Maternal and infant temperamental predictors of attachment: A meta-analytic review.

Abstract

Because of the possible implications of infant-mother attachment for later adjustment, we examined the extent to which it could be predicted by mother interactional variables and infant proneness to distress. The meta-analysis demonstrated that sensitive, responsive maternal interaction predicted the security of attachment in Ainsworth and Wittig's (1969) "strange situation." However, the strength of the relation was less than many narrative reviews have suggested. Proneness to distress, which is conceptualized as a temperamental variable, predicted resistance, which is a behavioral pattern in the strange situation that is thought to indicate one variety of insecure attachment. The strength of this association was low but was roughly comparable to that in the maternal domain. In both the maternal and infant domains, we attempted to predict the effect sizes by study characteristics such as method of assessment, sample composition, and age of subjects at the time of study. The pattern of results highlights several continuing methodological problems in the field and suggests that additional explanantory concepts are needed.Variation in maternal caregiving and infant temperament are two potential predictors of the quality of the mother-infant relationship in early life. Our review of these issues is motivated, in part, by the widespread belief that mother-infant attachment influences later adaptation, including the development of behavioral disorders (Sroufe, 1983). Although the cumulative evidence for the effect of variation in attachment quality on psychopathology is still inconclusive, serious anomalies of attachment that may result from maltreatment foreshadow poor outcomes (e.g., Cicchetti & Rizley, 1981; Oaensbauer & Harmon, 1982;Schneider-Rosen & Cicchetti, 1984). Study of the typical range of variation in attachment has yielded enough suggestive evidence to justify a volume called the Clinical Implications of Attachment (Belsky & Nezworski, in press). Although we must avoid uncritical attribution of later outcomes to early developmental influences (Rutter, 1981), the organizational construct of attachment is a potentially crucial determinant of later personality, both normal and disordered. Another potential determinant is the multidimensional construct of infant temperament, which also may influence later behavioral disorders (Bates, 1987;Chess & Thomas, 1984).Because these relationships are potentially relevant to later functioning, investigators have attempted to study the origins of We appreciate the efforts of many investigators who facilitated this report by sending us additional data from their studies. We also thank Loretta Rieser-Danner, who helped initiate the project. H. H.

The Effects of Daycare Reconsidered

“…Mothers of difficult boys returned to work earlier than mothers of difficult girls, and mothers with lower AFQT scores were more likely to work if the child was a boy; preexisting differences in maternal ability and infant temperament may explain some reported gender differences in daycare outcome (Rutter, 1981;Bronfenbrenner et al, 1984;Gamble and Zigler, 1986;Goldsmith and Alansky, 1987;Hoffman, 1989, Desai et al, 1989, and NICHD, 1997. )…”

Maternal sensory sensitivity, mother–infant 9-month interaction, infant attachment status: Predictors of mother–toddler interaction at 24 months

“…In contrast, Sroufe (1985) has argued that while temperament must be considered an important contributor to the mother-infant relationship, temperament per se is unimportant in explaining the construct of attachment itself. Others hold a more moderate position (e.g., Belsky & Rovine, 1987;Seifer, Schiller, Sameroff, Resnick, & Riordan, 1996;Stams, Juffer, & van IJzendoorn, 2002;Teti, Nakagawa, Das, & Wirth, 1991;Vaughn et al, 1992), and in a meta-analysis it was concluded that infant temperament variables are associated with attachment status to a similar degree as is maternal sensitivity (Goldsmith & Alansky, 1987). Together, these contentions suggest that temperament may be a confound in studies exploring relations between attachment status and variables of interest.…”

“…Because past research has shown that the infant temperament variable is associated with infant attachment status (Goldsmith & Alansky, 1987), dyadic synchrony (Vizziello, Ferrero, & Musicco, 2000), and with maternal behavioral sensitivity (e.g., Crockenberg & McClusky, 1986;Mangelsdorf, Gunnar, Kestenbaum, Lang, & Andreas, 1990) and sensory sensitivity (Donovan et al, 2005a;2005b), the temperament measures, ITQ and ECBQ, were initially entered as control variables in Step 1 of each regression analysis. The entry of the ECBQ measure was found to improve the significance of all the predictor variables, including sensory sensitivity for which its entry was necessary to obtain significance.…”

Assessing Attachment Security With the Attachment Q Sort: Meta‐Analytic Evidence for the Validity of the Observer AQS

“…Sufficient numbers of studies were available for testing the association between AQS and SSP (k 5 32), AQS and sensitive responsiveness (k 5 42), AQS and temperament (k 5 27), and AQS and socioemotional development (k 5 33). Earlier meta-analyses in the area of attachment were based on similar or smaller numbers of studies (e.g., Fox, Kimmerly, & Schafer, 1991;Goldsmith & Alansky, 1987;Van IJzendoorn, 1995).…”