Abstract:

The sensitivity of caregivers plays a crucial role in developing secure attachment relationships. According to the attachment theory, children internalize their interactions with attachment figures as the Internal Working Models (IWMs), which guide their information processing and behavioral performance later in intimate relationships. As a basic structure of IWMs, the attachment script is conceptualized as a set of attachment expectations. However, little has been known about how young children represent those attachment interactions. Studies have found that attachment figures’ response patterns are related to young children’s support-giving expectations. However, few studies have directly explored how attachment figures’ responses influence young children’s attachment expectations. The present study used a real-time interaction task to examine how attachment figures’ response patterns affected children’s expectations of attachment figures’ support-giving behavior and willingness.
In both experiments, the Attachment Expectation Task (AET) was used to manipulate attachment figures’ response patterns, and the Attachment Story Completion Test (ASCT) was used to measure children’s attachment styles. Experiment 1 adopted a 2 (response pattern: response/non-response) × 2 (attachment style: secure/insecure) mixed design to investigate how attachment figures’ response pattern influenced children’s expectations of support-giving behavior and willingness under deterministic conditions. A total of 161 children 5~6 years of age were recruited (82 boys, M_age = 5.66 ± 0.29 years). Experiment 2 adopt a 3 (response pattern: 20% / 50% / 80%) × 2 (attachment style: secure/insecure) mixed design to investigate the effects of response pattern on children’s support-giving expectations under probabilistic response conditions. A total of 95 children 5~6 years of age participated (45 boys, M_age= 5.46 ± 0.29 years). Participants in both experiments were asked to finish the AET and the ASCT in two sessions.
The results of Experiment 1 showed that in new intimate relationships, secure children were more likely to expect attachment figures would [t (149) = 2.45, p = 0.015, Cohen’s d = 0.495] and were more willing [t (149) = 2.68, p = 0.008, Cohen’s d = 0.304] to provide support than insecure children. Meanwhile, children’s expectations of support-giving behavior and willingness increased under response condition [behavior expectation: t (149) = 6.79, p < 0.001, Cohen’s d = 0.444; willingness expectation: t (149) = 5.57, p < 0.001, Cohen’s d = 0.338] but decreased under non-response condition [behavior expectation: t (149) = -20.53, p < 0.001, Cohen’s d = -1.837; willingness expectation: t (149) = -14.65, p < 0.001, Cohen’s d = -1.307] (see Table 1 and Figure 1 & 2). Further, non-response condition had a higher effect on children’s expectation revision of behavior and willingness than response condition [change of behavior expectation, C-BE: M_non-response = 0.68, M_response= 0.39, V = 5121.5, p < 0.001; change of willingness expectation, W-BE: M_non-response = 1.27, M_response= 1.06, V = 6206.0, p < 0.001].
The results of Experiment 2 showed that children’s expectation of support-giving behavior and willingness significantly decreased under 20% [behavior expectation: t (88) = -9.52, p < 0.001, Cohen’s d = -1.016; willingness expectation: t (88) = -8.49, p < 0.001, Cohen’s d = -0.880] and 50% [behavior expectation: t (88) = -5.56, p < 0.001, Cohen’s d = -0.518; willingness expectation: t (88) = -4.50, p < 0.001, Cohen’s d = -0.427] response condition, however, under 80% response condition, only expectation of support-giving behavior decreased significantly [t (88) = -2.71, p = 0.049, Cohen’s d = -0.296] (see Table 2 and Figure 3 & 4). Meanwhile, the 20% response condition [C-BE: M_20%-response = -0.30, W-BE: M_20%-response = -0.67] had a higher effect on children’s expectation revision of behavior and willingness than 80% [C-BE: M_80%-response= 0.064, V = 2047.5, p < 0.001; W-BE: M_80%-response= 0.24, V = 2445, p = 0.003] and 50% condition [C-BE: M_50%-response = -0.06, V = 2215.0, p < 0.001; W-BE: M_50%-response = -0.11, V = 2218, p < 0.001].
The results indicate that attachment styles influence 5-and 6-year-olds’ initial support-giving expectations for new attachment figures, and they can revise these expectations based on attachment figures’ response patterns. The current study enriches the empirical evidence on how attachment figures’ response influences children’s attachment expectation revision in interpersonal interactions and extends our understanding of the organization and development of attachment representation. These findings also have important implications for the mechanism underlying secure attachment development in children.

Key words: support-giving expectation, attachment expectation revision, response pattern, attachment style, young children