Representations of attachment in mothers and their one-year-old infants

The assessment of representational risk (ARR): Development and psychometric properties of a new coding system for assessing risk in the parent-infant relationship

There are few clinically valid tools that can be used to assess potential parent-infant relational risk. This study describes the development and initial validation of the assessment of representational risk (ARR) coding system to be applied to the parent development interview (Slade, A., Aber, J. L., Bresgi, I., Berger, B., & Kaplan, M. (2004). The Parent Development Interview - Revised. Unpublished protocol. New York, NY: The City University of New York.) for assessing potential risk in caregivers' representations of their infant, themselves as parents, and their relationship. The ARR was developed and validated in three samples in England. A review of the literature informed the selection of 10 items. It had a three-factor structure that was used to inform subscales: hostile, helpless, and narcissistic caregiving representations. The subscales and total risk scores showed good criterion validity for discriminating between high and low risk samples and good concurrent validity with measures of parental psychopathology and parent-infant interaction. The ARR is a potentially valuable coding system for identifying risk in early attachment relationships.

Maternal pre and perinatal experiences with their full-term, preterm and very preterm newborns

Background

Mothers’ reports about pregnancy, maternity and their experiences during the perinatal period have been associated with infants’ later quality of attachment and development. Yet, there has been little research with mothers of very preterm newborns. This study aimed to explore mothers’ experiences related to pregnancy, premature birth, relationship with the newborn, and future perspectives, and to compare them in the context of distinct infants’ at-birth-risk conditions.

Methods

A semi-structured interview was conducted with women after birth, within the first 72 h of the newborn’s life. A total of 150 women participated and were divided in three groups: (1) 50 mothers of full-term newborns (Gestational Age (GA) ≥ 37 weeks; FT), (2) 50 mothers of preterm newborns (GA 32–36 weeks; PT) and (3) 50 mothers of very preterm newborns (GA < 32 weeks; VPT).

Results

Mothers of full-term infants responded more often that their children were calm and that they did not expect difficulties in taking care of and providing for the baby. Mothers of preterm newborns although having planned and accepted well the pregnancy (with no mixed or ambivalent feelings about it) and while being optimistic about their competence to take care of the baby, mentioned feeling frightened because of the unexpected occurrence of a premature birth and its associated risks. Mothers of very preterm newborns reported more negative and distressful feelings while showing more difficulties in anticipating the experience of caring for their babies.

Conclusion

The results indicate that Health Care Systems and Neonatal Care Policy should provide differentiated psychological support and responses to mothers, babies and families, taking into account the newborns’ GA and neonatal risk factors.

Improving Maternal Representations in High-Risk Mothers: A Randomized, Controlled Trial of the Mom Power Parenting Intervention

A key mechanism of risk transmission between maternal risk and child outcomes are the mother's representations. The current study examined the effects of an attachment-based, trauma-informed parenting intervention, the Mom Power (MP) program, in optimizing maternal representations of high-risk mothers utilizing a randomized, controlled trial design (NCT01554215). High-risk mothers were recruited from low-income community locations and randomized to either the MP Intervention (n = 42) or a control condition (n = 33) in a parallel design. Maternal representations were assessed before and after the intervention using the Working Model of the Child Interview. The proportion of women with balanced (secure) representations increased in the MP group but not in the control group. Parenting Reflectivity for mothers in the treatment group significantly increased, with no change in the control condition. Participation in the MP program was associated with improvements in a key indicator of the security of the parent-child relationship: mothers' representations of their children.

Relationship Change at the Transition to Parenthood and Security of Infant-Mother Attachment

Attachment research has so far favoured maternal and child characteristics as determinants of secure or insecure attachment relationships between infant and mother. Several recent findings now suggest that the broader context of the family and the couple’s relationship should receive more attention as possible determinants.

The quality of parental relationships was assessed in a prospective longitudinal study using 28 women and 23 of their partners. A Partnership Questionnaire was administered at four points of measurement from the beginning of pregnancy to one year after the birth of the first child. At 13 months of age the infants and their mothers were observed in the Strange Situation Procedure. Statistical analysis revealed two systematic results. (1) Both wives and husbands from families with a secure infant-mother attachment judged the quality of their partnership as more satisfying than parents with insecurely attached infants: lower decrease in “tenderness”, relative absence of “quarrelling” as perceived by the wives, and husbands’ “general happiness” across the transition to parenthood were related to secure mother-child relationships. (2) After becoming parents, both parents perceived a decline in the quality of their marital relationship. The results also suggest that parental partnership and mother-child relationship are systematically interdependent and support the idea that early parenthood is a critical time for establishing different attachment qualities.

Negative and distorted attributions towards child, self, and primary attachment figure among posttraumatically stressed mothers: what changes with Clinician Assisted Videofeedback Exposure Sessions (CAVES)

This study found that within a non-referred community pediatrics clinic sample, the severity of mothers' trauma-related psychopathology, in particular, their interpersonal violence-related (IPV) posttraumatic stress, dissociative, and depressive symptoms predicted the degree of negativity of mothers' attributions towards their preschool age children, themselves, and their own primary attachment figure. Results also showed that mothers with IPV-related posttraumatic stress disorder (PTSD) as compared to non-PTSD controls showed a significantly greater degree of negativity of their attributions toward their child, themselves and their primary attachment figure during childhood. The study finally found a significant reduction in the degree of negativity of mothers' attributions only towards their child following a three-session evaluation-protocol that included a form of experimental intervention entitled the "Clinician Assisted Videofeedback Exposure Session(s)" (CAVES), for mothers with IPV-PTSD as compared to control-subjects.

Antecedents of maternal sensitivity during distressing tasks: integrating attachment, social information processing, and psychobiological perspectives

Predictors of maternal sensitivity to infant distress were examined among 259 primiparous mothers. The Adult Attachment Interview, self-reports of personality and emotional functioning, and measures of physiological, emotional, and cognitive responses to videotapes of crying infants were administered prenatally. Maternal sensitivity was observed during three distress-eliciting tasks when infants were 6 months old. Coherence of mind was directly associated with higher maternal sensitivity to distress. Mothers' heightened emotional risk was indirectly associated with lower sensitivity via mothers' self-focused and negative processing of infant cry cues. Likewise, high physiological arousal accompanied by poor physiological regulation in response to infant crying was indirectly associated with lower maternal sensitivity to distress through mothers' self-focused and negative processing of infant cry cues.

Associations between first-time expectant women's representations of attachment and their physiological reactivity to infant cry

Associations among 53 primiparous women's Adult Attachment Interview classifications (secure-autonomous vs. insecure-dismissing) and physiological and self-reported responses to infant crying were explored. Heart rate, skin conductance levels, and respiratory sinus arrhythmia (RSA) were recorded continuously. In response to the cry, secure-autonomous women demonstrated RSA declines, consistent with approach-oriented responses. Insecure-dismissing women displayed RSA and electrodermal increases, consistent with behavioral inhibition. Furthermore, insecure-dismissing women rated the cries as more aversive than secure-autonomous women. Nine months postpartum, secure-autonomous women, who prenatally manifested an approach-oriented response to the unfamiliar cry stimulus, were observed as more sensitive when responding to their own distressed infant, whereas women classified prenatally as insecure-dismissing were observed as less sensitive with their own infants.

EARLY POSTPARTUM PARENTAL PREOCCUPATION AND POSITIVE PARENTING THOUGHTS: RELATIONSHIP WITH PARENT-INFANT INTERACTION

Parenting behaviors and parent-infant emotional bonding during the early postpartum months play a critical role in infant development. However, the nature and progression of parental thoughts and their relationship with interactive behaviors have received less research. The current study investigated the trajectory of parental thoughts and behaviors among primiparous mothers (n = 18) and fathers (n = 15) and multiparous mothers (n = 13) and fathers (n = 13), which were measured at the first and third postpartum month. At the third postpartum month, the relationship between parental thoughts and parental interactive behaviors also was tested. Mothers and fathers showed high levels of preoccupations and caregiving thoughts during the first postpartum month that significantly declined by the third postpartum month. In contrast, positive thoughts about parenting and the infant increased over the same time interval. Mothers presented higher levels of preoccupations and positive thoughts than did fathers, and first-time parents reported more intense preoccupations than did experienced parents. Although maternal sensitivity was inversely related to maternal anxious thoughts, paternal sensitivity was predicted by higher levels of anxious as well as caregiving and positive thoughts.

In an idealized world: can discrepancies across self-reported parental care and high betrayal trauma during childhood predict infant attachment avoidance in the next generation?

Adult caregivers' idealization of their parents as assessed by the Adult Attachment Interview is a risk factor for the intergenerational transmission of the insecure-avoidant attachment style. This study evaluated a briefer screening approach for identifying parental idealization, testing the utility of prenatal maternal self-report measures of recalled betrayal trauma and parental care in childhood to predict observationally assessed infant attachment avoidance with 58 mother-infant dyads 18 months postpartum. In a logistic regression that controlled for maternal demographics, prenatal psychopathology, and postnatal sensitivity, the interaction between women's self-reported childhood high betrayal trauma and the level of care provided to them by their parents was the only significant predictor of 18-month infant security versus avoidance. Results suggest that betrayal trauma and recalled parental care in childhood can provide a means of identifying caregivers whose infant children are at risk for avoidant attachment, potentially providing an efficient means for scientific studies and clinical intervention aimed at preventing the intergenerational transmission of attachment problems.

Maternal relationship style, paediatric health care use and infant health

OBJECTIVE

The aim of the current exploratory study was to examine the relationships between maternal relationship style, paediatric health care use and infant health variables in a sample of middle-class and ethnically diverse mother-infant dyads.

METHODS

An initial cross-sectional cohort study obtained demographic and self-report data on mothers' relationship styles. As an extension of the original study, infants' patient files were reviewed for the year following initial recruitment to obtain data regarding the use of paediatric health care services and infant health. The final sample included 64 mothers and their infants.

RESULTS

Correlational analyses revealed that mothers' higher endorsement of a dismissive relationship style were associated with fewer acute care visits and fewer reported infant illnesses.

CONCLUSIONS

Compared with other relationship styles, mothers who highly endorsed a dismissive relationship style tended to use fewer acute paediatric health care services and reported fewer infant health problems. However, further longitudinal research is needed to clarify these relationships.

Adult attachment measures: a 25-year review

OBJECTIVE

Over the past 25 years, attachment research has extended beyond infant-parent bonds to examine dyadic relationships in children, adolescents, and adults. Attachment has been shown to influence a wide array of biopsychosocial phenomena, including social functioning, coping, stress response, psychological well-being, health behavior, and morbidity, and has thus emerged as an important focus of psychosomatic research. This article reviews the measurement of adult attachment, highlighting instruments of relevance to-or with potential use in-psychosomatic research.

METHODS

Following a literature search of articles that were related to the scales and measurement methods of attachment in adult populations, 29 instruments were examined with respect to their utility for psychosomatic researchers.

RESULTS

Validity, reliability, and feasibility were tabulated on 29 instruments. Eleven of the instruments with strong psychometric properties, wide use, or use in psychosomatic research are described. These include the following: Adult Attachment Interview (George, Kaplan, and Main); Adult Attachment Projective (George and West); Adult Attachment Questionnaire (Simpson, Rholes, and Phillips); Adult Attachment Scale (and Revised Adult Attachment Scale) (Collins and Read); Attachment Style Questionnaire (Feeney); Current Relationship Interview (Crowell and Owens); Experiences in Close Relationships (Brennan, Clark, and Shaver) and Revised Experiences in Close Relationships (Fraley, Waller, and Brennan); Parental Bonding Instrument (Parker, Tupling, and Brown); Reciprocal Attachment Questionnaire (West and Sheldon-Keller); Relationship Questionnaire (Bartholomew and Horowitz); and Relationship Scales Questionnaire (Grifiin and Bartholomew).

CONCLUSION

In addition to reliability and validity, investigators need to consider relationship focus, attachment constructs, dimensions or categories of interest, and the time required for training, administration, and scoring. Further considerations regarding attachment measurement in the context of psychosomatic research are discussed.

Socioemotional development in adolescents at risk for depression: The role of maternal depression and attachment style

We examined the impact on adolescent socioemotional functioning of maternal postnatal depression (PND) and attachment style. We also investigated the role of earlier aspects of the child's development-attachment in infancy, and 5-year representations of family relationships. Ninety-one mother-child pairs, recruited in the postnatal period, were followed up at 13 years. Adolescents were interviewed about their friendships, and their level of emotional sensitivity and maturity were rated. Emotional sensitivity was heightened in girls whose mothers experienced PND; notably, its occurrence was also linked to insecure attachment in infancy and raised awareness of emotional components of family relationships at 5 years. High emotional sensitivity was also associated with adolescent depressed mood. Raised social maturity was predicted by a secure maternal attachment style and, for girls, by exposure to maternal PND. Precursors of adolescent social maturity were evident in the narrative coherence of 5-year family representations. Higher social maturity in the friendship interview was also associated with overall good adjustment.

Stability and change in mothers' internal representations of their infants over time

This study examined predictors of stability and change in women's maternal representations of their children. Participants were 180 women, recruited from the community, half of whom had experienced domestic violence during pregnancy. Maternal representations of were assessed with the Working Model of the Child Interview (WMCI; Zeanah, Benoit, Hirshberg, Barton, & Regan, 1994) during the last trimester of pregnancy and again at the child's first birthday. Results indicated that when collapsed into balanced and non-balanced categories, 71% of the sample was stable over time, and women who had balanced representations had significantly more stable representations than women who had non-balanced representations (p < .001). Income, single parenthood, abuse status, and depressive symptomatology predicted change. In addition, women who became non-balanced postnatally benefited from having balanced representations while pregnant to buffer the quality of their interactions with their children.

Traumatized mothers can change their minds about their toddlers: Understanding how a novel use of videofeedback supports positive change of maternal attributions

This study explored the use of a brief experimental intervention that integrates principles of infant-parent psychotherapy, videofeedback, controlled exposure to child distress in the context of parental posttraumatic stress disorder (PTSD), and stimulation of parental reflective functioning (RF). The Clinician Assisted Videofeedback Exposure Session (CAVES) was applied to 32 interpersonal violence-exposed mothers of very young children (8-50 months) with respect to change of maternal perception of her child. While we found no significant reduction over two videotaped assessment visits with a mental health professional, we did find a significant reduction in the degree of negativity of maternal attributions towards her child following the videotaped visit focused on the CAVES (p<.01). Maternal RF, a mother's capacity to think about mental states in herself and her child, accounted for 11% of the variance in reduction of maternal negativity after accounting for baseline levels of negativity. Clinician-assisted videofeedback appears to support emotional self-regulation of mothers with violence-related PTSD. Focusing with a therapist on videofeedback of child separation distress exposes mothers to avoided mental states of helplessness and perceived loss of protection. Negative maternal attributions may mark violent trauma-associated emotion dysregulation and projected self-representations of the maltreated mother.

Understanding the link between maternal adult attachment classifications and thoughts and feelings about emotions

The purpose of this study was to examine the associations between maternal representations of attachment, as assessed with the Adult Attachment Interview (AAI; George, Kaplan, & Main, 1996), and mothers' thoughts and feelings about their own emotions and emotions emerging in their toddlers. Eighty-nine adolescent mothers completed the AAI and the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) and Katz, Gottman, Shapiro, and Carrere's (1997) meta-emotion interview for parents of toddlers. Autonomous mothers demonstrated the most open and flexible mindset around a variety of emotions in themselves and their toddlers. Dismissing mothers exhibited a tendency to minimize internalizing emotions in themselves and their children, while unresolved mothers described the most emotion regulatory difficulties.

Associations between parental and child attachment representations

In the present article on intergenerational transmission of attachment representations, we use mothers' and fathers' Adult Attachment Interview classifications to predict a 3-year-old's responses to the Attachment Story Completion Task (ASCT). We present a Q-sort coding procedure for the ASCT, which was developed for children as young as three. The Q-sort yields scores on four attachment dimensions (security, deactivation, hyperactivation, and disorganization). One-way ANOVAs revealed significant mother-child associations for each dimension, although results for the hyperactivation and disorganization dimensions were significant only according to contrast tests. Conversely, no father-child association was found, regardless of the dimension considered. Findings are discussed in terms of the respective part played by each parent in their children's emotional development.

Adult attachment style and parental responsiveness during a stressful event

Despite widespread use of self-report measures of adult attachment, relatively little research has explored the predictive utility of these measures in the domain of parent-child relationships. The present study examined the association between self-reported attachment style and parental responsiveness during a stressful event. Children and their parents were observed while children received an inoculation at a county immunization clinic. Children's reactions to the inoculation were rated and parents' responsiveness was assessed with the Emotional Availability Scales (EAS). Results revealed that children of parents scoring high on self-reported attachment avoidance were more distressed during the inoculation than children of parents scoring low on avoidance. Moreover, parents high on avoidance were less responsive when children were highly distressed, whereas this pattern was reversed among parents scoring low on avoidance. Finally, the influence of adult attachment on parental behavior and children's distress was found to be independent of children's temperament and parental personality. These findings suggest that self-report adult attachment measures may be useful in the domain of parent-child relationships.

Adolescents' Attachment Representations and Developmental Tasks in Emerging Adulthood

The association between attachment representations and adolescents' coping with 3 developmental tasks of emerging adulthood-leaving home, advancing in the capacity for mature intimacy, and developing individuation-was examined. Israeli male adolescents (N = 88) were administered the Adult Attachment Interview during their high-school senior year. A year later, they and their friends reported on the adolescents' adjustment to mandatory military service. Three years later, participants and their parents reported on the adolescents' capacity for intimacy using an in-depth interview and on their individuation. An autonomous state of mind was associated with better coping with basic training and with a higher capacity for mature intimacy but was not associated with markers of individuation. The results highlight the importance of attachment representations in shaping an individual's developmental trajectory.

The Impact of Maternal Characteristics and Contextual Variables on Infant-Mother Attachment

This prospective study examined the effects of maternal characteristics, social support, and risk factors on infant-mother attachment in a heterogeneous sample. Two hundred and six women between the ages of 18 and 40 were interviewed during their last trimester of pregnancy and 1 year postpartum. Structural equation modeling revealed that maternal attachment experiences were significantly related to prenatal representations of the infant and of the self as a mother, which were significantly related to infant-mother attachment assessed by the Strange Situation. Maternal risk factors were significantly related to prenatal representations, and social support from other women predicted infant-mother attachment. The overall model indicated a good fit. Thus, both individual and contextual factors were important in explaining infant attachment security.

Intergenerational Family Relations and Sexual Risk Behavior in Young Women

PURPOSE

To more fully understand the associations between family variables and sexual behavior of young women.

STUDY DESIGN AND METHODS

Forty-two female clients of an urban youth agency (16 to 25 years of age) were interviewed about intergenerational stories and contact with extended families. Individuation was measured by self-report using the intergenerational individuation subscale of the Personal Authority in the Family System Questionnaire. Self-report questions were used to gain information about the young women's sexual risk behavior. Data were analyzed using Pearson correlation, independent t-tests, and multiple regression. Family stories were analyzed for themes of resilience or vulnerability.

RESULTS

Young women who perceived their connection with previous generations in a resilient light tended to be more individuated and report less risky sexual behaviors.

CLINICAL IMPLICATIONS

These results may give direction in the design of family interventions for reducing sexual risk taking and enhancing positive health behavior. Interventions would include promoting a healthy degree of attachment between adolescents and their families, exploration of family of origin issues, and referral for more intensive services when needed.

Adult attachment stance and spouses' marital perceptions during the transition to parenthood

The relationship between adult attachment and marital perceptions was examined in couples making the transition to parenthood. The Adult Attachment Interview was administered to spouses prenatally, and marital perceptions were assessed prenatally and 3, 12, and 24 months postnatally. The couples were also observed during a marital interaction task. Spouses were assigned to groups based on their attachment security (secure/insecure) and mean level (low/high) of negative emotional escalation in their marital interactions. A series of two-way (attachment security by negative emotional escalation) repeated multivariate analyses of variance, followed by univariate analyses, revealed that insecure husbands and the wives of insecure husbands exhibited different patterns in their marital perceptions than secure husbands and the wives of secure husbands when in high negative escalation marriages. Insecure husbands and the wives of insecure husbands in high negative escalation marriages reported greater declines in positive marital perceptions or less positive marital perceptions overall across the two-year postnatal period than secure husbands and wives of secure husbands in similarly high conflict marriages. Such differences between secure and insecure husbands (and their wives) were not apparent in low negative escalation marriages. Findings are discussed in terms of the protective effect a secure attachment may convey upon spouses during stressful periods in their marriage.

The role of attachment and cognitive inhibition in children's memory and suggestibility for a stressful event

There has been increasing interest in children's abilities to report memories of and resist misleading suggestions about distressing events. Individual differences among children and their parents may provide important insight into principles that govern children's eyewitness memory and suggestibility for such experiences. In the present study, 51 children between the ages of 3 and 7 years were interviewed about an inoculation after a delay of approximately 2 weeks. Results indicated that parents' attachment Avoidance was associated with children's distress during the inoculation. Parental attachment Anxiety and the interaction between parental Avoidance and children's stress predicted children's memory for the inoculation. Cognitive inhibition was also a significant predictor of children's memory errors and suggestibility. Theoretical implications concerning effects of stress and individual differences on children's eyewitness memory and suggestibility are discussed.

Emotional availability in the mother-infant dyad as related to the quality of infant-mother attachment relationship

Dyadic emotional availability and infant-mother attachment relationship were examined in 687 Israeli dyads. Concurrent assessments used the Strange Situation procedure (Ainsworth, Blehar, Waters, & Wall, 1978) for evaluating infants' attachment relationship, and the Emotional Availability Scales (Biringen, Robinson, & Emde, 1993) for evaluating the quality of mother-child interaction. It was found that higher scores on the Emotional Availability Scales were associated with infant attachment security. In addition, it was found that the Emotional Availability Scales discriminated between insecure-ambivalent and secure attachment classification, but were not informative about unique characteristics of emotional availability in dyads with avoidant and disorganized infants. Our findings contribute to the cross-cultural validation of Emotional Availability Scales against infants' attachment security.

Motherhood: the emotional awakening

During her physical pregnancy, a mother also undergoes a psychological pregnancy in which she imagines her baby as having certain physical and intellectual attributes. Shortly after birth, a mother finds that her "imagined baby" encounters her "real baby"--a moment when her prepartum thoughts of an idealized baby may conflict with the reality of her newborn. This encounter contributes to the future of their relationship in many ways, some of which may have clinical implications. For example, in the case of a preterm or premature birth, the mother is also premature psychologically and may need professional support to effectively deal with her unexpected infant. The Brazelton Neonatal Behavioral Assessment Scale can be an effective therapeutic tool during these perinatal encounters.

Mothers' representations of their infants assessed prenatally: stability and association with infants' attachment classifications

The stability and predictive validity of classifications of mothers' representations of their infants as determined by the Working Model of the Child Interview (WMCI) were examined. Concordance between mothers' representations of their infants assessed prenatally and again one year later and infant Strange Situation (SS) attachment classifications at 12 months was also examined. WMCI classifications were stable over 12 months in 80% of mothers, compared to 51% expected by chance alone. Pregnancy WMCIs predicted infant SS classifications in 74% of cases, compared to 54% expected by chance. Concordance between 11-month WMCI and 12-months SS classifications was 73% (vs. 55% expected by chance). Problems with the skewed distribution of the sample, the low concordance between pregnancy and 11 months for one of the three classifications, and future directions for research are discussed.

Maternal attachment status, psychosocial stressors and problem behaviour: follow-up after parent training courses for conduct disorder

37 Mothers of children with conduct disorder who had received Parent Management Training (PMT) were interviewed using the Adult Attachment Interview, 13-43 months after treatment. Child behaviour levels at follow-up were strongly associated with the scores at referral in the unresolved attachment group, but not in the resolved group. This was reflected in a statistical interaction between referral behaviour score and resolved/unresolved attachment status in a regression model. This model, which also included a strong independent contribution from a composite psychological stress index, explained 66% of the variance in follow-up behaviour scores. The implications of these results for predicting the outcome of PMT are discussed.

Stability and transmission of attachment across three generations

Stability of adult attachment and transmission of attachment across 3 generations were examined in a longitudinal study of 96 infants, mothers, and maternal grandmothers. The Adult Attachment Interview (AAI) was used to assess attachment in mothers (during pregnancy and when infants were 11 months old) and grandmothers (any time during the study). The Strange Situation (SS) was used to assess attachment in infants at 12 months. Both the 3- and 4-category classification systems of the AAI and SS were used. Mothers' AAI classifications were stable over 12 months in 90% (3-category) and 77% (4-category) of mothers. Mothers' AAI classifications during pregnancy predicted infants' SS classifications in 81% (3-category) and 68% (4-category) of cases, and grandmothers' AAI classifications in 75% (3-category) and 49% (4-category) of cases. Using log-linear analysis, we show that a simple parent-to-child model accounts for transmission of attachment.