Unit 1: attachment
A two-way, enduring, emotional bond between two individuals, in which the following behaviours are displayed.
1. Seeking proximity, especially at times of stress. 2. Distress on separation and pleasure when reunited 3. Secure
base behaviour-so infants are happy to explore but regularly return to attachment figure.
What the specification says you have to know:










Caregiver-infant interactions in humans: reciprocity and interactional synchrony.
Stages of attachment identified by Schaffer.
Multiple attachments and the role of the father.
Animal studies of attachment: Lorenz and Harlow.
Explanations of attachment: learning theory and Bowlby’s monotropic theory. The
concepts of a critical period and an internal working model.
Ainsworth’s ‘Strange Situation’. Types of attachment: secure, insecure-avoidant and
insecure-resistant.
Cultural variations in attachment, including van Ijzendoorn.
Bowlby’s theory of maternal deprivation.
Romanian orphan studies: effects of institutionalisation.
The influence of early attachment on childhood and adult relationships, including the
role of an internal working model.
1
Caregiver-infant interactions in humans
Reciprocity
The word reciprocal means two-way, or something that is mutual. Infant and caregiver are both active
contributors in the interaction and are responding to each other. This is referred to as reciprocity.
Traditional views of childhood believed that a baby took a passive role, simply receiving care from an adult
however research shows us that a baby is very active and that mother child interactions are like a “dance”
Brazelton et al (1975). Mother and child take turn to initiate interactions and from birth babies and carers
spend a lot of time in intense and pleasurable interaction. Babies have alert periods that mothers pick up
on and respond to, although only about two thirds of the time (Feldman and Eidelman 2007). From three
months the reciprocal element kicks in as interaction becomes increasingly frequent and involves the carer
and child playing close attention to each other’s verbal signals and facial expressions.
Interactional synchrony
The word synchrony means a simultaneous action or occurrence.
Interactional synchrony relates to the timing and pattern of the
interaction. The interaction is rhythmic and can include infant and
caregiver mirroring each other’s behaviour and emotion. The infant and caregiver’s behaviours are
synchronised because they are moving in the same, or a similar, pattern.
Condon and Sander (1974) have investigated interactions between infants and caregivers in particular in
relation to responses to adult speech. In their paper they report “As early as the first day of life, the human
neonate moves in precise and sustained segments of movement that are synchronous with the articulated
structure of adult speech”. According to research by Meltzoff and Moore (1983) infants as young as 3 days
imitate the facial expression of adults. This implies that this ability to mirror is an innate behaviour.
Exam tip!
The AQA acknowledge that the textbooks tend to offer slightly different definitions and that the
two concepts do overlap. They say that candidates do need to be aware that the concepts overlap
but that they should also be able to define the terms separately and mention Condon and Sander
as a good example.
So how exactly do they overlap?
The concepts in this area are describing the metaphorical dance between infant and caregiver. It
looks, feels and acts like a conversation between infant and caregiver. Though no words may be
uttered (not on the baby’s part anyway) this two-way interaction has two active contributors (carer
and baby). The interactions will be rhythmic and mutual; infant and caregiver are likely to appear to
be in harmony as they take turns in this attachment promoting conversation which is likely to
involve the infant and caregiver responding to each other with similar sounds, emotions and
behaviours. Infant and caregiver are able to anticipate how each other will behave and can elicit a
particular response from the other.
What does this overlap look like?
The caregiver who smiles at their baby in response to a smile-like facial movement from the infant or who opens
their mouth in mock surprise when the infant scrunches up their face, is engaged in this two-way pattern of
interaction and is laying the foundations for an attachment to form. The infant that cries and
elicits a sad expression and look of upset on their caregivers face or the caregiver who laughs in
response to their infants giggling sound and tickles them, is experiencing synchronised
interaction.
2
Evaluation of caregiver-infant interactions
Supporting evidence
Both of these pieces of research show support for the concepts of reciprocity and
interactional synchronicity.
Meltzoff and Moore (1977) used 4 different stimuli; adults displayed one of three facial expressions and a
hand gesture and the child’s response was filmed and identified by independent observers who had no
knowledge of what the children had just seen. The behavioural catergories were mouth opening,
termination of mouth opening, tongue protrusion, termination of tongue protrusion. Each observer scored
the tapes twice so that inter-observer reliability could be calculated. All scores were greater than .92. An
association was found between the expression or gesture the adult has displayed and the babies actions.
Isabella et al (1989) observed 30 mothers and infants together and looked at synchrony and quality of
attachment and found an association between high levels of synchrony and better quality attachments.
This research supports because Meltzoff showed the existence of reciprocity consistently between
mother and child. Also Isabella found differences in the degree of reciprocity and intentional synchrony
shown i.e. infants displaying more synchronicity were more strongly attached which shows us not just
about the concepts existence but also about the purpose and importance of reciprocity and interactional
synchrony.
Other evaluation
Research
methodsobservations
Research
methodsobservations
Not universal
Practical applications
Practical applications
Observations of mother-infant interactions (such as above) are generally well
controlled procedures, with both mother and child being filmed, often from
multiple angles. This ensures very fine details of behaviour can be recorded and
later analysed and improves the validity of the research and the use of
independent observers improves the reliabilty of the research.
Infants don’t know or care that they are being observed so their behaviour
doesn’t change in a controlled observation which is generally a problem for
observational research. This is a strength as means the research has good
validity. However the behaviour of the mother may be subject to demand
characteristics.
- Interactional synchrony is not found in all cultures which weakens the support
for the idea that it is important in the development of attachments. Le vine
(1994) reported that Kenyan mothers have little physical contact or interactions
with their children but have high proportions of secure attachments.
Meltzoff and Moore (1983) demonstrated intentional synchrony with three day
old babies. Which suggest that babies should be placed next to their mothers in
hospital and should continue to remain in close contact for at least the first
three months to improve the quality of attachment.
Music therapists and other therapists who work with parents and
infants/children who have experienced disruption in their attachment
formation (e.g. when children are adopted and attachments with foster carers
need to be transferred to adoptive parents) recognise the importance of
reciprocity and interactional synchrony in their therapeutic work. For example,
caregivers are encouraged to mirror and share their child’s emotion by
imitating facial expressions and engaging in reciprocal behaviour in response to
music or other stimuli.
3
Stages of attachment identified by Schaffer
The 4 stages outlined by Schaffer and Emerson are outlined below:
Stage
Preattachment
Age
Birth-3 months
Indiscriminate 3 months to 7/8
attachment
months
Discriminate
attachment
Usually 7/8
months onward
Multiple
Attachments
9 months onwards
Description
Babies start to smile and become more “sociable”
and from around 6 weeks. They can tell people apart
and like to be in human company so begin to form
stronger attachments however these do not
progress much until the next stage as can be easily
comforted by any individual. At this stage, Schaffer
and Emerson found that they did not show a fear of
strangers.
Infants are recognising and forming bonds with their
carers through reciprocity and interactional
synchrony however their behaviour towards animate
(faces) or inanimate objects (teddies) is quite similar.
Towards the end of the phase they start to be more
content when in the presence of other people and
can be more easily calmed by familiar adults but will
allow strangers to handle and look after them.
The key things about this stage are that the infant
begins to show separation anxiety and “protests”,
usually by crying, when their primary attachment
figure leaves (the biological mother in 75% of cases)
They are said to now have formed a specific
attachment. The second key behaviour is that they
begin to show fear of strangers.
Shortly after infants show specific attachments
they begin to make multiple attachments (29%
within a month according to Schaffer’s study). This is
usually towards friends, grandparents and childminders/ nursery staff.
4
Evaluation of stages of attachment
Supporting evidence
Research by Schaffer and Emerson supports the stages of attachment
Schaffer and Emerson (1964) aimed to investigate the formation
of early attachments, when they started, their intensity and to
whom they were directed.
Method- A longitudinal study using 60 babies (31 male, 21
female) from Glasgow the majority from skilled working class backgrounds. They were visited at home
every month for the first year and then at 18 months. Mothers were interviewed to measure the infants
level of attachment asking questions about how their infants responded to seven situations e.g. adult
leaving the room (separation anxiety) and observations were conducted to investigate the level of distress
the presence of a stranger caused (stranger anxiety).
Findings- Timings-Specific attachment (signs of separation anxiety) 50% of infants by 7 months, 80% by 40
weeks and almost 30% displayed multiple attachments. By one year 78% had developed multiple
attachments with 33% having five or more multiple attachment figures.
The findings of this study were used to create Schaffer’s stages and so we can say that the stages have
longitudinal evidence supporting it.
Research methods evaluation-Longitudinal
The study looks at the same group of infants over a
much longer period of time than a lab experiment
or observation and so it can look at the real
development of attachment and is not just a snap
shot like most other research.
Schaffer’s study observed babies in their actual
home environment and so the behaviour was
unlikely to have been effected by the observers and
so can be said to have external (ecological) validity.
Schaffer and Emerson’s study is based on a limited
sample of participants. Whilst 60 babies is an
acceptable sample size, all the babies were from the
same area and were the same social class so is low
in population validity.
In some cultures multiple caregivers are the norm
so multiple attachments may develop immediately
instead of specific attachments and as child rearing
practices vary from culture to culture and over time
these results do not necessarily generalise well.
The stages link to reciprocity and interactional
synchrony particularly at the indiscriminate stage
and so further support the stages and the existence
of these two concepts.
Stage theories are inflexible. If a child doesn’t fit
into the particular stage at the “right” time then
they may be judged as abnormal or parents might
become worried if these stages are a standard by
which families are judged.
Validity
Validity
Generalisability
Links and supports other concepts
Stage theories are inflexible
5
The role of the father and multiple attachments.
The role of the father
So how important is the role of fathers in a child’s development?
Traditionally research looks at mother-child interactions and there is
limited research into this area. The research that has been carried
out focus on different areas and so it’s hard to come up with a
definitive answer however research does suggested that the father
plays a key role in the development of attachments.
Summary of the key findings
-Fathers play interactions are more exciting and pleasurable than mothers
-Mothers are more nurturing and affectionate
-Mothers are preferred when children are distressed and seeking comfort
-Fathers are preferred when children are in a positive emotional state and want stimulation (Lamb 1987)
-Fathers are less able than mothers to detect low levels of infant distress (Hrdy 1999)
-However fathers that do become the main care provider do quickly develop more sensitivity to children’s
needs and become a safe base for children to explore from suggesting that sensitive responsiveness is not
necessarily a biological ability limited to women Lamb (1987).
-Marital intimacy was linked to security of father-infant interactions. So the fathers with secure fatherinfant interactions had secure and intimate relationships with the child’s mother. (Belsky 2009)
Evaluation of the role of the father.
Research evidence
Evidence by Grossman (2002) has suggested that the father has a different, unique
role in attachment-one that is more to do with play and stimulation and less to do
with nurturing.
A longitudinal study was carried out looking at parents behaviour and its relationship to the quality of
children’s attachment in their teens. It was the quality of the mother’s attachment and not the fathers that
was related to attachment in their teens but the quality of fathers play with infants was related.
So this evidence shows that the father is still important in a child’s development and their role does
impact on the quality of their attachments in their teenage years but through play and not nurturing and
being the primary attachment figure.
There is evidence that contradicts the idea that fathers can play the same role as mothers successfully.
Field (1978) filmed 4 month old babies in face to face interactions with primary
caregiver mothers, secondary caregiver fathers and primary caregiver fathers.
They found that primary caregiver fathers like mothers spent more time smiling,
imitating and holding infants than the secondary caregiver fathers and that this
behaviour is important in the building of attachments.
This is suggesting the father is as capable as the mother of being a sensitive
and caring primary caregiver if given the chance and that it is the level of responsiveness and not the
gender of the parent that is the key to the attachment relationship.
6
Other evaluation
Contradictory findings
Biology?
Practical applications
Grossman suggests that fathers as secondary attachment figures have an
important role to play in a child’s development however research by Golombok
(2004) have found that children growing up in a single or same sex parent
household do not develop any differently and so questions the importance of the
role of the father in attachment.
Research suggests that a sensitive male carer is as capable of effective child
rearing as a female but why do men not tend to take on this role? It is of course
to do (in part) with traditional gender roles but it could be the fact that female
hormones create higher levels of empathy which biological pre-dispose women
to be the primary attachment figure.
The research can be used to help develop the quality of attachments within the
whole family as it shows that the parents relationships can impact on the fatherchild relationships. It can also be used to try to validate/increase the role that
fathers play in families as it shows how important it is to a child’s development.
Multiple attachments
So what does the evidence show us then about multiple attachments?
-29% of infants develop multiple attachments within a month of developing specific attachments (by 40
weeks)
- By the age of one 78% of children have multiple attachments
- By the age of one 33% of the 78% have 5 or more attachments (Schaffer)
-Fathers are one of the key multiple attachments that children form and that this attachment is important
for a child’s development.
Further evidence and ideas
Do all babies develop a single attachment to the main carer before they develop multiple attachments?
Bowlby developed the idea of monotropy – babies have one key attachment figure. This figure is usually,
but does not have to be, the mother. Secondary attachments then follow, such as to the father and
siblings.
If babies do develop multiple attachments after a primary attachment to one person, are multiple
attachments of equal strength to the primary attachment or not?
Rutter (1995) saw all attachments as being equal – so there is no such thing as primary and secondary attachments.
All a child’s attachments give the child an idea of how relationships work (an internal working model).
Is the picture clear on multiple attachments?
In some cultures multiple caregivers are the norm so multiple attachments may develop immediately
instead of specific attachments. For instance Sagi et al (1994) looked at children raised in a community
children’s home in kibbutzim where they
slept away from their parents, and
compared them to children with family
based sleeping arrangements. Attachment
to the mother was twice as strong in the
family based arrangements.
7
Animal studies of attachment.
In the early 20th century a number of ethologists conducted animal studies looking at attachment in infant
animals to inform their understanding of human mother-infant attachments.
Lorenz (1935)
To investigate the mechanisms of imprinting, where bird species mobile from
birth follow and form attachment to the first moving object they see.
Lorenz split a clutch of greylag goose eggs into two batches- one naturally hatched
by the mother, one in an incubator with Lorenz as the first moving object they
saw. He then recorded their behaviour.
He also marked the gooslings and placed them under an upturned box, then removed it and recorded their
behaviour.
Findings
The incubator group followed Lorenz everywhere whereas the control group followed their mother. When
the two groups were mixed up the incubator group still continued to follow Lorenz and the control group
the mother.
There was a critical period of between 4-25 hours (depending on species) and if imprinting did not occur
the chicks did not attach to a mother figure.
Lorenz subsequently reported that the goslings imprinted on humans would later attempt to mate with
humans.
Harlow (1959)
To determine whether food or close comfort was the important factor in attachment,
Harlow placed 16 baby rhesus monkeys in cages with two surrogate mothers; a harsh
wire mother or a soft towelling mother.
4 of the 16 monkeys were used in each of the following four conditions-wire mother producing milk, towel mother no milk
-wire mother no milk, towelling mother producing milk
-wire mother producing milk
-towelling mother producing milk
The amount of time spent with each mother was recorded as well as feeding time.
The monkeys were frightened with a loud noise to test for mother preference during stress.
A large cage was also used to test for degree of exploration.
Findings
Monkeys preferred contact with the towelling mother when given a choice of
surrogate mothers, regardless of whether she produced milk; they even stretched
across to the wire mother to feed whilst clinging to the towelling mother.
Monkeys with only the wire mother showed signs of stress as had diarrhoea.
When frightened by a loud noise, monkeys clung to the towelling mother in conditions where she was
available.
8
In larger cage conditions monkeys with towelling mothers explored more and visiting their surrogate
mothers more often.
Monkeys in adulthood
Monkeys in some of Harlow’s experiments were followed into adulthood and severe consequences were
found- more aggressive, less sociable, bred less often as were unskilled at mating. As mothers some of the
monkeys neglected their young and others attacked their children even killing them in some cases.
Evaluation of animal studies
Theoretical value
Difference in nature and complexity of
the bond (taken from mark scheme)
Problems of extrapolation to attachment
in human infants (taken form mark
scheme)
Ethical issues
Imprinting not
always
permanent -
Practical applications
Harlow’s findings have had a profound effect on psychologists
understanding of mother-infant attachment. Most
importantly Harlow showed that attachment does not
develop as a result of being fed by a mother figure but s a
result of contact comfort. He also showed us the importance
of the quality of early relationships for later social
development.
It is argued that it is not appropriate to generalise Lorenz’s
findings to humans as a mammal’s attachment system is quite
different to that of birds. Mammals show more emotional
attachment to their young and may be able to form
attachments at any time, albeit less easily in infants so the
bond in humans is much more complex.
Although Monkeys are more similar to humans than geese so
making it slightly more easy to generalise, they are still not
human and so it is argued that we can not necessarily
generalise Harlow’s or Lorenz’s research to human
attachments.
Harlow faced severe criticism for the ethics of his research.
The species were considered similar enough to humans to be
generalised to them and so the suffering they encountered
was presumably very human like there were sever long term
effects for all of the monkeys involved. Some argue the
findings were important enough to justify the ethical issues
though.
Lorenz work has been replicated and in some cases it was
found that imprinting was not always permanent. Guiton
(1966) found that when chickens were imprinted to a yellow
rubber glove they did at first imprint on it but learned with
experience to prefer mating with other chickens eventually so
it suggests that the impact of imprinting on mating behaviour
is not as permanent as Lorenz believed.
Harlow’s research has helped social workers to understand
risk factors in child neglect and abuse and so intervene to
prevent it (Howe 1998). They have also been important in the
care of captive monkeys and we now understand the
importance of proper attachment figures for baby monkeys in
zoo’s and breeding programmes.
9
Explanations of attachment: Learning theory and Bowlby’s monotropic theory.
Learning theory
The Behavioural explanation proposes that all behaviour is learnt rather than
inborn (innate). When children are born they are like blank slates and everything
they become can be explained in terms of the experiences they have. Learning
theory is put forward by behaviourist who focus solely on behaviour- what
people do rather than what they might be thinking. Behaviourists suggest all
behaviour and so attachment can be explained using the concepts of classical
conditioning and operant conditioning.
Classical conditioning
According to the classical conditioning part of learning theory attachment is formed partly due to learning
through association. Use the diagram to help you to understand the concept explained below.
Food (Unconditioned stimulus) naturally produces a sense of pleasure in a child (unconditioned response).
The person who feeds the infant initially (neutral stimulus) provides no natural response but over time the
“feeder” eventually produces the pleasure associated with the food; pleasure now becomes a conditioned
response and the feeder the conditioned stimulus. This association between an individual and a sense of
pleasure is the attachment bond.
unconditioned stimulus
unconditioned response
Neutral stimulus
No response
After enough pairings of the previous stage:
conditioned stimulus
conditioned response
10
Operant conditioning
Key terms
Positive reinforcement-when the consequences of your actions are pleasant, so you repeat the behaviour to again
get the positive reinforcer e.g. praise, money.
Negative reinforcement- when you repeat a behaviour not because you are receiving a reward or something
pleasant but are removing something unpleasant for example you take paracetamol to get rid of your headache
which is unpleasant.
Primary reinforcer-this is something which is automatically or directly reinforcing e.g. food, money, praise.
Secondary reinforcer-this is something that is indirectly reinforcing so through classical conditioning has become
associated with a primary reinforcer e.g. the mother giving a child food.
Dollard and Miller (1950).
Operant conditioning is learning through reinforcement of behaviour. Crying is an automatic response to
hunger. If an infant is fed each time she cries, she learns that crying results in food, which is rewarding, and
so repeats this behaviour more often - this is positive reinforcement. The caregiver’s attention results in
food, and so is also reinforcing. The primary reinforcer in this process is the food. The secondary reinforcer
is the caregiver. The infant learns that crying will maintain the caregiver’s attention and food supply, so
attachment behaviours such as separation distress are formed. The caregiver is also conditioned by the
infant. The caregiver feeds the infant, who stops crying. Feeding is repeated to avoid the discomfort of
hearing the infant cry. This is negative reinforcement for the caregiver.
Learning theory and drive reduction
The primary drive for the baby is hunger; babies are driven to have their hunger reduced. Attachment is a
secondary drive learned by an association between the caregiver and the satisfaction of the primary drive.
Hence learning theory predicts that babies want to be close to the caregiver who feeds them.
Evaluation of Learning theory
Plausible and scientific (from mark scheme)
Plausible and scientific as founded in an established
scientific theory, i.e. likely that association between
the provision of needs and the person providing
those needs can lead to strong attachments;
reinforcers are clearly shown.
Reductionist (from mark scheme)
The focus on basic processes (stimulus -response
links, reinforcement) is too simplistic to explain
complex attachment behaviours.
Counter evidence
Some animal studies have shown us that young
animals do not necessarily attach to those who feed
it. Lorenz’s geese imprinted before were fed and
maintained these attachments regardless of who
fed them and Harlow’s monkey study showed the
monkeys preferred the comfort monkey who gave
no food.
Counter evidence
Human research by Schaffer also shows that food is
11
not the key factor in attachment as they found that
more than half of infants were not attached to the
person who provided the primary care (i.e. fed
them) but instead were still attached to their
biological mother rather than this caregiver and so
there is no unconditioned stimulus involved.
Ignores other factors
It ignores reciprocity, interactional synchrony and
other factors associated with forming attachments.
Research (specifically Ainsworth) suggests that it is
the sensitivity of the caregiver that is important and
that the quality of the attachment is linked to
developing reciprocity and interactional synchrony.
If attachment was just purely about food then there
would be no need for these complex interactions.
You can compare with other explanations such a
Bowlby to highlight flaws and weaknesses.
Explanations of attachment: Bowlby’s monotropic theory.
Based on the work of Lorenz and Harlow, Bowlby proposed an evolutionary explanation of attachment.
Attachment is an innate system that is biological programmed into babies from birth in order to help them
to survive i.e Infants have an innate drive to survive.
Babies seek proximity to carer (mother) for safety as it protects them from hazards; millions of years ago
wild animals; today, cars, ponds, electricity. So security and secure attachment equals survival.
Monotropy- Bowlby proposed that infants have one special emotional bond, normally the biological
mother (but not always) and that this relationship was different and more important than any others.
Bowlby believed the more time spent with the primary attachment figure the better.
Internal working model-The importance of monotropy is that, for a child, this special relationship forms a
mental representation or a model for what relationships are like. It can therefore have a powerful effect on
the nature of a child’s future relationship and their ability to be a parent themselves. Individuals who are
strongly attached as children continue to be socially and emotionally competent in relationships and with
their own children whereas infants with poor attachments have more social and emotional difficulties in
childhood and in adulthood (the continuity hypothesis).
Social releasers are important for the development of this parent-infant attachment as they elicit
caregiving from the parent. Social releasers are innate mechanisms so natural characteristics or behaviours
of babies such as;
Baby faces/cuteness
Crying
Smiling
Clinging “pick me up!”
12
Critical period-Bowlby believed that if attachment hadn’t occurred by the age of two then a child will have
difficulty forming attachments later on in life.
Exam Tip!
To help you remember Bowlby’s theory use- CR.I.I.M.P.S
Cr. I .I . M. P .S
Critical period
Innate programming
Internal working model
Monotropy
Proximity
Social releasers
Evaluation of Monotropic theory
Research Evidence
Evidence from Harlow and Harlow (1962) supports Bowlby’s theory. They found that baby monkeys would
cling to a wire model covered with cloth rather than a wire model that could feed it. The monkeys spent
most of their time (22/24 hours) clinging to the towelling model, they went on to be bad mothers.
These findings support the concepts of innate programming and Monotropy because the monkey
instinctively sought to be close to one specific mother monkey even if it was a fake one and the fact if
favoured the cloth monkey (22/24 hours spent with it) further supports the idea of monotropy. It also
supports the idea of the internal working model because this monkey didn’t form adequate attachments
and so didn’t form relationships as an infant monkey and went onto become a bad mother, suggesting that
the monkey didn’t know how to look after infants because she had never learnt herself.
Evidence from Schaffer and Emerson (1964) which found that by 18 months old
only 13% of the babies were attached to one person and many of the infants had
as many as five attachment figures contradicts the idea of monotropy. Also Lamb
(1982) found that infants had different attachments for different purposes rather
than certain attachments being more important than others; fathers for play,
mothers for comfort for example. (Not grounded)
13
Hazan and Shaver (1987) discovered that infants who had been securely attached when children went on
to have happy, lasting and trusting relationships as adults, yet insecurely attached infants had less
successful adult relationships. This evidence supports the internal working model. (not fully grounded)
Bailey (2007) also supports the idea of the internal working model being passed through families as 99
mothers with one year old babies who when interviewed reported having poor attachments to their own
parents were more likely to be classified as having a poor attachment with their own babies during
observations. (not grounded)
See last section for further evaluation on the internal role model.
General evaluation
Economic implications/socially sensitive.
The concept of monotropy also included in Bowlby’s 1950s World health
organisation report suggested that babies needed constant care of the mother for
healthy social development which led to stay at home mothering and still has a
negative impact on mothers today. There is still the idea amongst some that mothers
should not work. Erica Burman (1994) and other feminists state that it places a
terrible burden of responsibility on mothers pushing them into particular lifestyle choices and placing the
blame on them if anything goes wrong in a child’s life.
Further evidence however shows that good substitute care either in a nursery, the father or a family
member does not have a detrimental effect on social development and so a mother can happily return to
work after having a child and remain economically active contradicting Bowlby’s theory.
Role of the father
The concept of monotropy suggests that the father can’t play the role of the primary caregiver as Bowlby
suggests it’s the mother or a female substitute which contradicts some research that says otherwise (see
role of father section)
14
Ainsworth’s ‘Strange Situation’. Types of attachment: secure, insecureavoidant and insecure-resistant.
Types of attachment
Secure attachment- This is a strong and contented attachment of an infant to his/her caregiver, which
develops as a result of sensitive responding by the caregiver to the infant’s needs. Securely attached
infants are comfortable with social interaction and intimacy. Secure attachment is related to subsequent
healthy cognitive and emotional development.
Insecure attachment- This is a form of attachment between infant and caregiver that develops as a result
of a caregiver’s lack of sensitive responding to the infant’s needs.
It may be associated with poor subsequent cognitive and
emotional development.
There are two types of insecure attachment; avoidant and
resistant (ambivalent).
Insecure-avoidant- A type of attachment that describes children
who tend to avoid social interaction and intimacy with others.
Insecure-resistant- A type of attachment that describes those infants that both seek and reject intimacy
and social interaction.
Ainsworth and Bell's (1970) The Strange Situation
Aims
To produce a method for assessing the security of an infant’s attachment by placing the infant in a mildly
stressful situation and observing the attachment behaviours that result.
Procedures
The strange situation took place in a laboratory. The original participants were American infants aged
between 12 and 18 months and their caregiver (usually their mother). The third individual involved in the
strange situation was a stranger. The same stranger was used all the time.
The procedure lasted for just over 20 minutes and the behaviour of the infants was closely observed to
assess the infant’s levels of exploring, playing and distress behaviours at separation and reunion with the
caregiver and when left with the stranger.
The eight stages of the strange situation are as follows:
Stage
1 (30 seconds)
infant
People in room
Caregiver, infant, researcher
Procedure
Researcher brings
and caregiver into room then
leaves.
2 (3 minutes)
Caregiver and infant
Caregiver sits, infant free to
explore room.
3 (3 minutes)
Caregiver, infant, stranger
Stranger comes into room, after
a while talks to the caregiver and
then to the infant. Caregiver leaves.
4 (3 minutes)
Infant and stranger
Stranger keeps trying to talk to
15
and play with the infant.
5 (3 minutes)
Infant and caregiver
Stranger leaves as caregiver
returns. At the end of this stage
the caregiver leaves.
6 (3 minutes)
Infant
Infant alone in room.
7 (3 minutes)
Infant and stranger
Stranger returns and tries to
interact with the infant.
8 (3 minutes)
Caregiver and infant
Caregiver returns and interacts
with the infant, stranger leaves.
If the infant became severely distressed in a particular stage, that stage was
shortened.
Findings
Below are the behaviours observed in three types of attachment; secure,
insecure-avoidant and insecure-resistant.
Secure Attachment (66%)
Harmonious and cooperative
relationship
High willingness to explore (using
caregiver as safe base)
Insecure-avoidant (22%)
Insecure-resistant (12%)
Avoid social interaction and
intimacy with others and treat
caregivers and strangers similarly
Both seek and reject intimacy and
social interaction
Low willingness to explore
High willingness to explore
(independently from caregiver)
High stranger anxiety
High stranger anxiety
Low stranger anxiety
Enthusiastic on reunion with
caregiver
Some separation anxiety but
maybe soothed
Indifferent/little/no separation
anxiety
Avoids contact on reunion with
caregiver
Very distressed on separation
from caregiver (high separation
anxiety)
Seeks and rejects reunion from
caregiver (angrily resisting being
picked up but seek proximity in
different ways)
Conclusion
The Strange Situation is a controlled way of measuring individual differences in attachment behaviour, and
these can be categorised into three broad types – secure (A), insecure avoidant (B) and insecure resistant
(C). Secure attachment is likely to be the most common and preferred type of attachment in North
America; Ainsworth and Bell argued that it linked to later healthy emotional and social development. They
also said that there was an association between the mothers’ behaviour and the infant’s attachment.
16
Evaluation of Ainsworth strange situations
General research methods-control, demand
characteristics, validity, sample
Remember to link it to the study generic
evaluations get limited marks!
Inter-rater reliability
Real world application
This study is a laboratory experiment using your research
pack you should be able to explain strengths and
weaknesses of this study. These can include; research in
controlled, reduced demand characteristics with the
babies, possible demand characteristics with mothers, and
investigator effects, internal validity, low ecological validity,
biased sample.
The study has inter-rater reliability. Different observers
watching the same children tend to agree on what
attachment type to classify infants as. Bick (2012) found it
to be as high as 94% which means we can be confident that
the attachment type of an infant observed by the strange
situation does not just depend on who is observing them
i.e. it is reliable.
The circle of security project teaches caregivers to better
understand their infant’s distress signals and to increase
understanding of what it is to be anxious. The project
showed an increase in infants classified as securely
attached from 32% to 40% which supports the research on
attachment types because such research can be used to
improve children’s lives which is a strength.
Depends on parent
Main and Weston (1981) found that children acted
differently depending on which parent they were with so
may have appeared insecurely attached in the study to
their mother but may well be attached to the father
illustrating the attachment types are linked to individual
relationships with carers and are not set characteristics of
children.
Culture bound
There is the belief that the study is culture bound, i.e. does
not have the same meaning outside of the USA and
Western Europe because children and caregivers may
respond differently to the strange situations depending on
their cultural experiences. Takahasi (1990) found that the
test does not work on Japanese children because Japanese
mothers are so rarely separated from their children that
they show very high levels of separation anxiety and in
observations Japanese mothers tended to race to their
children and scoop them up at the reunion stage meaning
the response was hard to observe. See next section on
cultural variations.
17
Cultural variations in attachment.
Culture is the whole way of living your life – your religion, your
language, your beliefs
A cross-cultural study is one that compares different cultures
Van Ijzendoorn and Kroonenberg (1988) cultural variations in attachment styles
Aims
To investigate variations in attachment styles between different cultures, using the results of studies
carried out in various countries. These studies always used the strange situation; hence this study is a
meta-analysis of those ones.
Procedures
The results of 32 studies that used the strange situation to measure attachment behaviour were analysed.
Research from 8 different countries was used, including Western cultures (e.g. USA, UK, Germany) and
non-Western cultures (e.g. Japan, China, Israel). The 32 studies yielded results for 1,990 children.
Findings
1.Secure attachment was the most common attachment style in all nations.
2. In Western cultures the dominant style of insecure attachment found was avoidant
3.in non-Western cultures the dominant style of insecure attachment was resistant
4. There was greater variation within cultures than between cultures.
Table to show percentages of children displaying attachment types in the 8 countries
Country
No. of
Percentage of each attachment type
studies
Securely
Avoidant
Resistant
attached
West Germany
3
57
35
8
UK
1
75
22
3
Netherlands
4
67
26
7
Sweden
1
74
22
4
Israel
2
64
7
29
Japan
2
68
5
27
China
1
50
25
25
USA
18
65
21
14
Conclusions
As secure attachment was the most common style of attachment, there may be
universal characteristics in infant-caregiver interactions. The variations in
attachment both between and within cultures may also show that child rearing
practice varies between and within cultures.
18
Evaluation of Cultural variations in attachment.
Issues with Cross-Cultural comparisons and research such as Van Ijzendoorn
1) Cultural Bias: There may be differences in how different cultures raise their children. It is possible
that behaviours that are valued in collectivist cultures differ from behaviours that are valued in
individualistic cultures. The Strange Situation is based on ideas about attachment from a Eurocentric
viewpoint, so people in the UK and USA may not value over-dependency on the caregiver (insecureresistant attachments), whereas this is valued in other cultures, such as Japan, thus incorrectly
interpreting a behaviour as insecure, when it is secure.
2) Culture does not mean country, and each country will have a number of different sub-cultures. We
must not conclude that all people from one country will act in a similar way. The few studies using the
Strange Situation in Japan only represent a very small proportion of the population, and results may
differ when other samples are used. For example, results in Tokyo seem to mirror findings from the
US, but a more rural Japanese sample found an over-representation of insecure-resistant attachments.
3) Can we be sure that the procedures are standard from one culture to the next? Can we be sure that
the tools used to measure behaviour are standardized in each study? This makes comparisons risky
unless they are exact replications of each other.
Cultural changes
Simonella (2014) conducted the strange situation in Italy very recently to see if the attachment levels were
similar to those recorded in past studies. 76, 12 month olds were used and they found much lower rates of
secure attachment than in previous studies 50% secure (66% in original study) 36% insecure-avoidant (22%
in original study). The researchers suggest it reflects the increasing number of mothers of very young
children working long hours and suggests that cultural changes over time can make a dramatic difference
to attachment patterns.
Oumar (2001) studied the Dogon people of Mali using the strange situation and
compared the findings to North American parents. They found no avoidant
attachments compared to 23% in North American and 67% secure compared to 55% in
North America. The style of child rearing is very different called natural parenting or
attachment parenting with infants constantly in physical contact, breastfeeding on
demand, co-sleeping and responding immediately to distress signals, so children aren’t
left to cry like many western “cry it out” sleep techniques. This study again reflects the
difference in results to the strange situation
depending on culture i.e. is culturally bound and
also highlights the differences in childrearing
around the world. Attachment parenting is becoming increasingly
popular in the west and so research needs to be carried out here to see
if this style of parenting in the west provided different results.
Culture bound
There is the belief that the study is culture bound, i.e. does not have the same meaning outside of the USA
and Western Europe because children and caregivers may respond differently to the strange situations
depending on their cultural experiences. Takahasi (1990) found that the test does not work on Japanese
children because Japanese mothers are so rarely separated from their children that they show very high
levels of separation anxiety and in observations Japanese mothers tended to race to their children and
scoop them up at the reunion stage meaning the response was hard to observe.
19
Bowlby’s theory of maternal deprivation.
Bowlby (1953) The Maternal Deprivation Hypothesis
1. In this hypothesis he suggested that “Mother love in infancy and
childhood is as important for mental health as are vitamins and
proteins for physical health” (1953)
2. if a child does not have a warm, intimate and continuous
relationship with his or her mother (or mother-substitute) then
they would have trouble forming relationships with others later on, and would be at risk of
behavioural disorders such as affectionless psychopathy (see later) but also risk of intellectual
problems e.g. low IQ.
3. If the separation occurs before the age of two and a half (without a substitute) the effects on the
emotional well-being of the child are particularly severe, (critical period)
4. although the child is at risk of deprivation up to the age of five.
5. Bowlby identified circumstances in which maternal deprivation could occur, which included the
mother being imprisoned, divorce or even the mother working full time.
Easy to remember summary
Continuous, warm mother love = Good mental health
Broken, poor mother love = risk of behavioural and intellectual issues, poor internal working model
Broken mother bond < 2 1/2 severe risk but still risks > 5
Bowlby based this theory on his own study the 44 thieves- you can use this study to evaluate his main
theory.
Bowlby’s 44 thieves
Bowlby studied 88 children aged 5 to 16 who had been referred to a child guidance clinic where he worked. 44 were
referred for stealing the other half were a control group; Bowlby diagnosed 16 of the 44 as affectionless psychopaths
– shameless and conscienceless. The control group had been referred for other types of behaviour but none of them
were diagnosed as affectionless psychopaths. Bowlby interviewed the children and their family to build a record of
their early life experiences. Bowlby discovered that 86% of the affectionless psychopaths had experienced early and
prolonged separation from their mothers; only 4% of the control group had experienced such separation. He
concluded that the separation had caused affectionless psychopathy.
Review of the main findings. From 44 thieves, 16 were diagnosed with affectionless psychopaths. 86% of the 16
had suffered early and prolonged separation from their mothers. None of the control group were considered to be
affectionless psychopaths, and only 4% of the control group had suffered early and prolonged separation.
20
Evaluation of Bowlby’s maternal deprivation hypothesis
Research methods - Bias
Research methods-interviews
Remember to link it to the study generic
evaluations get limited marks!
Deprivation versus privation
Individual differences
Real world application
Sample bias- the sample were children who were
had all been referred to his clinic for issues with
their behavior so we can’t generalize these results
to all children as what about children without issues
that were separated?
Researcher bias- Bowlby carried out this research
and came up with the term affectionless
psychopaths and decided who fitted into each
group so there is an issue with objectivity here.
Interviews used so issues with extraneous variables
and other problems related to interviews (see
research methods.
-Interviews used so Bowlby was able to go into
depth and find out rich detailed information on
what is a sensitive area.
All of the negative evaluation points reduce the
support for Bowlby’s maternal deprivation
hypothesis.
Rutter (1976) claimed that Bowlby was actually
mixing up the concepts of deprivation and
privation. He claims Bowlby never made it clear
whether the child’s attachment bond had actually
even been there in the first place and that in fact
the severe long-term damage that Bowlby
associated with deprivation was more likely to be
the result of privation.
Barrett (1997) reviewed various studies on
separation and concluded that actually securely
attached and more mature children may actually
cope better and be less affected than insecurely
attached children.
Bowlby’s theory and work by Robertson and himself
had an enormous effect on childrearing and led to
major social change in the way children were cared
for in hospitals. Before the research parents were
discouraged or even forbidden from visiting children
who had to go into hospital but his research and
footage showing how distressed these children
became from Robertson changed this (See
homework sheet for more details).
21
Romanian orphan studies: Effects of Institutionalisation
Institutionalisation refers to when children are
looked after somewhere other than a home – for
instance, they may be in some kind of orphanage or
children’s home. Often (and particularly in the past)
children had so many changes of carers in the
institution that it was impossible for them to form
any attachments. If the children had been in the
institution since they were babies, privation may
result where the child has formed no attachments.
In Romania many children were placed in orphanages, although they were not always orphans. The grim
conditions in the orphanages became known worldwide from 1989 and became a tragic opportunity to
look at the effects of institutionalisation. Your background
research will have explained why Romania was particularly
looked at.
Some of the Romanian orphans were adopted by British
families. A team of psychologists (called the English and
Romanian Adoptees Study Team) led by Michael Rutter have
studied these children and the effects of their early
experience. The studies are known as the ERA studies and
they are a series of natural experiments that have been published at various times; the children have been
studied soon after arrival, at 4, 6, 11 and 15 and there are plans to study them into adulthood, so they are
also longitudinal studies as they take place over time. Remember that a natural experiment is one in which
the IV (exposure to conditions in the Romanian orphanages or lack of exposure for the control group) is not
manipulated by the experimenter because it exists prior to the study; the experimenters simply measure
the effect of the IV.
Rutter (1998)
Procedures -111 Romanian children were assessed on a variety of measures of
physical and intellectual ability on arrival in Britain. Most of them had been in
institutional care from shortly after they were born but they were naturally split into
3 conditions
1-adopted before 6 months
2-adopted between 6 months and 2 years and
3-adopted after 2 years.
The orphans were assessed for height, head circumference and cognitive functioning on arrival in the U.K
and assessed again at 4. A control group of 52 British adopted children were also assessed to ascertain
whether negative effects were due to separation from carers or the institutional conditions of the
Romanian orphanages.
Findings-The children’s IQ was tested upon arrival in the UK and the average score for the Romanian
orphans was 63. For those adopted when over 6 months old, the average was 45. Physical development
was also poor, 51% of them being in the bottom 3% of the population for weight. They were also shorter in
height than was normal for their age and had smaller head circumferences.
22
The Romanian orphans were tested again at the age of 4 and compared to a control group of 52 Britishadopted children, all aged 4, who had showed none of the negative effects suffered by the Romanians.
-At the age of 4 orphans adopted before 6 months showed no significant differences in either intellectual
or physical development with the control group. All the children had improved though with the average IQ
of the Romanians increasing from 63 to 107. However for those adopted after 6 months, it had gone from
45 to 90. The older adoptees tended to do less well in terms of physical development too.
In follow up studies when the children were 6 and 11, Rutter (2007) found that many had normal levels of
functioning. However 54% of children who displayed disinhibited attachment at 6 years old still displayed
disinhibited attachment it at 11 and many of them were receiving help from either special educational and
or mental health services.
Disinhibited attachment is characterised by a lack of close, confiding relationships, rather indiscriminate
friendliness and clingy, attention-seeking behaviour, a relative lack of differentiation in response to adults
(treating them all alike, a tendency to go off with strangers and a lack of checking back with a parent in
anxiety-provoking situations.
Le mare and Audet (2006) carried out a longitudinal study on 36 Romanian orphans adopted into Canada.
They were looking specifically at physical growth and health and found that adopted orphans were
physically smaller than control group at 4 ½ years old but that this difference had disappeared by 10 ½ and
the same was true for physical health so shows that recovery is possible from the physical effects of
institutionalisation.
Evaluation of Romanian orphan studies
Reliability
Many other studies of Romanian orphans have
found similar results to Rutter. Morison (2005) who
studied Romanian orphans in Canada, and O’Connor
(1999) who found indiscriminate friendliness (part
of disinhibited attachments-treating strangers like a
primary caregiver) was positively correlated to
length of time spent in an institution. This suggests
that Rutter’s findings are reliable.
Longitudinal studies
The children were followed over many years and so
we are able to see the real life, long term effects of
institutionalisation; a very bleak picture would’ve
been painted after 4 years! These studies are also
continuing into the future so we will be able to see
even longer term effects.
Natural experiments and extraneous variables
As these pieces of research had to be natural
experiments there are of course going to be issues
with extraneous variables and difficulties in
establishing cause and effect. For instance Rutter
acknowledges that is was difficult to find out
information about the quality of care the children
received in the institutions and thus the level of
privation. Some children coped much better than
23
others and it is thought that some children may
have received special attention in the orphanages if
they maybe smiled more for instance and so did
have some early attachment experiences this is
another variable uncontrolled.
The findings of Romanian orphanage studies have
led to improvements in the way children are cared
for in institutions and so have been immensely
valuable in practical terms (Langton 2006). Children
in institutions now are assigned a key worker and
have perhaps only one or two other caregivers
responsible for them so that the children have a
change to develop normal attachments, avoiding
disinhibited attachments.
Real life application
Summary of the effects of institutionalization






Some of the negative outcomes shown by the Romanian
children could be overcome through adequate substitute
care. Remember at aged 11 just under half of the children in
Rutter’s study had normal levels of functioning.
But intervention should take place before 6 months of age as
those adopted after tended to have lower IQ’s and be less
physically developed.
Many children, especially those who had been adopted later,
displayed disinhibited attachment - characterised by a lack of close, confiding relationships, rather
indiscriminate friendliness and clingy, attention-seeking behaviour, a relative lack of differentiation
in response to adults (treating them all alike, a tendency to go off with strangers and a lack of
checking back with a parent in anxiety-provoking situations.
Some research shows that the negative physical effects of institutionalization can be reversed by 10
1/2.
Separation from mother alone is not sufficient to cause negative outcomes as British children had
been separated but were not developmentally delayed
The effects of Romanian orphans are still not fully clear as they still need to be followed into
adulthood to see if the negative effects can still be overcome with more time.
24
The influence of early attachment on childhood and adult relationships,
including the role of the internal working model.
The internal working model
We have covered this already when looking at explanations of
attachment. It was suggested by Bowlby that a child’s first
relationship with their primary attachment figure forms a mental
representation for the child and this relationship acts as a
template or shapes all future relationships whether they be
childhood, romantic, plutonic, or with their own children. The
continuity hypothesis is based upon the internal working model
and says that the specific attachment types of children are also
reflected in their adult relationships.
Childhood relationships: securely attached children have better friendships and are the least likely to
bully and be bullied.
Belsky (1999) found that 3-5 year old securely attached children were more curious, resilient, selfconfident, got along better with other children and were more likely to form close relationships.
Why?
Securely attached children have higher expectations that others are friendly and trusting and so enables
easier relationships with others and closer relationships as they are prepared to be trusting and let people
get close to them.
Smith (1998) assessed attachment type and bullying involvement
using standard questionnaires in 196 children aged 7-11 from
London and found that secure children are unlikely to be involved
in bullying. Insecure-avoidant children were most likely to be
victims and insecure-resistant were most likely to be bullies.
Why?
As securely attached children tend to be more confident (Belsky
199) they are less likely to be the target of bullies for fear that
they will stand up for themselves and as they have close
friendships bullies risk then having the support of other children.
Adult romantic relationships: securely attached children have longer-lasting romantic relationships
Hazan and Shaver (1987) The love quiz
Procedure
They analysed the responses from 620 respondents of a “love quiz” they published in the rocky mountain
news (an American small-town newspaper).
Sample-205 were from men, 415 women; 14-82 years old, 91% heterosexual, 42% married, 28% divorced
or widowed, 9% co-habiting and 31% dating (some fitted more than one category).
The questionnaire firstly asked questions to assess current or most important relationship, secondly it
asked questions about attitudes towards love as an assessment of the internal working model. Thirdly
investigated attachment history to identify current and childhood attachment types by asking respondents
25
were to pick which of three descriptions best applied to their inner feelings about romantic relationships
(see below)
Findings
The table below shows the percentage of respondents classified as secure, avoidant or resistant.
Attachment style
% of
Response
respondents
Secure
56
I find it easy getting close to others and am comfortable
depending on them and having them depend on me. I don’t
worry about being abandoned or about someone getting close
to me
Insecure-avoidant
23
I am uncomfortable being close to others; I find it difficult to
trust them, difficult to depend on them. I am nervous when
anybody gets close to and love partners want me to be more
intimate than I feel comfortable being.
Insecure-resistant
19
I find others are reluctant to get close as I’d like. I worry my
partner doesn’t really love me or won’t stay with me. I want to
merge completely with another person and this desire scares
people away.
Table below shows what percentage of each attachment type endorsed each statement.
Internal working model statements that
Secure
Avoidant
respondents agreed with
It is easy to fall in love. I feel myself beginning to fall
9%
4%
in love often
Intense romantic love is common at the beginning of 28%
41%
a relationship but rarely lasts forever.
It is rare to find someone you can fall in love with
43%
66%
The kind of head-over-heels love depicted in novels
13%
25%
and movies doesn’t exist.
Resistant
20%
34%
56%
28%
-Securely attached respondents tended to have a positive internal working model (see table)
-They also found a positive correlation between attachment type and love experiences; the more attached
a person was the more positive they found their love experiences. Securely attached respondents
described love experiences as happy, friendly, trusting and were able to accept and support their partner
despite their faults.
-Securely attached love relationships were more enduring- lasting on average 10 years compared to five
for resistant and 6 avoidant and if married they tended not to divorce.
-Both insecure types were vulnerable to loneliness with resistant being the most vulnerable
-Avoidant types tended to reveal jealousy and fear of intimacy.
Relationships with own children.
If the internal working model is to be believed then childhood attachment
type should affect parenting style i.e. if you are a securely attached child then
you should have a secure relationship with your children and vice versa. This
idea is supported by Harlow’s studies we looked at earlier as the motherless
monkeys went onto become bad mothers but do human studies find the
same thing?
26
Bailey (2007) considered the attachments of 99 mothers to their babies and to their own mothers using
the strange situation and an adult attachment interview and they found the majority of women had the
same attachment classification both to their babies and their own mothers. Quinton (1984) compared 50
women raised in institutions with 50 women raised as home and found that when the women were in their
20’s the ex-institutional women experienced extreme difficulties acting as parents and more of their
children has spent time in care.
Evaluation of early attachment on childhood and adult relationships
Research methods evaluation
Remember to link it to the study generic
evaluations get limited marks!
Most of the research mentioned here used
questionnaires or interviews and largely self-report
and so encounter the common strengths and
weaknesses of these methods and you can add
more evaluation for yourself (see research pack).
Social desirability bias
Some of the problems to consider are social
desirability bias especially in Hazan and Shaver as
participants may be reluctant to admit for instance
that they begin to find themselves falling in love
often. This means the validity of the research is
limited because they depend on respondents being
honest and having a realistic view of relationships.
Retrospective data
Most studies rely on asking adults to recall their
early lives in order to assess infant attachment but
such recollections are likely to be flawed as our
memories of the past are not always that accurate.
Causation
Also the research linking the internal working model
to relationships is correlational rather than
experimental so therefore we can’t claim that the
reason for the later relationship style is the earlier
attachment there may be other reasons such as the
child’s temperament. It also means fortunately that
if you have had a poor relationships as a child you
can still have good, healthy and secure relationships
as an adult. So these factors limit the internal
working.
Some longitudinal research contradicts the model
When people were tracked from childhood and
followed to adulthood (rather than self-report as
adults) then the results contradict the internal
working model. Zimmerman(2000) studied 44
German children at 12-18 months to assess
attachment style then again at 16 and found that
childhood attachment was not a good predictor of
adolescent attachment and that life events often
27
altered secure attachments to insecure.
Rutter (1999) also found the change could occur in
the other direction and in his study found that a
group of people who experienced problematic
relationships with their parents went on to achieve
secure, stable and happy adult relationship.
Low correlations
Steele (1998) found only a small correlation of 0.17
between having a secure attachment type in
childhood and early adulthood. Fraley (2002)
conducted a review of 27 samples where infants
were assessed in infancy and reassessed up to 20
years later and found correlations ranging from 0.50
to as low as 0.10. So such correlations suggest that
attachment type is not very stable.
28