Attachment Categories: The Strange Situation
Most developmental studies of attachment employ the Strange Situation paradigm where dyads of children and their caregivers (usually the mother) are observed in a laboratory (see, e.g., DeKlyen & Speltz, 2004). The lab is furnished as a playroom, and a similar procedure is followed across studies. The procedure follows phases where the child is exposed to gradually escalating stress, so the observer can take note of changes in behavior toward the caregiver (Lamb, Thompson, Gardner, & Charnov, 1985). The child’s behavior is observed when the mother is present, when a stranger is present, when the child is alone, and when the mother returns. The attachment relationship is inferred from the observable proximityseeking behaviors of the child (Ainsworth, 1973).
Preschool children are normally subject to conflicting impulses. They are drawn away from their mothers by the interesting, novel environmental stimuli, while at the same time drawn to her for safety (Ainsworth, 1973). In young children, the secure attachment category (denoted as category B) is assigned when a child is distressed when the mother leaves, and positive and happy when the mother returns. The securely attached child normally demonstrates stranger anxiety and is friendly to the stranger when the mother is present, but avoidant of the stranger when the mother is not in the room. A securely attached child is easily soothed by the attachment figure, is glad to see her return to the playroom and will seek proximity to her when she does.
Other patterns of responses are categorized as “insecure” with subcategories of “avoidant” (category A), or “resistant” (category C) (also referred to as “ambivalent”), and “disorganized” (category D). Avoidant infants do not show distress when their mothers leave the playroom and often ignore her return. The avoidant preschooler may even actively avoid contact with his or her mother (Fox & Hane, 2008). The avoidant infant or toddler is less aware of adult changes, expects adults to be unresponsive, and is not likely to cry (Howes, Matheson, & Hamilton, 1994). Resistant infants will employ proximity-seeking behaviors but will also resist their caregiver’s attempts to calm their distress (Fox & Hane, 2008). Finally, infants categorized as “disorganized” will display contradictory emotions and may appear confused or apprehensive (Fox & Hane, 2008).
Eleven studies report associations between categories of attachment security and physical aggression; six of these used the Strange Situation to determine the categories. Here we focus on comparisons of “securely attached” to other categories. In some cases, the authors provided significance tests; in others, they provided means, standard deviations, and sample sizes which allowed us to estimate test ratios which we used to complete the table. Six of the eleven studies reported that securely attached individuals (mostly young children in these studies) were less physically aggressive and violent than the other children. Four of these studies reported a PoC that was statistically significant (14 out of 31 total comparisons).
Studies that present unexpected findings, however, are not inconsequential in number, and it is worth looking at them in more detail. Studies frequently reported a comparison of the “secure” group (B) to the other groups. Many of the findings summarized in Table 6.1 that are in the wrong direction compare securely attached children to group C, insecure (resistant) children. In fact, in all three studies that report such comparisons, the resistant children have lower physical aggression than the securely attached children (Bates & Bales, 1988; Burgess et al., 2003; Erickson, Sroufe, & Egeland, 1985). By contrast, when securely attached children are compared to group A, the avoidant-insecure group, the coefficients were mostly in the right direction (3 out of 4 coefficients, 1 was statistically significant). This strongly suggests that attachment insecurity has adverse effects on violence for some children but not others, depending on the form this insecurity takes. Of course, we might speculate that attachment insecurity taking the form of avoidant behavior is more likely to be caused by parental insensitivity, which we would hypothesize could result in very bad consequences (see also the section on neglect in Chapter 8), but it may also be the case that the child’s temperament is strongly involved.
The findings in this category differed by gender. In 4 out of 5 studies of females, the findings were in the wrong direction or null. The findings for male subjects were more likely to be in the expected direction.
In summary, it does not appear that securely attached children are less physically aggressive than all categories of insecurely attached children. The findings suggest that while securely attached boys are commonly less aggressive than avoidant boys, “resistant” children may be suppressing aggression still more, though there are only a few studies reporting findings relevant to this point. Studies that use other measures to identify “secure” vs. “insecure” in broad categories tend to report that securely attached are less physically aggressive than insecurely attached children.
We conclude that using attachment categories as an indicator of attachment may not be the best way to predict violent behavior.