Arrangement in grey and black or Whistler's mum as it is usually known!



Home AS A2 Links


Explanations of Attachment

Bowlby's theory of attachments

Strange Situation

Deprivation and Privation

Day care and social development





Early social development





Fans of Tom and Jerry may recall a 1955 episode called ‘That’ my mommy’ in which a duck egg slips from a nest rolls under an unsuspecting Tom and hatches out to produce Quacker a new born duckling.  In the absence of his real mother Quacker assumes that Tom is mum and spends the rest of the episode trying to foster an attachment whilst an ever desperate Jerry attempts to rescue Quacker from the frying pan! 


Image:Quacker The Duck.JPG


Now you may think that even though we use the term ‘bird brained’ as a term of abuse, that no creature would be so stupid, however, the ‘and finally’ light-hearted items you get at the end of news bulletins would suggest otherwise since there are often tales of strange attachments of this type.  I seem to remember one from a few years ago in which an owl had adopted a young kitten for example. 

Behaviour of this sort in which a young creature forms an immediate attachment with a mother figure is called imprinting and seems to be the norm in many species, but can it be applied to humans?  Why do we form attachments in the first place and what happens if they are broken?  How does this relate to day care and affect social policy?  These are just a few questions hopefully answered in the pages that follow. 

What is an attachment?

There are a number of definitions offered by different psychologists but these two seem typical:

 Schaffer (1993)  ‘A close emotional relationship between two persons, characterised by mutual affection and a desire to maintain proximity.’

In a similar vein, Maccoby (1980) describes four characteristics of an attachment:

  1. Seeking proximity, the desire to be close to the person to whom you are attached.
  2. Separation anxiety, the distress that results from being separated from that person.
  3. Pleasure when reunited, relief and observable joy when reunited with them.
  4. General orientation of behaviour towards the caregiver, the child’s awareness of where the person is, and the reassurance they feel by them being close.

Remember these characteristics since they will reappear in later studies e.g. the ‘’Glasgow babies’ and the ‘strange situation.’

Why do we form attachments?

This is best broken into short term (immediate) benefits and longer term benefits.  Both, to some extent, can be explained in terms of benefits to the reproductive success of the species or individual (depending whether you favour Darwin or Dawkins; for the biologists).

Short term benefits

Most species emerge into the world unable to fend for themselves so require lots of assistance in the early stages of life.  This is particularly true of the human infant that is helpless for many early years of its life. 

Forming a close attachment with a caregiver therefore ensures that the offspring will be fed, protected from harm, educated in various techniques of survival and kept warm.  It seems likely that the infant’s need to form an attachment is innate. It is also worth considering that it is also in the interests of the parent(s) to protect their offspring from harm.  Again in evolutionary terms they, particularly the mother, have invested a lot of time and energy in producing offspring, it is in their best interests to see the fruits of their labours reach maturity and reproduce themselves.  It therefore seems likely that adults also have an innate tendency to form attachments with their offspring.


Young horses are on their feet within minutes

Humans are helpless for much longer

Long term benefits


These are not so apparent.  Bowlby (1969) proposed that early attachments provide a template or schema, or a set of expectations that allow us to build other attachments later in life.  He called this template the ‘internal working model.’  Early attachments are our first feel for what constitutes an emotional bond and we use this in later life as a basis for other attachments. 

It seems they also act as an anti-incest device.  Incest, as well as being morally repugnant in all societies, is biologically very dodgy, leading to greatly increased risk of genetic abnormality.  Any species or individual that avoids incest is therefore more likely to successfully propagate its offspring. 

The ‘Westermarck effect’

This is more for general interest than inclusion in an answer.  Westermarck (1891) found that children that spend a lot of time with each other in the first 6 years of life, will not go on to form sexual relationships with those same people when they reach maturity.  Westermarck believed that this is an anti-incest device and in normal circumstances prevents us forming sexual relationships with close relatives.

Shepper (1971) found that not one of the 3000 Israeli marriage records he studied was for couples who had been reared together, as children in a Kibbutz. This provides evidence for the Westermarck effect but also for this concept of early attachments influencing later emotional and romantic attachments.


 How do attachments develop?

It seems that different rules apply to animals and humans.


1. Imprinting as proposed by Konrad Lorenz

This idea comes from the work of ethologists on non-human animals, particularly birds.

Just as physical characteristics of various species develop at certain stages of growth, the ethologists claim that perhaps attachments will only form during similar critical periods.  The most famous examples of this are birds forming attachments to the first thing they see upon hatching.  Think of Quackers!

Ethologists refer to the phenomenon as imprinting.  It has the following characteristics:

It occurs during a critical period.  With ducklings the strongest tendency, according to Lorenz, is between 13 and 16 hours after emerging from the egg.  If no attachment has developed within 32 hours it’s unlikely any attachment will ever develop. 

It is irreversible: once the bond is formed it cannot be broken, nor can its effects.

It has consequences both for short term survival and in the longer term forming templates for later relationships. 



Konrad Lorenz (1935) split a clutch of goose eggs and got half to be hatched by their mother and the rest were placed in an incubator and saw Konrad on hatching.  The second group subsequently followed Konrad everywhere and became distressed if they were separated from him.

Although of fragile and amiable appearance, Lorenz’s politics did leave a lot to be desired.  In 1938, at the age of 35 he joined the Nazi party and devoted his research to the aims of the National Socialists.  Some of his later research supported the idea of ‘racial hygiene’ proposed by the Nazis. 

In later life, he joined the Austrian Green Party and distanced himself form his earlier politics.  Some have claimed that his research suggesting a genetic basis for many behaviours may have been inspired by his National Socialist beliefs though there is no evidence of any fudging or doctoring of results. 

Lorenz and his imprinted ducks

Immelmann (1972) imprinted newly hatched zebra finches on Bengalese finches.  Later in life the zebra finches ‘preferred’ to mate with Bengalese finches rather than their own species.  Evidence for the long term effects. 



Guiton et al (1966) disagreed with the irreversible nature of imprinting.  They imprinted newly hatched chickens onto yellow rubber gloves (Marigolds if you prefer).  In later life, just as the theory predicts, they did indeed try to mate with the gloves.  However, when they had chance to spend time with others of their own species, they developed a ‘taste’ for mating with these instead.

Most criticisms of the imprinting or critical period theory however, are based on its application to humans.

Few would argue for such a rigid period of attachments in humans.  However, some, for example Bowlby, have argued for a ‘watered down’ version, referred to as a sensitive period.  The idea being that there is a time in an infant’s life when it is most likely to form an attachment, but it can continue to form them outside this period.  Bowlby argued that our need to form attachments was innate and would occur in the sensitive period between the ages of 1 and 3 years.

Skin to skin hypothesis

Klaus and Kennel (1976) looked at two groups of newly born infants:

  • Group one allowed contact with mother during feeding in the first 3 days
  • Group two allowed extended contact with mother lasting several hours a day

One month later when they returned to the hospital mothers in group two were found to cuddle their babies more and make greater eye contact.   The effects were still noticeable a year later.

Klaus and Kennel believed that this showed that greater contact led to stronger and closer bond formation between mother and child and provided evidence for the sensitive period.


 Note:  This research led to a change in social policy with hospitals encouraged to room mother and infant together in the days following birth rather than the previous tendency to keep them apart.  Also fathers were encouraged to be present at birth so that they too could form an early attachment



Durkin (1995) pointed out that most of the mothers were unmarried and from poor families so results may be difficult to generalise to the general population.  Perhaps the closer bond was due more to the extra attention given to them during the experiment. 

(Note that this is another one of those ‘catch-all’ evaluation comments that can be applied in any situation when participants are chosen from a narrow grouping, e.g. students).

However, De Chateau et al (1987) repeated the procedure on middle class Swedish mothers and found very similar results. 

Even in non-human animal species it seems that imprinting is more flexible than Konrad Lorenz thought.  Sluckin (1965) believes that the sensitive period is a time when a young animal is most likely to form an attachment, but that such responses can be learned at any stage in life.


Schaffer and Emerson’s stages of attachment

So having decided that humans have a sensitive period for development we next consider the process by which the human bond develops


Schaffer & Emerson and the Glasgow babies (1964)



To find the age at which attachments start and how intense these were.


They studied 60 babies from a working class area of Glasgow, observing them every four weeks for the first year and then again at 18 months.


They measured strength of attachment by:

  1. Separation anxiety: how distressed the child became when separated from the main caregiver (which suggests an attachment has been formed) and
  1. Stranger anxiety: distress shown when the child was left alone with an unfamiliar person (which suggests that the child can recognise familiar and unfamiliar people).


(Compare these to Maccoby’s defining characteristics of an attachment (page 1) and to Mary Ainsworth’s characteristics for measuring type of attachment in the strange situation).


The first specific attachment was formed by 50% of infants between 25 and 32 weeks

Intensity peaked in the first month following the onset of the first attachment.

Multiple attachments began soon after the first attachment had been formed.  By 18 months 31% had five or more attachments, e.g. to grandparents etc.


Conclusion: human attachments develop in three distinct stages:


Stage and age



(0-6 weeks)

This is short lived.  Attention seeking behaviour such as crying and smiling is not directed at anyone in particular, suggesting attachments could be made with anyone.

Indiscriminate attachment

(6weeks to 7 months)

Similar in that the child seeks attention from anyone and is happy to receive attention from anyone.  However, preferences are shown to familiar faces that elicit a greater response from the infant.

Specific attachments

(7 to 11 months)

Child is primarily attached to the main caregiver.  If they are separated the child becomes distressed and the child is wary of strangers.  (See later notes on the Strange situation).



Since babies were observed in their own homes (a natural environment) we can assume that the study is high in ecological validity; the findings can be generalised to the real world.

However, accuracy of data collection by parents who were keeping daily diaries whilst clearly being very busy could be questioned. 


Explanations of attachment

Why do babies form attachments?

Learning Theory

According to behaviourists, behaviour is not innate but learned.  Learning can be due to associations being made between different stimuli (classical conditioning) or behaviour can be altered by patterns of reinforcement (reward) and punishment (operant conditioning). 

Neo-behaviourists suggest that we learn by watching others (social learning theory or SLT).  Social learning of this sort is particularly powerful when we see others being reinforced or punished for their actions. 

Note: think of conditioning as learning. 

Behaviourist explanations of attachment:

Operant conditioning

Dollard and Miller (1950) suggested that the attachment was due to drive reduction.  Hunger and cold have a strong motivating affect on the child, driving the child to satisfy its need by eating or seeking warmth.  Obtaining food or warmth results in drive reduction which in itself provides reward for the child. 

Hunger and cold (discomfort) are referred to as primary drives and food and warmth are primary reinforcers.  The person supplying the food and warmth (usually the mother) becomes associated with the food and warmth and acts as a secondary reinforcer.  The attachment occurs because the child wants the person providing the food and warmth. 

Note: When the child is cold and hungry it cries.  This is unpleasant for the mother (punishment) who is likely to feed and cuddle the child.  The child stopping crying acts as a negative reinforcer for the mother (something unpleasant has been taken away).  Negative reinforcers make the mother’s behaviour, feeding and cuddling, more likely in future!

Classical conditioning

This offers a similar but simplified explanation of how food provides attachment.  The child simply associates food and mother together, much as Pavlov’s dogs associated bell and food together. 

If you want this in technical terms:

  • Food is an unconditioned stimulus that produces an unconditioned response (pleasure).
  • At the outset, mother is a neutral stimulus who produces no response (pleasure)
  • However, because she is continually paired with the unconditioned stimulus (food) she slowly becomes associated with it until eventually mother alone can produce pleasure.
  • Mother has now become a conditioned stimulus and the pleasure she brings is a conditioned response.

(Again think of conditioned as learned whereas unconditioned is something that was there all the time). 


As always the behaviourist explanation is reductionist because it takes a complex human behaviour and tries to explain it in the simplest terms possible.  It does not consider any internal processes or seek to explain the emotional nature of attachments simply how they arise as behaviours. 

The behaviourist theories of attachments (and Freud’s psychodynamic) are sometimes referred to as cupboard love theories because of their emphasis on food and feeding.


Is food needed to create an attachment?

In the Glasgow babies study (Schaffer and Emerson): 39% of the babies formed their first attachment with someone other than the person who fed them (e.g. grandparent).  This would suggest that food is not the main requirement for forming attachments as the behaviourists suggest.

Harlow’s monkeys (1959): ‘The origins of love.’

Harry Harlow used rhesus monkeys in his research into learning and noticed that many of the young monkeys kept in isolation became distressed when he cleaned out their cages.  It seemed that the monkeys were forming an attachment with the sanitary towels he used to line the base of the cages. 


To find out whether food or comfort was more important in developing attachments. 


Harlow carried out a number of variations using sixteen young isolated monkeys.  Some where kept in cages with both a wire surrogate mother and a softer one covered in Terry cloth whilst others were kept in cages with just one.  Sometimes the monkeys would be fed by the wire mother and other times by the softer cuddlier mother.  However, the important variation was the one with a monkey in a cage with a wire mother that provided food and a Terry cloth mother that didn’t (providing the monkey with a choice; food or comfort). 



 Harlow noticed that the monkeys would spend most time clinging to the cloth mother and occasionally feeding from the wire mother.  When the monkeys were stressed by a mechanical toy banging a drum the monkeys would always run to the cloth mum for safety suggesting an attachment.  Also the monkeys with only wire mothers produced water faeces which was attributed to stress. 


The evidence suggested that warmth and comfort rather than food were more important in nurturing an attachment and provided scientific evidence against the behaviourist (and psychodynamic) cupboard love theories. 


Clearly because Harlow used monkeys it is difficult to generalise the findings and conclusions to humans.  However, the Glasgow babies study also suggested that babies didn’t always form attachments with the food provider.

There are also serious ethical issues with this study.  The young of an intelligent species are reared in isolation and not allowed to form attachments with their own species.  The monkeys grew up unable to socialise with other members of their species and were bullied as a result. 


Social Learning Theory (SLT)

This is similar in some respects to learning theory, in that both emphasise the role of reinforcement (an action that is rewarded being more likely to be repeated).  However, SLT emphasises the role of imitation.  We watch others and if they are rewarded for their behaviour we are likely to copy it ourselves.  Hay and Vespo (1988) suggested that attachments develop because parents teach their children to love them.  This can be achieved in three ways:

  1. Modelling: children copy the affectionate behaviour that they see between their parents.
  2. Direct instruction: parents teach their children to be affectionate.
  3. Social facilitation: parents watch their children and encourage appropriate behaviours.


Durkin (1995) does not believe that SLT can explain the intensity of emotion that the attachment produces.

On the plus side, the theory can be said to be influential in that it has stimulated a lot of research into the interactions that take place between parents and their children.


Bowlby’s theory of attachment

It is impossible to study attachment and child development without considering Bowlby’s work.  His theories on attachment and on maternal deprivation have been some of the most influential writings on the topic.  Given the nature of his theory I thought it would be worth including an abridged biography of Bowlby’s early life which sheds light on his motivation to produce the sort of theories he did.

Bowby’s theory

Bowlby worked for many years as a child psychoanalyst so was clearly very influenced by Freud’s theories and child development.  However, he also liked the work of Lorenz on the innate nature of bonds through imprinting and combined these two very different ideas to produce his own evolutionary theory of attachments.  Bowlby believed that attachment is innate and adaptive.  We are all born with an inherited need to form attachments and this is to help us survive.  In line with Darwin’s theory of natural selection, any behaviour that helps you survive to maturity and reproduce yourself will be maintained in the gene pool.  In human terms, the newborn infant is helpless and relies on its mother for food, warmth etc.  Similarly the mother inherits a genetic blueprint that predisposes her to loving behaviour towards the infant.

Bowlby believed that an attachment promotes survival in 3 ways:

  1. Safety: the attachment keeps mother and child close to each other.  Separation results in feelings of anxiety.
  2. Safe base for exploration: the child is happy to wander and explore (necessary for its cognitive development) knowing it has a safe place to return to if things turn nasty.  This also develops independence necessary in later life. 
  3. Internal working model: This was based on Freud’s idea of the mother-child relationship acting as a prototype fro all future attachments.  Bowlby believed that this first relationship forms a template or schema that gives the child a feel for what a relationship is.  It uses this in future years to develop other relationships and is particularly important in determining the parenting skills in later life. 

Other aspects of the theory (deary)

Sensitive period

Being innate Bowlby believed there would be a period in which they were most likely to develop, similar to the critical period for imprinting.  Unlike a critical period (the only time in which an attachment may form), a sensitive period suggests a time when they are most likely to occur.  Bowlby believed that for the human infant this was between the fourth and sixth month.  After this it becomes ever more difficult for the child to form a first attachment.


Once made the attachment cannot be broken. 

Social releasers

Being innate the child has built in mechanisms for encouraging care-giving behaviour from parents.  Children have ‘baby faces’ and their noises and facial expressions such as smiles encourage contact.  It seems that adults are genetically primed to respond to these releasers by offering care and affection. 


Babies’ smiles are powerful things leaving

mothers spellbound and enslaved.  Who

can doubt that the baby who most readily

rewards his mother with a smile is the one

who is best loved and best cared for?’ 

                                   Bowlby 1957. 


Continuity hypothesis

The internal working model ensures that early attachments are reflected in later relationship types.  For example, a secure attachment as a child leads to greater emotional and social stability as an adult, whereas an insecure attachment is likely to lead to difficulties with later relationships.  As already mentioned this is likely to be reflected in the parenting style when the child matures and has children of their own.


The argument here is not as straight forward as it first appears.  On the face of it the debate is between many attachments or just the one.  However, Bowlby, who was in the ‘one’ camp, did not actually believe that only one attachment was formed, rather that there was only one primary attachment.  The ‘many attachments’ approach believes there are many attachments and that they are all similarly important to the child.  The text also claims that Bowlby did not believe that the main attachment had to be the mother, saying that his words ‘maternal’ and ‘mothering’ were not intended to mean mother!                       

Thomas (1998) has made it clear that children will benefit from a variety of attachment styles provided by different caregivers, so for example an attachment to a father figure will provide benefits to the child that a mother alone could not provide.  In Caribbean and European culture, infants seem to form many equally important attachments to different people.

Bowlby (1969) claimed that there was a hierarchy of attachments, with a primary caregiver, usually the mother at the top.  The Efe, an African tribe, share the care of their children so that women in the village breast feed each another’s children.  However, the infants still go on to form their primary attachment with their biological mother.

Schaffer and Emerson’s Glasgow babies study found that nearly a third of infants had five or more attachments by the age of 18 months. 


Other evaluation points to make when discussing Bowlby

Bowlby’s theory has been very influential.  It has been widely studied with some researchers agreeing, some suggesting modifications.  It has been widely applied in practical situations, particularly in hospitals, children’s homes and fostering policy.


 Bowlby seems to concentrate on the role of the mother and neglected the father believing the latter to be of little significance.  Later research has shown that the father can play a useful role and lamb (1983) suggests that often children prefer the rough and tumble play they get with the father. 



Similarly Bowlby seemed to overlook the relationships the child develops with its brothers and sisters.  Schaffer (1996) describes these as horizontal relationships as opposed to the vertical relationships with parents, teachers and other adults. 

Research for the internal working model is at best mixed.  Zimmerman et al (2000) assessed attachment style of children ages 12 to 18 months and then in a longitudinal study checked again at the age of 16 years (using interviews to determine the relationship the child had with its parents).  They found that early attachment style was not a good predictor of later relationships and also discovered that life events such as parental divorce had a much greater impact.

A poor early start can be overcome by positive experiences at school and good adult relationships (Rutter & Quinton 1988).

The following study does provide support for the idea of an internal working model but is itself a very poor piece of research:


Hazan and Shaver (1987): The Love Quiz’ (pronounced lerrrrve obviously)


The researchers asked people to volunteer to take part in the study. 

They were given 2 questionnaires, one to determine their early relationships with parents, the second their later, adult romantic attachments.


They found three basic types of childhood attachment and related these to later adult attachments:

Type of bond

Childhood experience with parent

Adult experience with partner



Close warm relationship with parents and between parents

Secure, stable and loving relationship with partner



Mother was cold and rejecting


Fear of intimacy, emotional highs and lows, jealousy


Anxious ambivalent

Father was perceived as unfair


Obsessive, jealous and emotional highs and lows



Early attachments do affect later, romantic attachments.

Evaluation of the Love Quiz

As I said a poor piece of research because of the following reasons:

Self selecting sample:  the participants volunteered after reading an advert in the Rocky Mountain News.  This is a poor way of selecting participants since you are not getting a cross section of the public.  Using this sampling technique, for example, you are going to get people with an ‘axe to grind’ or with extremes of experience or opinion.

Questionnaire:  People tend not to answer truthfully, particularly on issues of relationships, instead wanting to make themselves look good.

Retrospective:  As we saw in memory our recollection of past events is not reliable, so it seems unlikely that people’s memory of their childhood experiences will be accurate

Cause and effect:  The researchers have shown a relationship between early attachments and later ones and are assuming that the childhood experience has caused the adult experience.  However, other factors could be involved.  Kagan (1984) suggested the temperament hypothesis.  Children with a pleasant disposition are more likely to form warm relationships with parents and later in life, assuming they maintain their ‘niceness’, will form more loving relationships


Note: from a practical point of view, poor research like this is good when it comes to writing essays or discussing points.  You can explain why the study offers support for Bowlby’s IWM but then criticise it using one or more of the points above and then importantly suggest that these criticisms question the extent to which it can really be seen as reliable support. Easy AO2 marks!


A better study but with similar findings is McCarthy (1999).  Forty women (aged 25 to 44) who had experienced insecure attachments as children were given various tests (including Hazan and Shavers) to determine the quality of their adult romantic attachments.  Women classed as avoidant as children tended to have less successful romantic attachments whereas those classed as ambivalent were more likely to have problems forming non-romantic friendships. 



A brief biography: for background interest only

John Bowlby’s father (Sir Anthony Bowlby) lost his own father at the age of five and spent much of the rest of his life caring for his mother (John’s grandma).  John himself, as was common for wealthy families at the time, was reared by a nanny til the age of four when she left.  According to Bowlby, his mother was cold and reacted to his needs in the very opposite way that you’d expect a mother to react.  At the age of seven, again as was, and still is common, John was sent to boarding school, so again was separated from friends and family.  In the introduction to one of his many books Bowlby quotes Graham Greene;

‘Unhappiness in a child accumulates because he sees no end to the dark tunnel.  The thirteen weeks of a term may just as well be thirteen years.’

It is very clear that his young life was not happy.  He experienced many separations, including his father going off to war when he was seven.

He studied psychology at Cambridge but took time off, spending six months in a school for maladjusted and delinquent children.  He later referred to this as the most important six months of his life.  Whilst there he noticed how many of the children had lost their mothers at a very young age.

He later trained in medicine but didn’t enjoy the experience of medical practice.  In 1939 he raised concerns about the desirability of evacuating young children and separating them from their mothers. 

Unusually for a psychoanalyst (or any other psychologist for that matter) he was keen to incorporate other approaches into his theories.  He was particularly fond of the work of Konrad Lorenz (mentioned earlier) and his ethological work on the evolutionary advantages of attachments.  According to Ainsworth, Bowlby’s theory appeared as a flash of inspiration after he had read Lorenz’s “King Solomon’s Ring.”




Types of attachment including insecure and secure attachment

Much of what we’ve seen in Psychology so far looks at ways in which we are similar.  It has tried to find models or theories that explain human behaviour but without considering ways in which our behaviours differ.  An important area in the subject however is individual differences which, as the title suggests seeks to explain the differences we find between people.

This section seeks to explain how:

·         Individuals differ in the types of attachment they form

·         Different cultures influence the types of attachment we form

The Strange Situation

This is a method devised by Ainsworth and Bell to measure the type of attachment that a child has formed.  It uses many terms and concepts that you should already be familiar with from earlier work in the topic, e.g. Glasgow Babies and the Love Quiz.

The experiment is set up in a small room with one way glass so the behaviour of the child  can be observed.  Children were aged between 12 and 18 months.  Each phase of the procedure lasts 3 minutes and a session progresses as follows:

·         Parent (or caregiver) enters room with child, child explores for 3 minutes

·         A Stranger enters and joins the parent and infant, talks to mother

·         Parent leaves the infant with the stranger

·         Parent returns and the stranger leaves.  Parent settles the infant.

·         Parent leaves again

·         Stranger returns

·         Parent returns and stranger leaves.

In all the stranger enters on average eight times, more if the child is okay, less if it isshowing signs of distress.

Throughout the procedure the child is observed by a team of researchers who make notes every 15 seconds about the following behaviours:

Proximity and contact-seeking behaviours

Proximity and contact-maintaining behaviours

Interaction-avoiding behaviours

Contact and interaction-resisting behaviours

Search behaviours

To get more reliable results Ainsworth and her co-workers combined the results of several studies so that a total of 106 different child observations were included in the final report.  This combining of studies is called a meta-analysis.

The observations made allowed Ainsworth to judge the child’s reaction to the following three variables:

Separation anxiety: how the child reacts when mother leaves

Stranger anxiety: how the child reacts to being alone with a stranger

Reunion behaviour: how the child behaves when mum returns


Findings and conclusions

From her observations Ainsworth (Mary) concluded that there were three types of attachment.



Secure Attachment




Separation anxiety

Distressed when mother leaves


Infant shows signs of intense distress

Infant shows no sign of distress when mother leaves

Stranger anxiety


Stranger is able to offer some comfort

Infant avoids the stranger

Infant is okay with the stranger and plays normally when stranger is present

Reunion behaviour


Runs to mother and greets her enthusiastically

Child approaches mother but resists contact, may even push her away

Infant shows little interest when mother returns.



Infant cries more and explores less than the other 2 types

Mother and stranger are able to comfort infant equally well

% of infants





However, after reviewing a further 200 tapes of children in the strange situation Main and Solomon (1986) added a fourth type of attachment that they referred to as ‘disorganised.’  The infant’s behaviour is not consistent and shows signs of indecisiveness and confusion.  Sometimes the child will freeze or rock back and forth.


Evaluation of the Strange Situation

Subsequent studies that have used the 'Strange Situation' have found it to be reliable and valid. 

·    Reliability refers to whether you can produce the same results if tested again. 

·    Validity refers to the extent to which the 'Strange Situation' actually measures what it is supposed to measure.


Reliability of the 'Strange Situation' was demonstrated by Main, Kaplan and Cassidy (1985): They tested babies at 18 months and then retested them at 6 years of age.  They found that 100% of the secure babies were still classified as secure and 75% of the avoidant babies were still under the same classification. 

This is called test-retest reliability and checks for consistency over time. 

Ainsworth herself also tested inter-rater reliability (the extent to which different observers score a behaviour in a similar way).  This was also found to be very high.

Validity of the 'Strange Situation.'  Some have argued that the strange situation only measures the relationship between the child and one other person (usually the mother) so rather than measuring attachment type it is simple measuring relationship.  However, Bowlby argued the case for montropy, the idea that there is only one primary attachment figure (the mum) and all others are of lesser importance anyway.

If the test is a valid measure of attachment type then we should be able to use its findings to predict the future stability of a child’s relationships (again assuming Bowlby was right about the internal working model).  Secure attachments in childhood should result in more stable adult relationships. 

Evidence here is not so conclusive, different researchers reaching different conclusions:

Sroufe (1983) found support for its validity.   Infants that were rated as secure went on to become more popular, have higher self esteem, and be social leaders. 

However, Bates, et al (1985) disagreed.  They claimed that early attachment styles did not predict the presence of behaviour problems at 3 years of age.

The strange situation study seems to imply that attachment types influence personality and therefore affect later attachments.  However, the strange situation may actually be testing the relationship between the infant and the caregiver. 

Main and Weston (1981) found that children behaved differently depending on which parent they were with.  This suggests that attachment type is not consistent. 

Other issues with the strange situation

The test was devised by Ainsworth in the USA using American children.  The test is therefore culturally biased.  Desirable attachments in the USA may not be seen as desirable elsewhere.  Nevertheless the test has been used worldwide and used to judge infants in other cultures.  This is an example of imposed etic when we create a theory, test or construct in one culture (usually Western) and impose it on the rest of the world! 

The strange situation also seems to exaggerate behaviours.  Children over-react when placed in the strange situation so do not behave as they would normally in the real world. 

Finally Ainsworth is criticised for over-simplification in her belief that children can be categorised into only three groups.  Other studies have suggested that there big individual differences between children within in attachment group. 

Explaining the different attachment types

Sensitive responsiveness

Ainsworth herself believed that the kind of attachment the child develops is due entirely to the mother.  Secure children have mothers who respond appropriately to the child’s needs by picking up on the signals.  Insecure children on the other hand have mothers that are less responsive and the attachments they develop are coping strategies that enable them to deal with this lack of response. 

Temperament hypothesis. 

Perhaps the reason for a relationship between early attachment and later relationships has nothing to do with the type of attachment formed.  Kagan (1984) believed it was all down to the temperament of the child.  Those who are naturally good at forming relationships do so early in life and form close relationships with parents and this is true later in life as well; because of their pleasant temperament they are more popular with people in later life too. 

Thomas and Chess (1977) thought that children were born with a certain personality type and suggested three main categories;

Easy: eat and sleep regularly and accept new experiences (under 50%)

Difficult: eat and sleeping a problem, as is accepting anything new or different (10%)

Slow to warm up: Take a while to get new to new experiences (15%)


Easy babies go on to form secure attachments.  Slow to warm up babies require a lot more encouragement so will only form secure attachments with patient mothers. 

Belsky and Rovine (1987) found that babies in the first few days of life have certain physiological characteristics that seemed to match later attachment styles.  Calmer and less anxious children at this age were more likely to develop secure attachments a few months later. 

Cross cultural studies

Bowlby believed that attachments were innate so the need to form this initial bond should be genetic and as a result experienced by the infants of every culture.  However, the kind of attachment formed may vary between societies and between cultures depending upon the child rearing techniques seen most desirable within that community.  This section (a favourite for examiners) looks at different patterns of attachment found in other cultures and possible explanations for the differences. 

Ainsworth carried out most of her research in the USA but others have found broad agreement with her findings in other parts of the World (worth mentioning in an essay!).  The ones I’ll mention below are exceptions in that they are different and we shall consider possible explanations for this.

Van Ijzendoorn and Kroonenberg (1988) carried out a meta-analysis combining the findings of 32 other studies of the strange situation from a variety of countries and based on the observation of over 2000 children. 

Note: if the question asks you to describe the procedure of a study into cross-cultural differences in attachment describe the strange situation but emphasise it was carried out in a variety of countries. 



The most notable finding was the similarity in types of attachment across most countries.  Secure attachment is the norm in the overwhelming majority of cultures. 

















However these three countries stood out from the rest.  Germany has a high percentage of avoidant attachments and both Israel and Japan a much higher than normal percentage of resistant



Despite the large number of studies combined in this meta-analysis over half (18 of the 32) were still US.  Only five of the 32 were carried out in *collectivist cultures. 

We also have the issue of imposed etic.  The strange situation was designed by an American, using American children for use on other Americans.  Many researchers have therefore questioned whether it can possibly be suitable for testing the children of other cultures.  Mary Ainsworth assumed that separation anxiety was an indication of secure attachment and this may be the case in some countries such as Britain and the USA.  However, separation anxiety in other societies and cultures may represent other factors.  The strange situation may therefore not always be a suitable measure of attachment and may in fact be culturally specific.  This development of a test for one culture, then being used in unfamiliar cultures is referred to in psychology as imposed etic, and is most controversial in the testing of IQ.

Rothbum et al (2000) suggest that although the need for protection appears to be universal (the same across all cultures) other factors differ.  They believe psychologists should be working at an emic level, producing theories of attachment tailored to fit each individual culture.  For example, the Japanese have the concept of ‘amae’ (depending on another’s love) which according to Doi (1973) leads to a sense of oneness between mother and child.

This might explain why so many infants develop a seemingly more clingy resistant attachment style and become so distressed when separated from mum.

Van Ijzendoorn and Kroonenberg report that differences in attachment within a culture are far greater than those found between cultures.  They conclude that it is wrong to think of everyone in a culture having the same practices.  Within a culture there are many sub-cultures, all with their own way of rearing children.  These may be ethnically or racially based but also may be class specific, for example in the UK the so called ‘middle classes’ having different child-rearing techniques to the ‘working classes.’  The upper classes traditionally have left child rearing to nannies!

However, in support of the strange situation, Bee (1999) believes that the most striking feature of the cross cultural studies is their similarity.  With the exception of the countries mentioned earlier, most countries do seem to have a similar pattern with most infants forming secure attachments and the rest being split equally between avoidant and resistant.


Individualistic versus *Collectivist cultures:

A distinction that crops up frequently in psychology (we’ll see it again in stress and again when we look at abnormality…oh and again when we look at conformity). 

 We are individualistic!  Western society encourages its citizens to be independent and not reliant on others (some would say greedstern (and increasing Eastern) Europe and Australia are all classed as Individualistic. 

Collectivist cultures place the emphasis on the family or on collective rather than the individual.  People are more dependent on those around them, more likely to share worksponsibility and see the family as being more important.  Asian and African countries tend to be collectivist.  Japan is a combination being financially individualistic but collectivist when it comes to social situations (including child rearing). 


 Explanations of these cross cultural differences

 Israeli children were reared in a Kibbutz so were used to being separated from their mother.  As a result they do not show anxiety when their mother leaves.  However, they are not used to strangers so get distressed when left alone with the stranger.  This explains the high percentage of resistant behaviour.  Fox (1977) tested children in the strange situation using either their mother or the metapelet (Kibbutzim nurse).  The children appeared similarly attached to both except for reunion behaviour where they were more pleased to see mother. 

Japanese children show similar patterns of attachment to the Israeli children but for different reasons.  Japanese children are very rarely left by their mother.   So the distress they show when she leaves is probably more due to shock than it is to insecure attachment.  The distress they show when left alone with the stranger is also more likely to be due to absence of the mother.

The German study highlights a high percentage of avoidant behaviour, typical of independent children.  This is not surprising given that Grossmann et al (1985) say that German parents seek ‘independent, non-clingy infants, who do not make demands on parents, but obey their commands.’   Ah if only it were true in Britain! 

Being able to explain the cultural differences is very useful AO2 stuff.   

But don’t forget to emphasise the similarity between most other cultures! 


Disruption of attachment

There are a number of situations in which an attachment can be broken either temporarily, for example by hospitalisation or permanently through death.  A broken attachment like this is referred to as deprivation.   

Unfortunately there are also cases of children being so badly treated, perhaps being kept in total isolation for many years, that they never have the opportunity to form an attachment and this is called privation.

Bowlby did not differentiate between the two in his maternal deprivation hypothesis.  Recent studies however suggest:

1.       Deprivation and privation are distinct, believing that the long term consequences of privation are far more severe than the long term consequences of deprivation.

2.       Children are generally far more resilient to early separation than Bowlby originally proposed.  Bowlby himself later changed his views adopting this line.


Deprivation (separation)

Short term effects


Robertson & Bowlby (1952) PDD Model

It is important to bear in mind that the initial distress varies according to the temperament of the child, especially its security.  For example if a child is used to short term separations they know the carer will return so the effect of the separation is less severe.

The research for the PDD model was carried out on children aged between 1 and 4.

Protest:  Child cries and calls for its mother.  Panic is usual.  This can last from a few hours to a few weeks!

Despair:  Child becomes apathetic (i.e. uninterested in what is happening around them).  They continue to cry occasionally and call for mother.

Detachment:  The child cries less and is more interested its surroundings.  Onlookers may think that the child is getting over the separation, whereas in fact the child is hiding its feelings.  When the mother returns the child shows little interest and may even be angry or push the mother away.  However, the attachment is soon rebuilt.

It is important to note that the carer may also suffer similar patterns of distress


James and Joyce and the not so famous five

As is often the case researchers in psychology get all loved up with fellow workers.  James and Joyce Robertson fostered four children (Jane, Kate, Lucy and Tom) whilst their mothers were in hospital.  They carried out a detailed observation of how the children coped with this temporary separation.  Four of the children settled well and soon adapted to the new regimen which the Robertson’s had tried to keep as similar to the children’s home environment.  However, a fifth child, John, was placed in a residential nursery for nine days.  At first all appeared well but after two days John’s behaviour deteriorated has he became attention seeking and spent lots of time crying.  Not getting the attention he needed he developed an attachment with a teddy bear.  Whereas during the first week he had appeared pleased to see his father by the second week he seemed apathetic and would largely ignore him.  When his mother finally returned from hospital John screamed and tried to distance himself from her.  


Barrett (1997) looked at Robertson & Bowlby’s original study and found the situation to be more complex than they had suggested.

a.       If the child is securely attached it copes better than the model predicts

b.       If the child is avoidant then it gets the full blown PDD effects.

The effects are not inevitable as was initially predicted.  The study by Robertson and Robertson shows that if steps are taken to minimise the separation, for example discussing their real mother and keeping to familiar routines, the effects can be small.  

As a result of research by psychologists, the care of children in hospitals is very different to how it was 30 years ago.  It is recognised that good psychological care is vital for physical rehabilitation so steps are taken to minimise bond disruption.  More regular, even around the clock, visiting hours are now available.


Long term effects  

Bowlby’s 44 Thieves and MDH

Bowlby interviewed the children, and their families, who attended a clinic where he worked.  He compared the backgrounds of 44 juvenile thieves with the background of 44 other non-delinquent children.

Findings: 32% of the thieves were diagnosed by Bowlby as having affectionless psychopathy, the main symptom of which is lack of moral conscience.  Most of these had experienced separation for at least one week before the age of 5.

Conclusion:  Separation in early life led to long term ill effects, particularly adversely affecting emotional development.

Evaluation: The data collection is retrospective (i.e. the children and their parents had to think back many years to the child’s younger days).  This can produce inaccuracies (as you will appreciate being experts on the memory process!).

Some of the children were only separated for short periods, so it is difficult to believe this could have caused such emotional disturbances.

The results are correlational, so we cannot prove cause and effect.  Bowlby assumed that the early separation had caused the later disturbance, but many other factors could be responsible.  For example children from poor backgrounds are more likely to be hospitalised.  Children from poor families are also more likely to become delinquent.  Attachments may not be the cause but poverty may be.


From this study Bowlby developed his Maternal deprivation hypothesis:

Breaking of bonds in early life leads to intellectual, social and emotional problems in later life.  Note, by ‘maternal’ it is usually assumed that Bowlby meant mother figure.  Bowlby originally believed the effects to be permanent and irreversible.

ADDIDDAS (a mnemonic for the characteristics of Maternal deprivation):

Aggression, Delinquency, Dwarfism (no jokes!), Intellectual retardation, Depression, Dependency, Affectionless, Social maladjustment.


Support for the Maternal Deprivation Hypothesis (MDH)

Bowlby’s own 44 thieves study (see above).

Spitz (1945) study of orphanages in S. America.  The children showed little affection with many showing all the symptoms of anaclitic depression e.g. loss of appetite and apathy.

Spitz & Wolf (1946) study of 100 children who had become depressed after hospitalisation.  A full recovery was made if stay in hospital was less than 3 months.  Note this study was carried out in the 1940s when hospital care of children was very different to that found today.

Goldfarb (1947) orphanage study.  The research involved 2 groups of children:

Group 1: spent the first few months in an orphanage and were then fostered.

Group 2: spent 3 years in an orphanage before being fostered, (i.e. had little opportunity to form attachments in early life).

Both groups were tested to the age of 12.


The children who had spent 3 years at the orphanage performed less well on IQ tests were less social and more likely to be aggressive.


Evaluation of maternal deprivation hypothesis

Reversing the irreversible

Later studies have shown that many of the effects of early deprivation can be overcome.  They are not so permanent and irreversible as Bowlby seemed to assume. 

Deprivation or conditions?

Children reared in institutions (Spitz, Goldfarb) were not only separated from parents they were also kept in relatively poor conditions.  This is likely to have added to the effects.

Rutter (1981) in his book ‘Maternal Deprivation Reassessed’ (available in the department) criticises Bowlby on two key counts:

Cause and effect and the issue of poverty

We have a classic case of a study not being able to establish cause and effect.  Bowlby showed that children who are separated from their parents in early life (usually due to spending time in hospital) are more likely to be delinquent when they grow up.  From this he assumed that early breaking of the maternal bond will disrupt later development.  This cannot be proven using Bowlby’s method.  An association or correlation between two variables (in this case deprivation and delinquency) does not establish a definite cause and effect relationship. 

We know, sadly, that poorer people are more likely to resort to crime (delinquency).  We also know, sadder still, that children who are reared in poverty are far more likely to end up in hospital.  Therefore it seems natural to conclude that a third variable, poverty, caused many of the issues for the 44.  Poverty led them to be hospitalised and poverty led them to crime. 

Deprivation or privation?

Bowlby did not distinguish between the effects of separation and privation.  We now know that the long term effects of privation are far more severe than the effects of deprivation.  Rutter believes the effects Bowlby attributed to separation may well have been due to privation.  Children in institutions such as orphanages never had the opportunity to form attachments because of the high turn over of staff so were effectively suffering privation as opposed to the deprivation that Bowlby had assumed.



Privation is the failure to form an attachment and can be caused following the death of both parents (most likely during times of war) resulting in children being raised in institutions or it can be caused by extreme neglect in which children are raised in isolation.  We shall look at the latter first and consider the findings, the ethical issues raised and also the methodological problems and benefits of using case studies in psychology.


Case studies


Genie (as reported by Curtiss 1977)

Found at the age of 13, she had been kept tied to a potty chair for much of her life.  She had been severely punished for making a noise.  When found she had the appearance of a six or seven year old.  Curtiss described her as ‘unsocialised, primitive and hardly human.’

Following her discovery she continued to be mistreated at the hands of doctors and psychologists who were more interested in furthering their own careers than in Genie’s welfare.  She never acquired full language skills and failed to adjust socially.

 Unfortunately we have no way of knowing whether Genie was, as her father suggested, brain damaged at birth.  If this had been the case this could partly explain her lack of progress.




Czech twins, PM and JM (as reported by Koluchova 1976).

PM and JM were male identical twins born in the former Czechoslovakia 1960.  Their mother died at birth.  They spent 11 months in a children's home before being reared by their father and stepmother.  The father was of low intellect and the stepmother was particularly brutal in her treatment of the twins.  They were kept in a small closet or cellar.  They were discovered at the age of 7.  Their speech was poor and they had rickets (due to vitamin D deficiency caused by poor diet), so consequently could not walk.  They were subsequently adopted by two sisters and were well cared for.  They were tested at the age of 14 and showed no long term ill effects.  In later life they both found employment and ‘enjoyed warm relationships.’

Clearly the outcome of these two cases is very different.  However, it does appear that given favourable care a near full recovery from early privation is possible.  There are a number of reasons why Genie’s outcome was not good:

·         The possibility that she may have been brain damaged at birth as her father had suggested

·         The later age at which she was discovered.

·         She had been reared alone whereas the twins had each other.

·         The better care the twins received after being rescued.


Freud and Dann (1951) and the concentration camp victims

Six children whose parents had been murdered in concentration camps were adopted into English families after the war.  Despite having no adult attachment figures in their early lives, having no speech and witnessing all types of atrocities the children went on to make reasonable recoveries. 

If asked to describe the procedure of a study investigating privation do not use case studies! 

Case studies; the issues



Case studies provide ideal opportunities for psychologists to study situations that could not be created in any other way.  However, in so doing they need to keep the best interests of their participants in mind.  This doesn’t appear to have been the case with Genie.  In the case of privation we usually have young children who because of their age or mental state are unable to give full consent to research.  Carers may also feel pressurised into allowing access to psychologists even though it may not be in the child’s interest.  Psychologists on the other hand have a double-obligation dilemma.  They have a duty to carry out research into this area but also a duty of care to those being studied. 


Case studies can provide lots of detailed and intimate information about a particular condition and its causes.  However, by their nature they are one-offs and involve patients who have been exposed to a unique set of conditions and are therefore suffering unique symptoms.  Although a case study can provide detailed information about one individual case it is difficult to then generalise findings to others or to come up with a theory based on what is likely to be one disturbed individual.  The two studies above highlight this perfectly.  Two cases of extreme privation but with very different outcomes and no way of knowing why. 

Cases of institutionalised privation provide less detailed information and appear less interesting, but do allow for much greater levels of control and probably tell us more as a result.



Bowlby famously claimed that a bad home was better than a good institution because of the poor psychological care children receive in such places.  Skodak & Skeels (1949) compared two groups of mentally retarded children brought up in an institution.  One group were transferred to a home to be cared for, the other group remained in the institution.  Those removed to a home showed improvement in their IQ (up from 64 to 91), those remaining in the institution showed a drop in IQ (down from 87 to 61).  I don’t know why their IQs were so different at the start!  Twenty years later the difference was still present.  This would appear to lend support to Bowlby.  However, most studies suggest Bowlby was wrong:


Hodges and *Tizard (1989))

A very useful study both in its findings and in its design! 

This is a natural experiment.  All experiments have an IV and a DV.  Usually the IV is manipulated by the experimenter (for example time spent counting backwards in the Brown-Peterson procedure).  However, with a natural experiment researchers take advantage of an IV that changes naturally, in this case children in care either being fostered or being returned home. 

It is also longitudinal.  The researchers study the same group of children on a number of occasions at different stages in their development. 

Hodges and Tizard (1989)


The aim of Hodges and Tizard's study was to examine the effect of institutional upbringing on later attachments.


Sixty five children in a care home were assessed over a 16 year period.  The participants in the study were all aged 16 and had all been in institutional care until the age of four.  During this time they had not been able to form attachments because of the high turn over of staff.   By the age of two the children had on average 24 different carers each! 

At the age of four:

  • 25 of the children were returned to their biological parents
  • 33 were adopted
  • 7 remained in the institution with occasional fostering

The above categories (form of care) are the IV for this experiment. 

Five main methods were used to collect data on all the adolescents (including those in the comparison groups):

  1. An interview with the adolescent;
  2. An interview with the mother (in some cases with their father present);
  3. A self-report questionnaire concerning 'social difficulties';
  4. A questionnaire completed by the participants' school teacher about their relationships with their peers and their teachers;
  5. Rutter 'B' scale which is a type of psychometric test which identifies psychiatric problems such as depression.


At 16 the majority of the adoptive mothers (17/21) felt that their child was deeply attached to them, whereas only a half of the restored children were described as deeply attached. Adopted adolescents were also more often said by their mothers to be attached to their father than the restored group.

Ex-institutional children had greater problems with siblings than a comparison group.

There were no differences regarding the number of contacts with opposite sex friends, or whether the 16 year-old currently had a boy/girl friend compared to non-institutionalised adolescents.

However, ex-institutional children had poorer relationships with peers than a comparison group. Teachers rated the ex-institutionalised group as more often quarrelsome, less often liked by other children and as bullying other children more than the comparison group.


Hodges and Tizard believed that their findings demonstrate that children who are deprived of close and lasting attachments to adults in their first years of life can make such attachments later, although this does depend on the adults concerned and how much they nurture such attachments.  Hodges and Tizard offer an explanation for why the adopted children were more likely to overcome some of the problems of early institutional upbringing better than the restored children.  The financial situation of the adoptive families was often better, they had on average fewer children to provide for, and the adoptive parents were particularly highly motivated to have a child and to develop a relationship with that child. The biological parents in Hodges and Tizard's sample seemed to have been 'more ambivalent about their child living with them'.



Being a natural experiment this is very high in ecological validity. 

However, being a natural experiment the researchers would have had little control over confounding variables.  For example in this study at the age of four the children were split with some returning to parents and others being adopted whilst seven stayed mostly in care.  It is unlikely that this would have been a random process!  It is most likely that the more personable children with the better social skills would have been fostered.  The ones with the most problems are likely to have remained in care.  As a result it is difficult to be certain that the resulting behaviours at the age of sixteen were down to type of care.  They could have been due to temperament of the child.

Longitudinal studies can suffer from attrition.  Not all participants starting the procedure see it through to the end.  Families move to other areas, no longer want to take part or simply can’t be traced.  In the case of Hodges and Tizard only 51 of the original 65 were questioned at the age of eight.  The ones who are left may not be representative of the initial sample. 

The effects of institutionalisation

Reactive detachment disorder

Ever seen ‘Good Will Hunting?”  Matt Damon plays a character with this condition.  An extreme lack of sensitive responsiveness from a parent in early life can lead to a child growing up unable to trust or love others.  They become isolated and very selfish and unable to understand the needs of others can become sociopathic without a conscience. 

Disinhibited attachment..

A condition in which children select attachment figures indiscriminately and behave in an overly familiar fashion with complete strangers.  It seems to be caused by long periods of institutional care in early life.  They often have other behavioural disorders too including attention seeking. 

Rutter et al (2007) and the Romanian orphans

This is an on-going longitudinal study which began in 1998. 

111 Romanian orphans were adopted into British families.  Rutter wanted to see if good care could compensate for the privation the children had suffered before the overthrow of the Communist dictator Ceaucescu. 

Again this has been run as a natural experiment with age of adoption being the naturally occurring independent variable (IV).  Rutter is studying three groups:

·         Adopted before the age of 6 months

·         Adopted between 6 months and 2 years

·         Adopted after the age of two (late adoptees).

By the age of six years children were making very good recoveries, however, those adopted later (older than two years) had a much higher level of disinhibited attachment.  In 2007 Rutter returned to the children (then aged eleven years) and found that some had made recoveries but about half of those diagnosed with the condition at the age of six still had it at the age of eleven.



Children exposed to privation are more likely to make a fuller recovery if adopted into a caring environment at an earlier age. 


How does day care affect children’s social development

The issue (bless you)

Bowlby believed that short term separation in the first few years of life could produce long term and irreversible negative effects on the child’s social, emotional and cognitive development.  If this really were the case then surely placing children in day care would be damaging for child. 

In recent years the issue of day care has become increasingly important with more mothers of young children working longer and longer hours. In the 1950s it was almost unheard of for mums to work.  By the 1970s 50% of women in the UK between the ages of 25 and 44 were in work or seeking work. By the 1990s this had risen to 75 %.  More importantly 80% of non-working mothers would work if they could find and afford good quality day care.

Recent figures from the USA show that 80% children below the age of six spend an average of 40 hours a week in non-parental care. 

This section of the topic looks at the types of day care available and crucially at the evidence for and against day care being damaging. 

Types of daycare

Day nurseries look after and educate children from a few months old to 5 years old through play opportunities. There are different types of day care including: private, community, workplace or local authority run. All of them will be registered and inspected by OfSTED, to ensure quality care for your child. Some nurseries offer after school and holiday care for children of school age.

Day nurseries are staffed by qualified and experienced people and can offer either part-time or full-time day care.

A childminder is someone (other than a parent, relative, person with parental responsibility, or foster parent) looking after one or more children under 8 years of age, in their own home for over two hours a day, for reward. Under the terms of the 1989 Children Act childminders must register with Ofsted. It is illegal for an unregistered person to look after children for reward

Childminders provide care for babies, toddlers and school age children in the childminders own home. Childminding is a professional career providing consistent high standards of care in partnership with parents.

Playgroups operate in a variety of community venues and provide integrated care and education for children aged 2 to 5 years on a sessional basis during term time. Sessions usually last 2.5 hours to 4 hours and provide a wide range of activities which offer children the opportunity to learn through play. Some groups will offer wraparound care to extend the sessions. Through well-planned play, children will develop skills such as language, listening, concentration, learning to work together and to co-operate with other children. They will also learn skills to help them with reading, writing and numbers. Pre-schools are registered and regularly inspected by OfSTED.

Others include crèches, nannies and play schemes, nursery classes and out of school clubs


The effects of day care on social development.

This is the main thrust of this section and as already stated is based on the work of Bowlby who would have predicted that daycare is bad for the emotional attachment and future development of the child.  In 1994 Violata and Russell reviewed the findings of 88 previous studies (this is called a…….?) and concluded that regular day care for more than 20 hours a week had an ‘unmistakeably negative effect on socio-emotional development, behaviour and attachment of young children.’

In fact the evidence is not clear with some studies agreeing with Violata and Russell and others coming down in favour of day care. In fact it is difficult to come to clear conclusions for all sorts of reasons and mentioning some of these will qualify you for those pesky hard to get AO2 marks! 

Some texts lump social development together as one amalgam, however for the purposes of revision I think it would be easier to do as other texts suggest and break it down into its component parts: effects on; attachment, aggressiveness and peer relations.

Effects on attachment

- (negative)

Belsky and Rovine (1988) found support for Bowlby’s theory of permanent harm.  If the child is in day care for more than four months in their first year they are significantly more likely to develop insecure attachments.  Sroufe (1990) suggested therefore that children should not be placed in daycare until at least the age of one. 

+ (By positive here I mean there are no negative effects)

Clarke-Stewart et al (1994) in a study of 500 children found that children receiving up to 30 hours a week of day care were no more distressed than other children who had attended much lower intensity day care when separated from parents in the strange situation.  Roggman et al (1994) compared infants who had attended day care in the first year with those who had remained at home and found no difference in attachment with mothers.

These studies suggest that day care has no ill effects on attachment or social development. 

In a recent study Erel et al compared day care-attenders with home-reared on six measures including type of attachment with mother, self-esteem and interaction with peers.  Again they found no significant differences between the two groups suggesting no negative (or positive) effects from attending day care. 

Effects on peer relations

Common sense would suggest that early opportunities to mix with lots of other children of similar age would help with peer relationships and most evidence seems to support this idea:


Shea (1981) videotaped children in a day nursery and compared the behaviours of those attending for different lengths of time.  (Interesting for two reasons: clear ethical issues with taping young children playing and also the very early days of video!).

He found that children who attended more regularly were more active, more sociable in that they went looking for people to talk to, and made more contact with others.  This increase was greatest in those attending day care most often (every week day) suggesting a correlation between time spent in day care and sociability. 

Clarke-Stewart et al (1994), already covered, found that increased time in day care seemed to speed up social development, so children who had experienced more day care learned their social skills at an earlier age.


There is no direct evidence suggesting that day care harms peer relations.  However, Sroufe et al (2005) in their Minnesota longitudinal study did find support for Bowlby’s continuity hypothesis.  This would suggest that children forming secure attachments with parents were more likely to form close relations with others later in life.  Since day care may increase the likelihood of insecure attachments it seems logical to assume that day care may have a negative impact on relations with peers later in life.



Effects on aggression



Vandell and Corasaniti (1990) found that eight year olds who had spent their early years in day care were rated as more ‘non-compliant’ by both their teachers and their parents.  A number of studies e.g. Belsky (1999) have tended to support this finding that long periods of day care in the first five years can lead to raised levels of aggressive behaviour in later childhood.  Haskins (1985) found that children kept in larger groups were more likely to be aggressive



However, as always, things aren’t quite so clear cut.  Borge et al (2004) carried out a questionnaire study of over 3000 Canadian children comparing day care children with children reared at home.  Mothers were asked to rate their lovelies on measures such as frequency of hitting and biting other children and how they react when accidentally hurt by others.  These researchers found that children kept at home appeared to be more aggressive.  However, the results are difficult to interpret since those being kept at home were more likely to be from disadvantaged backgrounds.  As a result it is difficult to know whether the day care is preventing aggression or poverty is causing it!

Clarke-Stewart (again) argue that much of the research into aggression (e.g. Vandell and Corasaniti) fail to distinguish non-compliance from assertiveness from aggression.  What is being reported as more aggressive behaviour in the daycare children could simply be children that have greater confidence and have learned to assert themselves better.


Comparing studies and drawing conclusions

This is the AO2 stuff!

The studies covered above have come to a variety of sometimes complimentary and sometimes conflicting conclusions.  However, when we look at the studies we are not comparing like with like for all sorts of reasons:

The types of daycare used differ between studies, sometimes looking at childminders and sometimes nurseries etc.  Because of this the adult-child ratios and number of peers present will vary between studies and will impact on the child’s behaviour.

Quality of daycare is a crucial variable, sometimes being provided by qualified and experienced professionals and unfortunately sometimes not.  Much research has shown that only high quality daycare will be beneficial to the child.  Children who get good quality daycare tend to come from well-educated middle class backgrounds (Marshall 2004).

Age of commencing daycare and time spent in daycare also vary between studies with some looking at children who started in their first year and are spending 40 hours a week in daycare whilst others start later and spend perhaps only a couple of days a week away from parents.

The personality (temperament) of the child is hardly ever mentioned in the studies but is clearly going to affect how the child reacts to separation.

Family backgrounds for those attending and not attending daycare are likely to be very different.  Wealthier middle class children are more likely to spend longer time in good quality care whilst children from poorer and one-parent families are more likely to remain at home.  Therefore as with the Borge study, comparisons between daycare and home reared are not comparing like with like. 


But there is some good news!

The NICHD (National Institute of Child Health and Human Development) are carrying out a HOOOOOGE longitudinal study of the effects of daycare and hopefully taking many of these variables into account. 

The NICHD study began with over 1000 children from a variety of backgrounds in 1991 and being a longitudinal study is ongoing with results still coming in. 

Children aged five:

More time in daycare resulted in more assertive, non-compliant and aggressive children.

Children in full-time daycare were three times more likely to show behavioural problems such as tantrums, lying and arguing.

Higher quality daycare was associated with higher levels of cognitive development (reading etc).

The sensitivity of the mother was a much better indicator of social competence and aggression than time spent in daycare.  More sensitive mothers were far more likely to be producing children with fewer social problems. 

The latest findings from the study (reported by Jay Belsky 2007) with the children aged twelve years, still found a link between extended daycare and aggressiveness. However, latest findings by Belsky also reinforce earlier conclusions that the home environment is far more important. 


Campbell, Lamb and Hwang (2000)

Another longitudinal study, this time of Swedish children, some of whom attended daycare and some who’s parents had requested daycare but been denied. 

At the outset (children aged 18 months) before starting daycare, the children were assessed on a variety of measures including quality of parenting, play behaviour, home environment.  They were then observed after they had just started daycare.  These measures provided a baseline for future observations.

At the ages of 21/2 and 31/2 these observations were repeated.

At 6 their carers were asked to judge the child’s social competence.

At 8 their teachers were asked to judge their social behaviour

At 15 they were given questionnaires to assess their social style, development and friendship quality.


Children who had spent long days in daycare were found to be less socially competent

Conversely, those who spent shorter periods but many times a week in daycare were found to be more socially competent.

Quality of daycare was vital.  Those attending good quality daycare before the age of three had increased social competence.

Crucially they found a high correlation between social competence aged 3 and aged 15.  This suggests that social abilities are formed at an early age and once developed are unlikely to change in later life.  We call these stable characteristics. 


Good quality early daycare is essential for the development of social skills.


Individual needs of the child

Overall evaluation

All the above studies are suggesting that early daycare is related in some way to later social development (either in a positive or negative way).  However, all the studies are correlational so we cannot assume a causal link.  As an example: children who attend daycare become more socially competent and outgoing.  We therefore assume daycare is causing extrovert behaviour in later life. 

 However, it could be that shy children have shy mothers (perhaps introversion is inherited) who prefer not to work but stay at home and care for their children.  As a result only outgoing children attend daycare in the first place!

Egeland & Heister (1995) studied 70 children from poor backgrounds.  Half entered day care before the age of one, the other half stayed at home with mothers.  The Strange situationess the se attachment of the childrn.  They were assessed at the age of one and 3 1/2. 

Findings:  Day care seemed to have negative effects for securely attachedildren.  It had a positive effect for insecure children. 

Conclusion:  Insecure children needed the extra attention so benefited from day care.  Securely attached children did not need the extra attention. 


Implications of research into attachments and day care

In the UK in 2004 the Government announced the National Childcare Strategy offering free state nursery to all three and four year olds and providing guidelines for what they should be taught!  Much of this and other government and state initiatives in the past 50 years has stemmed from psychological research into the importance of attachments and what constitute good day care.

 Children’s hospital

If as a child you were unfortunate enough to be hospitalised, even as late as the 1960s, there would be strict rules applied to parental visiting times and limits placed on our long parents could spend with their children.  Following the findings of Robertson and Robertson’s study that children require continuing emotional care and as much contact as possible with natural parents, visiting hours were extended.  Today hospitals sometimes allow around the clock visiting hours. 

Deferred adoption?

Mothers who were intent on having their children adopted used to still be encouraged to nurse them in the first few weeks and months of life before they were adopted.  This would result in either a broken bond when they were moved or in no attachment being made at all because of the lack of sensitive responsiveness on the part of the unwilling mother.  Today, thanks to research into attachment behaviour, adoption usually occurs in the first few weeks of life. 

Teaching mother to suck eggs

Some psychologists have suggested that knowledge of attachments can be used to improve parenting skills.  For example:

Juffer et al (1997) conducted a study aimed at seeing whether adoptive parents could be taught sensitive responsiveness.



Ninety families were split into three groups of thirty.  Thirty received no training, thirty received training via a self-help booklet and thirty received training via being filmed interacting with children and then watching the footage with an expert who advised on better techniques. 

Findings and conclusions

The first two groups control and training via a booklet showed no differences in responsiveness to the general population, whereas those trained by human intervention were significantly improved which crucially led to a more secure bond being formed.  The researchers concluded that not only could sensitive responsiveness be taught but it did indeed improve quality of attachment.

Improving day care

 Small groups

High staff to child ratio

Well trained staff

Sensitive staff

Lots of adult-child interaction

Low staff turnover

Consistency of care

Lots of stimulating toys

Mixed age groups





 Improving day care

The most researched area of policy based on attachment research however has been into ways of improving day care.  What constitutes good care isn’t easy to define (Schaffer) but the following variables seem important:

1.       Plenty of verbal interaction, especially between the child and carer which seems to result in higher quality day care.  However, it is important to mention that verbal interaction between parent and child tend to be more complex and beneficial simply because they can provide undivided attention to their children rather than have to divide it between man y.

2.       Sensitive responsiveness of the carers is, according to the NICHD, the single most important factor.  In one survey they discovered that only 23% of carers provided ‘highly sensitive’ care with 20% being ‘emotionally detached’ from the children in their care.

3.       Consistency of care is also important.  Low staff turnover provides opportunities for children to form attachments with the care givers whereas, higher staff turnover can result either in failure to form an attachment or in broken attachments if a regular care-provider leaves. 

4.       Staff should be well qualified in caring for children.  Sylva et al (2003) found it was particularly important for the day care centre manager to be wee trained. 

5.       Child/staff ratios should be kept low.  The NICHD recommend no higher than 3 children to each carer.  Similarly children should be kept in small groups so fewer strangers need to be dealt with.

6.       Groups of mixed ages are also good for the child’s social development since, according to Clarke-Stewart et al (1994) younger children get the opportunity to watch and learn from the social interactions of the older children.

7.       Similarly plenty of educational toys will help improve the child’s cognitive development.  We will look at this briefly in year 13 when we consider the environmental factors of intelligence.


Child/day care fit

Some children will benefit more from day care than others.  Similarly the type of day care required will vary from child to child, depending on their temperament and upon the home environment.  Parents should therefore, whenever possible try to find the appropriate day care for their child.

For example, more aggressive children may be better off placed in day care homes or receive at home care rather than be placed in larger day care centres. 

However, children from very poor backgrounds seem to be less aggressive when in larger groups than when kept at home.  Perhaps for these larger day care centres may be beneficial. 

Children who attend group day care tend to develop better social skills since they are mixing with larger groups of peers.  An only-child may benefit more from this opportunity than a child from a larger family who has lots of siblings to play with. 


Finally… a thought for the mothers

It isn’t just the child who may suffer the effects of day care but also the parents and particularly the mother.  Mothers who return to full-time employment within a year of giving birth often feel guilt and anxiety, worrying that their child may become less securely attached.  However, by staying at home they run the risk (Brown and Harris 1978) of becoming depressed, presumably because of the social isolation they experience. 

The provision of workplace nurseries may help reduce the worst effects providing opportunities for mothers to spend more time with their kiddies whilst at work.