may be an unconscious automatic response.

• Social sensitivity is a concern when investigating child rearing techniques as some parents may find their life choices criticised, such as mother who return to the workplace shortly after giving birth. Findings may lead new parents to blame themselves if their attachment is not strong or their children do not develop according to the models (see below).

• Schaffer and Emerson (1964): The Glaswegian baby study- collected the data of 60 babies and their families over the course of a year via monthly observations and interviews, with a follow-up visit at 18 months. Two types of behaviour were looked at, stranger distress (signs of discomfort when around a stranger and ability to distinguish between familiar and unfamiliar people) and separation anxiety (signs of discomfort when the caregiver moved to another room, showing the development of an attachment bond). They found separation anxiety in the majority of babies by 25-32 weeks, with stranger anxiety occurring in most babies approximately one month later. In the 18 month follow-up 87% of babies had developed multiple attachments. The strongest attachments tended to be to the mother, particularly those with consistent caregiver-infant interaction.

• This suggested to the researchers that attachment occurs in stages in all babies, perhaps being a biologically controlled process. Also the quality of caregiver interaction has a direct impact on the strength of the attachment for infants and their caregivers.

Evaluation

• This study used only white, working class Scottish babies from Glasgow, so cannot be generalised to babies of other classes or cultures (or even babies living in other cities in Scotland) so lacks external validity.

• Study was carried out in the 1960s so now may lack temporal validity as childrearing practices have changed significantly over the last 50 years.

• Has high ecological validity as children were studied in their own homes, and high mundane realism as strangers visiting the home would have been a fairly normal occurrence for the babies, as would have been the caregiver leaving the presence of the infant

• By incorporating self-report into the study researchers were able to use the process of triangulation (using two different research methodologies -interview and self-report- and finding the same results), so greater internal validity.

Stages of attachment identified by Schaffer

• Schaffer and Emerson identified four stages of attachment through their longitudinal observation study. The named these Asocial, Indiscriminate, Specific, and Multiple.

1. The Asocial (pre-attachment) stage occurs from 0-6 weeks of age. Babies in this stage respond to objects in a similar way to humans, such as by smiling, however babies tend to be more content with certain individuals and humans in general compared to objects.

2. The Indiscriminate (diffuse attachment) stage occurs from 6 weeks to 7 months. In this stage babies can be handled by strangers without distress, however, can discriminate between familiar and unfamiliar individuals, with preference for familiar adults. No separation or stranger anxiety.

3. The Specific (or discriminate attachment) stage occurs from 7-9 months onwards. Babies in this stage experience separation and stranger anxiety, and they demonstrate a preference for one primary caregiver such as their mother.

4. The Multiple stage occurs after 9 months of age. Babies in this stage can be observed to be attached to more than one individual, including brothers, sisters, and grandparents. Fear of strangers also decreases.

Multiple attachments and the role of the father

• Schaffer found that the primary attachment figure was the mother 65% of the time. In 30% of cases the mother and someone else (i.e: the father) were both primary attachment figures, but only in 3% of cases was the father the primary attachment figure. This may be due to cultural and temporal reasons (i.e: is it part of living in Glasgow or in the 1960s?) or potentially biological reasons.

• The role of mothers and fathers have changed in Western cultures since the 1960s, with mother more likely to work and father more likely to help with a greater proportion of the childrearing that previously was done by the mother. This is likely to have an impact on attachment patterns in infants.

• Bowlby (1988) reasoned that if patterns of attachment are a product of how their mother has treated them, it could be anticipated that the pattern they develop with their father is the product of how their father has treated them. Bowlby suggests that fathers can fill a role closely resembling that filled by a mother but points out that in most cultures this is uncommon. Bowlby argues that in most families with young children, the father's role tends to be different. According to Bowlby, a father is more likely to engage in physically active and novel play than the mother and tends to become his child's preferred play companion.

• Field (1978) conducted research which compared the behaviours of primary caretaker mothers with primary and secondary caretaker fathers. Face-to-face interactions were analysed from video footage with infants at 4 months of age. Overall, it was observed that fathers engaged more in game playing and held their infants less. However, primary caretaker fathers engaged in significantly more smiling, imitative grimaces, and imitative vocalizations than did secondary caretaker fathers and these were comparable with mothers’ behaviour. This suggests that men who take on the role of primary caregiver change their interactional style to be more like that of a mother in terms of comfort and sensitive responsiveness.

• This point can be paralleled with Ainsworth’s Caregiver Sensitivity Hypothesis (1979). Although this was based on observations of mothers’ behaviour, it predicted that responsive and sensitive care predicted attachment types.

• Verissimo et al (2011) observed preschool children’s relationships with their mothers and fathers and these were then compared to assessments of later social interactions at nursery. A strong attachment to the father was the biggest predictor of ability to make friends at nursery, suggesting that fathers play an important role in the socialisation process.

Evaluation

• Research on the importance of the father in social development finds that males can effectively take on a more maternal role could provide confidence to fathers taking on the role of primary caregiver and for single gender families (i.e: those with two dads) becoming more common in modern society.

• Infants are unable to communicate their thoughts and emotions, so findings are based on inferences about the internal mental states of the infant based on observations of the infant’s behaviour. This is unscientific as it is subjective and so may suffer from observer bias, an interpretation matching the observer’s preconceptions. This may have occurred in Schaffer and Emerson’s study as they only used one observer during home visits and recordings were not possible at the time, so we can’t review their observations.

Animal studies of attachment: Lorenz and Harlow

• Imprinting is when an animal such as a bird will strongly attach to the first object (usually the mother) that they encounter. The infant animal will then follow this object.

• Lorenz (1932)- Lorenz aimed to investigate the mechanisms of imprinting. Lorenz took a large clutch of goose eggs and kept them until they were about to hatch out. Half of the eggs were then placed under a goose mother, while Lorenz kept the other half in an incubator beside himself for several hours.

• When the geese hatched the young birds regarded Lorenz as their mother and followed him accordingly. The other group followed the mother goose.

• Lorenz found that geese follow the first moving object they see, during a 12-17 hour critical period after hatching. This process is known as imprinting, and suggests that attachment is innate and programmed genetically.

• Imprinting has consequences, both for short term survival, and in the longer term forming internal templates for later relationships. Imprinting occurs without any feeding taking place. If no attachment has developed within 32 hours it’s unlikely any attachment will ever develop.

• To ensure imprinting had occurred Lorenz put all the goslings together under an upturned box and allowed them to mix.  When the box was removed the two groups separated to go to their respective 'mothers' - half to the goose, and half to Lorenz.

• Imprinting does not appear to be active immediately after hatching, although there seems to be a critical period during which imprinting can occur.  Hess (1958) showed that although the imprinting process could occur as early as one hour after hatching, the strongest responses occurred between 12 and 17 hours after hatching, and that after 32 hours the response was unlikely to occur at all.  Lorenz and Hess believe that once imprinting has occurred it cannot be reversed, nor can a gosling imprint on anything else.

• Lorenz concluded that the process of imprinting is a strong biological feature of attachment in certain birds, and that imprinting occurs based on the first object rather than on other cues such as scent or sound. This would explain why goslings imprint after a matter of minutes due to their increased mobility; human babies are born immobile and therefore there is less call for them to form an attachment straight away, and so, this develops later (7-9 months). It supports the view that having a biological basis for an attachment is adaptive as it promotes survival. The longevity of the goslings’ bond with Lorenz would support the view that, on some level, early attachment experiences do predict future bonds.

• Harlow (1958)- Harlow aimed to study the mechanisms by which newborn rhesus monkeys bond with their mothers. 16 monkeys were separated from their mothers immediately after birth and placed in cages with access to two surrogate mothers, one made of wire and one covered in soft terry towelling cloth.

• Eight of the monkeys could get milk from the wire mother

• Eight monkeys could get milk from the cloth mother

• The animals were studied for various length of time.

• Both groups of monkeys spent more time with the cloth mother (even if she had no milk). The infants of the second group would only go to the wire mother when hungry. Once fed they would return to the cloth mother for most of the day. If a frightening object was placed in the cage the infant took refuge with the cloth mother. The infant would explore more when the cloth mother was present.

• Then Harlow observed the difference in behavior between the monkeys who had grown up with surrogate mothers and those with normal mothers. He found that (for the surrogate mother monkeys):

1. They were much more timid (though monkeys with access to a cloth mother were less timid than those without, who were very distressed).

2. They didn’t know how to act with other monkeys.

3. They were easily bullied and wouldn’t stand up for themselves.

4. They had difficulty with mating.

5. The females were inadequate mothers.

• These behaviors were observed only in the monkeys who were left with the surrogate mothers for more than 90 days. For those left less than 90 days the effects could be reversed if placed in a normal environment where they could form attachments.

• Harlow concluded that “contact comfort” (provided by the cloth mother) was more important than food in the formation of attachment. This also shows that contact comfort is preferable to food but not sufficient for healthy development.

• He also concluded that early maternal deprivation leads to emotional damage but that its impact could be reversed in monkeys if an attachment was made before the end of the critical period. However if maternal deprivation lasted after the end of the critical period then no amount of exposure to mothers or peers could alter the emotional damage that had already occurred.

• Harlow found therefore that it was social deprivation rather than maternal deprivation that the young monkeys were suffering from.  When he brought some other infant monkeys up on their own, but with 20 minutes a day in a playroom with three other monkeys, he found they grew up to be quite normal emotionally and socially.

• This study suggests that Rhesus macaques and potentially other primates such as humans have a biological (nature) need for physical contact and will attach to whatever provides comfort rather than food, going against the behaviourist theory of attachment (see below).

Evaluation of animal studies

• There are serious ethical concerns about the level of suffering that primates endures in Harlow’s experiments as he intentionally orphaned infants and subjected them to high levels of stress. Some of Harlow’s other experiments were more extreme and lead in part to a negative view of psychology as a field of research, though this also lead to changed ethical standards. 

• Similarly, as Lorenz’s experiment had a permanent and irreversible negative effect on the geese (for example those that imprinted to him were less able to mate with their own kind) so was very unethical.

• There are problems with generalising findings on attachment from animal studies to humans infants, as geese are very different in evolutionary terms, and whilst monkeys are similar genetically to humans, there are still significant differences in both biology and cultural/social environments. 

• Knowledge gained form Harlow’s studies has been applied effectively to the early childcare of human infants. For example contact between mothers and babies is encouraged in the first few hours after birth to promote attachment, and social workers investigate cases of infant neglect as they understand the long term damage that can result from this. It has also been argued that the long-term benefit to millions of human infants resulting from Harlow’s research justifies the studies in terms of cost-benefit analysis.

Explanations of attachment: Learning theory and Bowlby’s Monotropic theory

• Dollard and Miller (1950) proposed the Cupboard love theory which suggested that the reason children become attached to their caregiver is because they learn that the caregivers provide food and meet their other physiological needs.

• This learning theory suggests that all behaviour including attachment behaviour can be explained by classical and operant conditioning.

• Classical conditioning is learning by association. When two stimuli are presented multiple times, such as food (unconditioned stimulus) and the mother (neutral stimulus), the feeling of pleasure (unconditioned response) starts to become associated with the mother, so the mother becomes the conditioned stimulus and pleasure becomes the conditioned response. So pleasure now happens whenever the mother appears.

• Operant conditioning is learnt through the consequences of trial and error, so through patterns of reinforcement. Pleasurable consequences for crying such as receiving food act as positive reinforcement, making crying behaviour when hungry more likely to happen. Also stopping crying (removing the negative stimulus) by providing food works as negative reinforcement for parents, making it more likely that parents will provide food the next time the baby cries.

• Primary drives (desire to complete an action such as feeding and sleeping) are instinctive according to behavioural psychologists as they are based on a biological need. Secondary drives are desires to complete actions that develop due to a learned process (i.e; we desire money), in which they are associated with a satisfying primary drive such as food. This would class attachment as a secondary drive as it is learnt by the baby through access to food.

Evaluation

• Learning theory is a clear and believable explanation for attachment , and the underlying theories on which it is based are backed up by significant amounts of well-controlled research. Also the theory has face validity. It ‘makes sense’ that babies would cry more if they learnt that it would gain them attention.

• Learning theories applied to human feelings of attachment can be seen as environmentally reductionist in explaining the complex interactions between primary caregivers and their infants as a result of simplistic processes such as stimulus-response links and patterns of reinforcement.

• The operant conditioning explanation for caregiver attachment to the baby is questionable in evolutionary terms as if the baby is repeatedly producing a negative effect (crying) then it would logically make more sense to remove the negative stimulus in a more permanent way such as by abandoning the baby somewhere. The removal of this negative stimulus would negatively reinforce the caregiver to abandon any other baby that they were supposed to care for, which goes against the survival of a species (if all babies are abandoned then very few will survive into adulthood). Therefore, given the continual survival of the human species, there must some other explanation for the caregiver’s attachment to the baby.

• There is a lot of evidence that rejects the cupboard love theory. Harlow’s research on monkeys showed that infant monkeys did not attach to the surrogate wire monkey that provided milk, but attached to the cloth mother that did not provide milk but instead provided some level of contact comfort that whilst not necessary for its survival, did seem to fulfill some other innate need.

• There are alternative theories of attachment that don’t depend on learning theory, such as Bowlby’s monotropic theory (see below). This gives a reasonable evolutionary explanation as to why infants would have an instinctual drive to form an attachment with the primary caregiver.

Bowlby’s monotropic theory

• Bowlby’s monotropic theory is an evolutionary explanation of attachment. It states that babies have an innate attachment drive in order to survive, as security results in survival. This is shown by the fact that babies will stay close to the caregiver (usually the mother) for safety. Bowlby’s theory is based on Lorenz’s theory of imprinting and Harlow’s contact comfort theories.

• Monotropy is the unique strong attachment that infants have to a single caregiver (usually the mother).

The concepts of a critical period and an internal working model

• Bowlby proposed that there is a critical period for attachment. Attachment must happen in the first 2 ½  years after birth, or else they will never form an attachment and this will result in long-lasting negative social consequences. This idea is based on Lorenz’s work.

• Bowlby also proposes that the first strong attachment a child has (its attachment with its primary caregiver) acts as a blueprint for all of the child’s future relationships. This internal working model acts as a guide on how to conduct future relationships such as if other people can be trusted, or if relationships are loving. This schema is based on the work of both Freud and Harlow.

• Bowlby suggests that stronger attachments will form if the care is consistent and weaker attachments form from long and/or frequent separations.

• As well as this, Bowlby states that infants with a good attachment to their primary caregiver will use the caregiver as a safe base from which to explore their environment, but will show distress if their carer leaves them or if a stranger approaches.

• Finally, Bowlby suggests that babies instinctively use signals that attract a caregiver’s attention (such as crying, smiling, vocalisations), and caregivers are biologically programmed to find certain behaviours (such as smiling) cute and others (such as crying) distressing. These behaviours are called social releasers and are used to activate caregiver attachment.

• To remember Bowlby’s theories use SCAMI: Social releasers, Critical period Attachment-consistency, Monotrophy, and Internal working model.

Evaluation

• Bowlby’s work is based in part on Lorenz’s work. Imprinting studies on geese indicate the potential strength of attachment to a single attachment figure, explained by evolutionarily significant survival advantages. However these are findings of animal studies applied to humans, and geese and humans are very different in evolutionary terms, so Lorenz’s findings may lack the external validity with which to be applied to human infants.

• Bowlby’s theory is criticised as suffering from alpha bias as it overexaggerates the differences between men and women by suggesting that it is/should only be the mother who acts as the primary caregiver and that the father’s role is simply to provide for the family. This lacks temporal validity as in modern society men are expected to do a much more equal amount of the childcare compared to women. Bowlby’s theory also has negative implications for working mothers or those bringing up a child who is are not the child’s mother (i.e: grandparents, gay male couples ..etc).

• There are alternative explanations of attachment. Behaviourists claim that it is the environment, not biology, that creates attachment. The cupboard love learning theory suggests that attachment is based on patterns of reinforcement and the association of the mother with the food that she provides. Learning theory is well supported by numerous experiments.

• The continuity hypothesis (see below) suggests that the types of relationships that we have as adults can be predicted by the attachments we have as infants due to the development of the internal working model. This is highly deterministic as people like to believe that they have full conscious control over their own relationships, suggesting responsibility for the success or failure of a relationship is set in infancy has socially sensitive implications such as potentially justifying domestic abuse (ie: because the man’s mother was violent to him this caused him to behave violently in all other relationships) which would have implications for the criminal justice system.

• Ainsworth (1970) was a colleague of Bowlby who identified a range of behaviours indicating attachment strength from her observation of Ugandan and American mothers. These were proximity to mother, exploration/secure base, stranger anxiety, separation anxiety, reunion response, and sensitive responsiveness of the mother to the infant’s needs.

Ainsworth’s ‘Strange Situation’

• Ainsworth and Bell (1971) aimed to measure different forms of infant attachment. They conducted a controlled observation recording the reactions of a child and mother (caregiver), who were introduced to a strange room with toys. In the strange situation about 100 middle-class American infants and their mothers took part. The infant’s behavior was observed during a set of predetermined activities.

• The Strange Situation procedure involved the child experiencing eight ‘episodes’ of approximately 3 minutes each.

• In this procedure the child is effectively observed playing for 20 minutes while caregivers and strangers enter and leave the room, recreating the flow of the familiar and unfamiliar presence in most children's lives. Observers noted the child’s willingness to explore, separation anxiety, stranger anxiety and reunion behavior.

• Ainsworth & Bell observed from the other side of a one-way mirror so that the children did not know that they were being observed.

Stage 1 – Mother and child enter the playroom

Stage 2 – The child is encouraged to explore

Stage 3 – Stranger enters and talks briefly to the mum, then attempts to interact with the infant

Stage 4 – Mothers leaves while the stranger is present

Stage 5 – Mother enters and the stranger leaves

Stage 6 – Mothers leaves

Stage 7 – Stranger returns and interacts with infant

Stage 8 – Mother returns and interacts with child, stranger leaves

• The observers assessed the infant’s willingness to use its mum as a safe base and explore the environment in stages 2 and 5, assessed separation anxiety in stages 4 and 6, assessed reunion response in stages 5 and 8, and assesses stranger anxiety in stages 3,4, and 7.

Types of attachment: secure, insecure-avoidant and insecure-resistant

• The findings provided evidence for three distinct attachment types that ainsworth called Insecure-Avoidant, Insecure-Resistant, and Secure. These types also seemed to correlate with the level of sensitive responsiveness of the parents, suggesting that secure attachment develops due to the attention of a consistently positive and responsive mother.

• 15% of infants were classified as Insecure-avoidant (Type A). These infants consistently kept a distance away from their mothers, and did not use her as a secure base, though still explored freely. They had low stranger anxiety and little separation anxiety. They also didn’t attempt to seek comfort from their mothers when they returned. The mothers of these babies appeared to show little sensitive responsiveness.

• 70% of infants were classified as Secure (Type B). This is the ideal attachment type. Secure infants used their mothers as a safe base as they explored the environment. These infants had high stranger anxiety and separation anxiety, but had a happy reunion response that allowed them to settle quickly back into exploration. Mothers of these infants showed sensitive responsiveness.

• 15% of infants were classed as Insecure-resistant (Type C). These infants would not explore the environment and were inconsistent about wanting closeness with or distance from their mum. They had high stranger anxiety and separation anxiety, but were unable to settle when reunited with their mothers, rejecting her attention. The mothers of these infants seemed inconsistent with their sensitive responsiveness.

• Conclusions: Most US children appeared to be securely attached. The results highlight the role of the mother’s behaviour in determining the quality of attachment.

• This led to the conceptualisation of the Caregiver Sensitivity Hypothesis, which suggests that a mother’s behaviour towards their infant predicts their attachment type.

Evaluation of Strange Situation

• Replicable/high inter-observer reliability- As the research is highly operationalised, observers have a clear view of how a securely attached infant should behave, due to the 4 specific criteria that Ainsworth used. For this reason, the research had high inter-rater reliability.

• It is also reliable as it was a highly controlled observational study with a clear standardised procedure and behavioural categories, so could easily be replicated with the same results obtained.

• The Strange Situation procedure has been replicated across many different cultural groups, enhancing the generalisability of its findings.

• The Strange Situation is a well respected and is the standard diagnostic tool used to measure the strength and type of relationships between mothers and their infants.

• Reliability of classifications- Waters (1978) assessed 50 infants at 12 and at 18 months of age using the SS procedure. Waters found clear evidence for stable individual differences using Ainsworth’s behavior category data. The greatest consistency was seen in reunion behaviours after brief separations. 48 of the 50 infants observed were independently rated as being classified in the same category at 18 months.

• A criticism of the study is that it has low ecological validity, and the results may not be applicable outside of the lab. The environment of the study was controlled and the eight scripted stages of the procedure (e.g. mum and stranger entering and leaving the room at set times) would be unlikely to happen in real life.

• Categories are not always applicable- A further classification group (disorganised) was subsequently identified by Main & Cassidy (1988), which would suggest that infants do not all fit into the three categories introduced by Ainsworth.

• Procedure is culturally biased- The SS was designed by an American according to observations of US children. Consequently, the criteria used to classify infants are based on US values, relating to child-parent behaviour. It could be argued that this is Ethnocentric, so observations of non-Americans will judged according to American standards. E.g. Japanese infants were judged as being resistant due to high levels of distress that were observed but this reflects their lack of experience being alone during the “infant alone” part of the research as Japanese infants spend most of their time being carried by their mothers. German children were more likely to be classified as avoidant because Germans value independence in their children so discourage secure base seeking behaviour.

• Observations were only a snapshot of behaviour, not taking into consideration wider factors such as the child’s relationship with other family members or their behaviour in familiar environments.

• The term ‘insecure’ is a value judgement in that these relationships are considered to be worse than secure ones are. This ethnocentric method is used elsewhere can effectively label entire cultures as defective but also stigmatises mothers who may struggle to balance the demands of work and childcare making the research socially sensitive.

Cultural variations in attachment, including van Ijzendoorn

• Individualist cultures value independence with each working to their own individual goals e.g. USA and Europe (Western Cultures).

• Collectivist cultures value cooperation with each working towards the family or group goals e.g. Japan and Israel (Eastern Cultures)

• Van Ijzendoorn & Kroonenberg (1988) aimed to investigate if attachment styles (secure and insecure) are universal (the same) across cultures, or culturally specific (vary considerably from place to place, due to traditions, the social environment, or beliefs about children).

• They analysed data from other studies using a method called meta analysis. Data from 32 studies in 8 different countries was analyzed, involving a total of around 2000 infant participants.

• All the 32 studies used the strange situation procedure to study attachment. Using a meta analysis (statistical technique) they calculated the average percentage for the different attachment styles (e.g. secure, avoidant, resistant) in each country.

• Average findings were relatively consistent with Ainsworth’s original research - Secure 65% - Avoidant 21% - Resistant 14%. Intra-cultural variation was nearly 1.5 times greater than the cross-cultural variations. Van Ijzendoorn speculated that this was linked to differences in socio-economic factors and levels of stress that varied between samples used within each country. 6/8 countries produced findings that were proportionally consistent with Ainsworth & Bell. Non-western collectivist countries like Japan (27%) & Israel (29%) revealed a higher incidence of resistant than avoidant children, whereas individualistic western countries like Germany (35%) had a higher incidence of avoidant than resistant. Chinese findings revealed the lowest rate of secure attachments (50%) with the remaining children falling into the other categories equally, and Britain had the highest rate of secure attachments (75%). Secure attachments were the most common (50% or more in all cultures) and insecure-resistant were least common. It was concluded that the modest cross-cultural differences reflect the effects of mass media, which portrays similar notions of parenting.

• The general trend towards secure attachment suggests that it is the globally preferred attachment style, suggesting it may be the innate/natural way for a caregiver to interact with their baby.

Evaluation

• Comparison is aided by the standardised methodology. The use of the strange situation as a procedure means that a comparison can be made across cultures, and the reliability is therefore high

• Meta-analyses include very large samples, so a strength of this study is that any poorly-conducted study or unusual results only have a small effect on the overall results, increasing the validity of the findings.

• As the dominant attachment style was secure, this may be evidence for Bowlby’s theory that there is a biologically instinctive drive to parent in a way that produces secure attachment.

• The study was not globally representative -Van Ijzendoorn & Kroonenberg recognised that data from less Western-oriented cultures were required to establish a more global perspective attachment classifications, pointing out that Africa, South America, and Eastern European socialist countries were not represented.

• Overall findings are misleading - As a disproportionately high number of the studies reviewed were conducted in the USA (18/32), the overall findings would have been distorted by these. This means that the apparent consistency between cultures might not genuinely reflect how much attachment types vary between cultures. Results for Germany misleading as German families value independent behaviour, so secure children may have been socialised out of using mother as a secure base. Misrepresentative for Japan as Japanese infants are almost never separated from their mothers in the first two years of life, so are very unused to being left alone. Israeli results misleading as the sample used for that country were being raised in a Kibbutz, a communal home where they would rarely encounter strangers (raised by a nurse) but would often be away from their mothers.

• Applying Strange Situation procedures and behavioural categories is ethnocentric – Cross-cultural research using the Strange Situation judges and categorises infant behaviour according to behavioural categories that were developed following observations of middle-class American infants. This means that when researchers interpret non-American infant behaviour, it is being judged against an American standard. Eg. an infant exploring the playroom by themselves would be classed as avoidant based on American standards but is valued as reflecting independence in Germany

• Takahashi (1990) replicated the Strange Situation with 60 middle class Japanese infants & mothers using the same standardised procedure and behavioural categories. Takahashi’s observation revealed distinct cultural differences in how the infants responded to the 8 stages of the procedure. The findings were 0% insecure-avoidant. Infants became severely distressed in the “infant alone step”; this situation was quite unnatural and broke cultural norms for the infants, 32% insecure-resistant, 68% secure. 90% of infant-alone steps had to be stopped due to excessive infant anxiety.

• Simonelli (2014) conducted the Strange Situation with 76 middle class Italian infants and found that there was a much lower rate of secure attachment than in historical findings (50%) and a much higher rate (36%) of infants with an insecure-avoidant type.

• Researchers suggested that this change is due to the changing demands of modern life, with more frequent separations from the primary caregiver and more complex childcare arrangements due to work commitments. They suggested that rather than being a problem, the relationship was still healthy, just shifted to the insecure-avoidant as a coping mechanism by the infants in order to not be constantly showing extreme emotion when separated. Perhaps in modern society this is the new secure style?

Bowlby’s theory of maternal deprivation

• Bowlby proposed the idea of Monotropy, that is, that infants form a single unique attachment bond with their mother, and that this is needed for healthy psychological development in infants.

• Maternal deprivation is not receiving suitable emotional care from a maternal figure.

• If attachment is disrupted/not formed within 2 ½ years since birth then the baby will never form an attachment and the lack of stimulation will have permanent negative intellectual and emotional consequences due to deprivation which will be permanent.

• ADDIDDAS to remember the effects of maternal deprivation: Aggression, Delinquency, Dwarfism, Intellectual retardation, Depression, Dependency, Affectionless Psychopathy, Social maladjustment.

• Three of consequences of deprivation that Bowlby outlines are delinquency (due to delayed social development these children often show behaviour outside of acceptable norms such as petty crime) , affectionless psychopathy (due to delayed emotional development children are unable to show caring behaviour to others or empathy for people’s feelings, and feel little guilt over harmful actions), and low IQ (due to delayed intellectual development their general cognitive abilities are lower than those of peers).

• The continuity hypothesis states that prolonged separation/deprivation from the mother limits the development of an effective schema (internal working model) for later relationships. If this schema fails to develop properly then it will lead to unsuccessful childhood and adulthood relationships as well as issues with their own parenting skills.

• Bowlby (1944) The 44 juvenile thieves: their characteristics and home life

• Aim: To investigate the long-term effects of maternal deprivation.

• Procedure: He selected an opportunity sample of 88 children attending his clinic.

• Group 1- thief group: 31 boys and 13 girls in the ‘theft group’ were referred to him because of their stealing.

• Group 2- control group: 34 boys and 10 girls were referred to him because of emotional problems.

• The two groups were matched for age and IQ.

• The children and their parents were interviewed and tested by a psychiatrist (Bowlby), a psychologist and a social worker focusing specifically on their early life experiences.

• Findings: 14 children from the theft group were identified as affectionless psychopaths, 12 of those had experienced prolonged separation of more than six months from their mothers in their first two years of life whereas only 5 of the 30 children not classified as affectionless psychopaths had experienced separations. Out of the 44 children in the control group, only 2 had experienced prolonged separations and none of them were affectionless psychopaths.

• This suggests that the affectionless psychopathy may have led to criminal/delinquent behaviour and may be linked to the periods of separation that thieves experienced in early life.

Evaluation

• However this research was only correlational, so a third factor such as extreme poverty, criminal relatives, or a family history of mental issues like psychopathy could be related both to the early separation and to the children’s behavioural difficulties.

• This was an opportunity sample, so these delinquents weren’t necessarily a representative sample of children who have experienced long separations. They were likely brought to Bowlby’s clinic because they were extreme cases, meaning that the external validity of this study is questionable.

• Bowlby’s theories come from assuming findings from animal studies should be applied to humans, but there are significant biological and psychological differences between humans and even our closest evolutionary relative, meaning that such findings can’t be validly generalised to humans.

• Bowlby's Maternal Deprivation theory is supported by Harlow's (1958) research with monkeys. He showed that monkeys reared in isolation from their mother suffered emotional and social problems in older age. The monkey's never formed an attachment (privation) and as such grew up to be aggressive and had problems interacting with other monkeys.

• Due to Bowlby’s theory a number of real life applications have been made: In orphanages now they have to take account of emotional needs, fostered children have to be kept in one stable home rather than being moved around. In maternity units mothers are now allowed to spend more time with their babies as well as if they have a sick child the visiting hours in hospital have been extended, parents can even stay overnight if they wish.

• Critics such as Rutter have also accused Bowlby of not distinguishing between deprivation and privation – the complete lack of an attachment bond, rather than its loss. Rutter stresses that the quality of the attachment bond is the most important factor, rather than just deprivation in the critical period.

• Bowlby assumed that physical separation on its own could lead to deprivation but Rutter argues that it is the disruption of the attachment bond rather than the physical separation. This is supported by Radke-Yarrow (1985) who found that 52% of children whose mothers suffered with depression were insecurely attached. This figure raised to 80% when this occurred in a context of poverty (Lyons-Ruth,1988). This shows the influence of social factors. 

• Bowlby did not take into account the quality of the substitute care. Deprivation can be avoided if there is good emotional care after separation. Hodges and Tizard's research (on privation / institutional care) shows that the effects of deprivation can be reversed to some extent.

• Bowlby’s work focuses on the unique importance of the mother-child bond but Schaffer’s work does not show this. Instead it shows how children quickly move from specific attachments to multiple attachments. Also in the Glaswegian baby study 35% didn’t have the mother as a primary attachment figure even before they reached the multiple attachments stage.

• Studies such as Rutter (2011, see below) highlight the importance of a sensitive period rather than a critical period. That is that while deprivation has long term damaging effects, some of these can be reversed.

Romanian orphan studies: effects of institutionalisation

• Institutionalisation is living away from a family environment, such as in a children’s home or hospital.

• Extended stays in institutions can result in children permanently behaving according to the rule of the institution and losing their personal identity (deindividuation), affectionless psychopathy, delinquency, and low IQ.

• Romanian orphans- The fall of the communist government in 1989 revealed the terrible conditions of up to 300,000 orphans. Contraception had been banned so many families had given up excess children to orphanages. The orphans lacked physical and emotional care from the staff and many were malnourished and abused.

• Whereas deprivation is not receiving suitable emotional care from a primary attachment figure, and can happen with both frequent and/or extended absences, privation is the total lack of any attachment bond (this happened to the Romanian orphans).

• Rutter (2011) aimed to understand the impact that this privation had on the romanian orphans. Studied 165 Romanian orphans adopted into British families. The children were followed in four groups; Group 1 were 58 children under the age of 6 months, Group 2 were 59 children between the ages of 6 and 24 months, Group 3 were 48 children over 24 months, and Group 4 was a control group of 52 british adoptees. At the start of the observation half of the Romanian children were severely malnourished and had low IQ compared to children of the same age. Each group was assessed at ages 4, 6, 11, and 15.

• At age 6 children adopted after 6 months showed disinhibited attachment and were overly friendly with strange adults. At 11 just over half (54%) of the Romanian adopted children that showed disinhibited attachment at six still displayed the behaviour. Children adopted after 6 months showed significant delays in intellectual development, scoring lower on IQ tests and with difficulty concentrating, with those adopted after 24 months showing an average IQ of 77. In a small number of cases (only in the Romanian orphans), quasi-autism tendencies were identified, with children having problems understanding the meaning of social contexts. Intellectual problems continued at the 15 year follow-up.

• Rutter concluded that adoption within the first six months of life is important to stop the effects of deprivation and privation from becoming permanent, however there may be some recovery as children develop, suggesting that the critical period is more of a sensitive period.

Evaluation

• Research on the negative effects of institutionalisation changed policies around adoption and care in orphanages and other institutional settings. Key workers in institutions ensured a higher level of care, and adoption at an early age and suitable care from the new family ensured by regular follow-ups were prioritised.

• This study provided detailed measurements through the use of interviews and observations of the children’s behaviors. The problem is that it is not easy to find out information about the institutional experience for the child and therefore we don’t know the extent of early privation experienced by these children.

• Another problem with this type of study is that once the children were adopted they may not wish to take part in the study anymore so the results would not be representative

• Children not randomly assigned to adoption and control groups, but were selected by the new parents, so it could be that the more sociable children were more likely to be adopted.

• This research is longitudinal so still ongoing, so long-term effects on Romanian orphans are still to be found (i.e: it could be that the children with disinhibited attachments at 11 catch up to their peers in development in early adulthood).

• Goldfarb (1947) compared 15 children brought up in children’s homes until beyond 3 with another group that had been fostered at around 6 months. At 12 years old he found that early fostering led to significantly higher levels of mental ability (IQ) and social skills.

Evaluation

• Small sample used, so individual differences may have had a disproportionate effect, children not randomly assigned

• Hodges & Tizard (1989) conducted a longitudinal natural experiment using 65 children had been placed into institutional care before the age of 4 months, where there was a no-attachment policy.

• By 4 years, 24 had been adopted, 15 returned home & the remaining 26 were still in the institution. Assessments were taken at ages 8 & 16 years. A control group of normal children were used. Data was obtained through interviews with the adolescents and their mothers (and sometimes fathers).

• A self-report questionnaire on social difficulties was completed by adolescents and finally, teachers completed a postal questionnaire, focusing on the adolescents' relationships with teachers and peers. The findings revealed that maternal deprivation was overcome to a large extent by adopted children, with them going on to develop stronger and longer lasting attachments to adoptive parents once placed in families in comparison to restored and institutionalised groups who made limited recoveries.

• However all three groups were more oriented towards adult attention, and had more difficulties with peers and fewer close relationships than a matched control group of adolescents, suggesting that early institutional experience had damaging long-term effects, but that some effects could be overcome with appropriate substitute care.

• It can be concluded that it does not support the Maternal Deprivation Hypothesis as the findings demonstrate that recovery was possible given the correct circumstances in the environment. 

• The children who were adopted may have had adoptive new families that made a considerable effort to attach to them, where as the restored children returned to the same difficult circumstance.  This demonstrates that the subsequent after care can affect the recovery from privation.

• It can be concluded to support the Maternal Deprivation Hypothesis as the findings show that both groups of children have difficulties outside of the family. 

• As they both experience this difficulty it shows that privation has a lasting impact upon later development.

Evaluation

• Hodges and Tizard’s research can be seen to be ethical, as it used a natural experiment meaning that the independent variable was naturally occurring, rather than being deliberately manipulated by a researcher.

• Lacks random allocation – As children were already placed in the institution, participants were not randomly allocated to conditions, which means that individual differences between the children could influence the findings in unanticipated ways. 

• The research appears to suggest that positive subsequent care can minimise the harmful effects of privation, however, the adopted children might have been adopted because of personal characteristics such as apparent resilience or being more attractive in some way. These characteristics might explain why they made a partial recovery rather than the fact that they were adopted, which lowers the internal validity of the research.

The influence of early attachment on childhood and adult relationships, including the role of an internal working model

• According to Bowlby (1969) later relationships are likely to be a continuation of early attachment styles (secure and insecure) because the behavior of the infant’s primary attachment figure promotes an internal working model of relationships which leads the infant to expect the same in later relationships. This is the continuity hypothesis.

Evaluation

• Considerable evidence has supported this view. For example, the Minnesota study (2005) followed participants from infancy to late adolescence and found continuity between early attachment and later emotional/social behavior. Securely attached children were rated most highly for social competence later in childhood, were less isolated and more popular than insecurely attached children.

• An alternative explanation for continuity in relationships is the temperament hypothesis which argues that an infant’s temperament affects the way a parent responds and so may be a determining factor in infant attachment type. The infant’s temperament may explain their issues (good or bad) with relationships in later life. For example Manlove et al. (2002) fathers are less likely to be involved with their infant if the infant has a difficult temperament.

• Hazan and Shaver (1987) - Hazan and Shaver aimed to understand if childhood attachment type would predict the behaviours and attitudes of adults with regards to romantic relationships. Hazan placed a 'love quiz' in a newspaper which measured individuals' attachment experiences as children and current attitudes to love and romantic relationships on over 600 American males and females. 56% of respondents were secure, 25% avoidant (insecure-avoidant), and 19% were anxious (insecure-resistant). They were judged secure if they balanced closeness and independence, avoidant if they were mostly independent, and anxious if they were not very independent.

• There was a strong correlation between childhood attachment type and adult attachment type. Securely attached adults believed in enduring love and were less likely to get divorced, Insecure-avoidant types were fearful of emotional closeness and believed love was not long-lasting, and Insecure-resistant types were preoccupied by love, fell in love easily but had trouble finding real satisfaction with partners. Percentage of types was similar to those typically found in Ainsworth’s strange situation.

• This supports the hypothesis that parenting styles create an internal working model which influences attachment type which has long-lasting consequences on personality and motivation. 

• So Hazan and Shaver concluded that adult relationships are a continuation of the infant types according to Ainsworth's secure and insecure resistant and avoidant categories.

Evaluation 

• Hazan's study has the advantage of having a large sample meaning we can generalise results, however, as it was conducted in the USA, it only measures western-style relationships and is, therefore, ethnocentric. 

• It is correlational so does not show cause and effect.

• There could also have been demand characteristics with participants giving socially desirable answers: e.g. exaggerating how positive their relationships were rather than being honest. 

• This study supports that our adult relationships are dictated by our early experiences, an example of determinism, which ignores the free-will that people have to overcome past attachment experiences and create positive relationships with partners. 

• Not representative as is self report and a volunteer sample.

• McCarthy (1999) studied 40 women who had been assessed for attachment style while in infancy and found that those assessed as secure in infancy were more likely to be in secure adult relationships (both friendships and relationships) and also had higher self-esteem. Insecure-avoidant had particularly poor romantic relationships, while resistant had poor friendships. Both insecure groups were more likely to have lived with a deviant partner. This suggests that early attachment style is linked to the quality of both adult relationships and friendships.

• By using women assessed in infancy we can avoid the issues of accurate recall found in Hazan and Shaver’s study

• Poor Parenting - Harlow's monkeys went on to become poor parents themselves. 

• Research by Bailey (2007) found that the majority of women had the same attachment classification both to their babies and their own mothers, suggesting that attachment style is passed on intergenerationally.

• Zimmerman (2000) studied a group of children growing up in Germany, and found that child attachment type did not predict adult attachment type. Life events such as the divorce of parents or parental illness or death had much more influence on later security. This suggests there are individual differences in the influences of early attachment type on later relationships. It suggests that there are other, perhaps more important, influences on adult relationships. 

• Knowledge of the internal working model can be used by schools to help children alter their IWM and to address issues such as bullying and loneliness, and potentially to help people with relationship problems.