Social & Cognitive - Lecture 3

Attachment II

hypotheses derived from attachment theory - universality and normativity

all infants will become attached to one or more caregivers.

• secure attachment is common

hypotheses derived from attachment theory - continuity

patterns of attachment are stable over the lifespan

hypotheses derived from attachment theory - sensitivity

early attachment security is dependent on caregivers’ responsiveness to children’s signals

• being aware of signals & responding appropriately

• quality matters not who

hypotheses derived from attachment theory - competence

secure attachment leads to positive outcomes in a variety of domains

family life in the 21st century - office for national statistics (2020)

• caregivers are working

  • being responsive is not always possible

• majority of men in full-time

• around 50% of women in full-time

• parent primarily at home in attachment

different family forms

• single parents

  • 1 in 6 solo parents

  • 1 in 5 father led

• older sibling care

• adoption, foster

• lesbian, gay

  • 1 in 200 same-sex use assisted reproduction (IVF)

• grandparents

  • 2 in 3 responsible for providing care

• heterosexual, 2 parent

why do we need to study attachment in different family forms - theoretical and practical

• theoretical

  • certain parenting affects a child’s attachment, age of adoption can affect, universality & normativity

  • help to advance theory, should apply to more individuals/cases

• practical

  • can remove stigma on certain types of caregivers, can influence the law

  • diversity, inclusivity, generalisability

madigan et al. (2023) - does the gender of the caregiver matter?

• mothers and fathers

• no significant difference in the amount of secure-attached

• supports normativity

• insecure patterns also significantly different

van den fries et al. (2009) - does the biological relatedness of the caregiver matter?

• most data comes from UK & USA (Western) - issue

• meta-analysis of 17 studies of adopted children

  • no difference in security between adopted/non-adopted children (<12 months)

    • d = 0.08 (N = 524)

  • some evidence for lower levels of security among adopted children (>12 months)

    • d = 0.80 (N = 198)

• meta-analysis of 11 studies of foster children (N = 300)

  • no evidence of difference in security between foster children and biologically related children (d = 0.06)

• adopted in the first year of life showed no difference compared to biological

  • may be lower for adopted after

  • possibly due to institutional care, prior experiences

madigan et al. (2023) - does adoption and fostering matter?

• no significant difference in secure attachment

• can form attachments

• adopted, avoidant - is less likely, may be ‘used’ to feeling neglected

• adopted, disorganised - is more likely, possibly due to history before adoption

different combinations of relatedness

golombok et al. - does adoption and assisted reproduction matter?

• women-led families

• compared attachment security in 4-8 yo children across different family forms

• no difference in secure attachment

• affected british legislation

golombok et al. (1997) - does it matter - lgbtq families

• 3-6 years old

• all biologically related

• during this time had laws on lesbians

• no difference in attachment between lesbian mothers & single heterosexual mothers

• stronger attachment in single heterosexual/lesbian mothers than in heterosexual couples

mcconnachie et al. (2020) - does it matter - lgbtq families

• 10-14 years

• more recent, newer laws on lesbian couples

• adopted

• no difference in attachment security between gay fathers and lesbian mothers

• difference in heterosexual parents

van rijn et al. (2025) - does the caregiver type matter - same-sex parents

• mothers (DI)

• fathers (surrogacy)

• heterosexual (IVF)

• absolutely no difference in attachment

• first study using the strange situation for same-sex couples

golombok et al. (2018) - does the type of caregiver matter? - conclusion

“The number, gender, sexual orientation and biological relatedness of parents to their children are less influential in children’s psychological development than are family processes such as the quality of family relationships and the social environment in which children are raised.”

• changed british legislation - story is not over yet…

• type of caregiver does not matter

  • supports universality and normativity

deneault et al. - calling for greater research

• “a call to represent the current diversity of family forms in attachment research”

• <1% of all studies on attachment focus on diverse family forms (need more)

• grandparent-child attachment

  • is a large portion of caregivers so very important

  • time spent with grandparents and attachment security with grandparent

• multi-generational households

  • co-parents (e.g. live-in grandparents, uncles, aunts)

• omission: researchers often don’t ask parents about gender/sexuality

• majority of studies in the US & UK

  • different family forms in different cultures

Feugé et al. (2020) carried out a study of gay fathers and their adopted children. The researchers found that 75% of children in the study were securely attached, and 88% of the fathers in the study exhibited sensitive parenting. Parental sensitivity was positively associated with secure attachment. These results support the _____ hypothesis. The results also suggest that _____ matters more for attachment security than _____.

1. normativity

2. sensitive caregiving

3. caregiver gender or seuxal identity

fearon & roisman (2018) - the sensitivity hypothesis

“Attachment theory makes the bold claim that the causes of variation in attachment security are largely if not entirely environmental, and that caregivers’ sensitivity to infants’ attachment cues and communications is the primary environmental determinant.”

• randomisation important so everything is equal

• causal theory (experience —> attachment)

ingredients of a causal theory

• how strong is the association (needs covariation)

• is the association spurious (driven by something else)

• what is the temporal order (need to know to establish causal effect)

is caregiver sensitivity the primary environmental cause of attachment security? - how strong is the association

• parental sensitivity related to the child’s attachment security

• ainsworth et al.

  • 16 hours observation per family

  • how infants performed in the strange situation

  • very small sample

  • enormous +ve result

• de wolff & van ijzendoorn

  • meta-analysis

  • 30 studies, late 70s - late 90s

  • smaller +ve result but still sig.

• lucassen et al.

  • small +ve result

  • fathers

• many varying estimates about how strongly they relate at age 2

madigan et al. (2024) - child attachment security - paternal and maternal sensitivity

• meta-analysis of every single study since 1978

  • 174 studies, 24000 infants

• caregiver sensitivity on performance in the strange situation

• caregiver sensitivity & child-caregiver attachment: r = .25

  • mother-child dyads: r = .26 (159 studies)

  • father-child dyads: r = .21 (22 studies)

• no effect of geographical region (but majority of studies European/American)

madigan et al. (2024) - child attachment security - based on the effect size, would you say that sensitivity is the primary determinant?

• inconclusive

• small effect (still sig.)

• other things could be influencing the relationship

• it is related, but do not know if the primary determinant

• most studies still carried out in US & Canada

is caregiver sensitivity the primary environmental cause of attachment security? - is the association spurious?

• could something else explain this

• parental sensitivity —> child’s attachment security

  • environmental influence of child’s attachment security

• possible other factor - infant temperament

  • individual differences in how reactive & fearful

  • may affect how the parent responds to them

is the association spurious - infant temperament

• Temperament: Individual differences in reactivity and self-regulation (Rothbart & Ahadi, 1994)

  • Tendency to approach/avoid novel stimuli - anxious/fearful

  • Lability/stability of mood (positive/negative affect) - e.g. crying a lot

• Temperament is heritable in infancy and toddlerhood (~50%) (e.g., Goldsmith et al., 1999).

groh et al. - is temperament related to attachment security?

• Variation in attachment security is related to caregiving environment OR

• Variation in attachment security is related to endogenous characteristics in the child OR

  • temperament factors

• Temperament influences the type of insecure attachment that infants have with insensitive caregivers

  • a mix of both

• temperament can be measured behaviourally (response to a stimulus), social smiling (recording initial reaction)

• pulled together all available studies where they measured temperament & attachment security

groh et al. - is the association between attachment and temperament spurious?

• no differences between secure and insecure children in temperament

  • not influencing factor

  • challenges idea endogenous characteristics might influence behaviour

• insecure resistant and temperament

  • significant effect size

  • however still a small effect size

• clumping together all insecure children is a strange way to analyse data

madign et al. - is the association spurious - genetics

• some may be explained by temperamental factors, but possibly simple passive genetic transmission

• adoptive parents’ sensitivity is associated with child attachment security

  • e.g. Stams et al. (2002), Juffer et al. (2005), Schoenmaker et al. (2015)

• large difference in a no. of studies

  • do not have as much info

• adoptive studies - not biologically related

  • isn’t just passive genetic transmission, but an environmental association

baker mans-kranenburg et al. - what is temporal order?

in first two years of life, experiences with caregiver lead you to develop a model/view of a caregiver

• meta-analysis (K = 70, N = 9957)

• random allocation to two different conditions

  • intervention - parental sensitivity training

• 51 RCTs (N=6,282): Intervention improved sensitivity (d = 0.33)

• 23 RCTs (N=1255): Intervention improved attachment security (d = 0.20)

• both small effect sizes

• causal evidence between the two

  • randomisation rules out potential contaminants

• a lot of studies haven’t had many long-term follow ups

  • does the intervention last?

the competence hypothesis - bowlby

“It appears that there is a very strong case indeed for believing that prolonged separation of a child from his mother (or mother substitute) during the first five years of life stands foremost among the causes of delinquent character development”

• children getting clinical support for, e.g. criminal behaviours

• causal claim

the competence hypothesis - ainsworth

“With secure infants, confidence in theaccessibility and responsiveness ofthat figure enables the child to venture forth to learn about his surroundings… He builds up … a sense of competence, which gives him confidence that he can have an effect on the world around him…”

• “people care for me”, “i’m important”

internalising

depressive, anxious symptoms

• turn the distress on yourself

externalising

turn our emotional distress outwards onto the world

• e.g. aggression, cognitive problems, hyperactivity

fearon et al. - why might attachment matter for mental health?

• sensitive caregiver

  • tune into signals & respond appropriately

  • go to caregiver for comfort & support

• working model

  • view others as available/supportive

  • view self as worthy of love/care

  • positive expectations of others

  • self confidence/self-regulation

• security/insecurity & mental health theories

  • insecure-resistent (fearful of new & social situations) —> internalising

  • insecure-avoidant (struggling to form intimate relationships) —> externalising

  • disorganised —> internalising/externalising

fearon et al. - attachment and externalising problems

to further the table:

• avoidant vs. others (SSP only, N = 3054) — d = 0.13

• resistant vs. others (SSP only, N = 2910) — d = 0.05

all longitudinal studies using the strange situation

• relationship between security at age 10 & mental health problems at age 70

secure vs. insecure / secure vs. insecure (male) / disorganised vs. others - elevated risk of externalising

groh et al. - attachment and internalising problems

• meta-analysis

• did the same as fearon et al.

• went against previously thought, as resistant found a non-significant result

• attachment and problems

fearon et al. & groh et al. - is attachment related to later mental health?

results:

• resistant —> internalising X (not even externalising)

• avoidant —> externalising / (some evidence)

• disorganised —> internalising X

• disorganised—> externalising /

conclusion

• attachment is related to both a small degree, but the specificity hypotheses are not consistently supported

considerations about meta-analyses

• Confounding factors (e.g., temperament, demographic factors) that could explain outcomes have not been considered.

• Focus on biological parents: potential genetic confounds.

  • Mental health has a high heritability

• Studies are not ‘true’ longitudinal designs: direction of association?

  • have not measured both occasions

• No causal evidence: does change in attachment produce a change in outcome?

  • more elaboration

• Specificity

  • Role of distinct types of insecurity is not clear.

  • Security is associated with a wide range of outcomes

stams et al. - addressing potential confounds

• longitudinal study of adopted children

• meaured in childhood and followed up at 7/8yo

• adoptive parents & children, controlled for confounds (e.g. temperament)

• temperament does not play a role in externalising and internalising

• early parental sensitivity is not sig. related

• attachment security with: social competence & externalising - did have an effect

van ijzendoorn et al. - establishing causal connections - sensitivity and attachment security

• Meta-analysis of interventions to improve parental sensitivity.

  • training studies can improve sensitivity

• Intervention improves parental sensitivity (r = .18) and child attachment security (r = .23).

• No clear (causal) effect of intervention on child externalising (r = .07).

• most studies only included measure of externalising few months or right after training

  • different a year later?

de klyen & greenberg - considerations about competence hypothesis

• Attachment security is unlikely to be a necessary or sufficient cause of mental health problems.

• Multiple pathways are likely to exist:

  • Different risk factors can lead to the same outcome (equifinality)

  • A given risk factor (e.g., resistant attachment) can give rise to multiple different outcomes (multifinality)

  • avoidant —> externalising & externalising

• The influence of attachment security on mental health may be moderated by other factors.

• Links between risk factors and mental health problems may be nonlinear.

• The link between attachment and mental health may vary across different points in development.

when it comes to attachment, does the quality of caregiving matter more than the type of caregiver?

• evidence using different family forms

• responsive/sensitive caregiving

many of the longitudinal studies involve measurement of attachment in infancy only & measurement of mental health at a later timepoint only. what can we conclude from these studies?

cannot really establish temporal connection (need measure of both twice)