Poster #98 - Trust as a safety conditioning process: extinction and the association with BIS and BAS sensitivity

• In Event: Poster Session 01

Abstract

According to attachment theory, attachment develops through experiences of sensitive and responsive caregiving (or lack thereof) that become internalized into internal working models (Bowlby, 1969). However, caregiving cannot fully explain individual differences in attachment (e.g. Verhage et al., 2016). In addition, it remains unclear how experiences of sensitive and responsive care become internalized into internal working models (Thompson, 2016). To help explain these gaps in our understanding of attachment development, Bosmans et al. (2020) formulated the learning theory of attachment (LTA). Accordingly, attachment development can be understood as a safety conditioning process through which caregivers become associated with the expectation that children’s distress will be regulated. The LTA proposes that the likelihood that caregivers become associated with the expectation that children’s distress will be regulated depends on the contingency (probability that caregivers successfully regulate children’s distress over caregiving interactions) and child factors that influence how easily the contingency affects attachment learning. In line with LTA, previous research has shown that the contingency predicts trust (Bosmans et al., 2019). In a trial-by-trial learning experiment we further tested he LTA, by examining (1) extinction in trust learning and (2) whether reward sensitivity (Behavioral Activation System; BAS) and punishment sensitivity (Behavioral Inhibition System; BIS) are associated with the rate of extinction. During the experiment, children opened doors to find their friend and an avatar gave advice on the door with the friend. The experiment consists of a learning phase and an extinction phase where the contingency or likelihood that avatar gives correct advice changes. Participants (N=89; 61.8% girls; 9-12 years) were randomly assigned to two conditions in which the contingency is manipulated: the 25%75% condition (correct advice on 25% of the learning phase trials and 75% of the extinction phase trials) and the 75%25% condition (correct advice on 75% of the learning phase trials and 25% correct advice the extinction phase trials). Sensitivity to punishment and reward were assessed using BIS/BAS-Scales for Children. With respect to extinction, we predict significant increases in trust in the 25%75% condition and significant decreases in trust in the 75%25% condition. For children with high levels of reward sensitivity, we predict faster extinction in 25%75% condition and slower extinction 75%25% condition. For children with high levels of punishment sensitivity, we predict faster extinction in the 75%25% condition and slower extinction in 25%75% condition. We found evidence for extinction in both conditions. Changes in contingency predicted significant changes in trust (F(1,80) = 83.08, p < .001) with increases in trust in the 25%75% condition (t(48) = -8.12, p < .001) and decreases in trust in the 75%25% condition (t(32) = 5.19, p < .001). Reward sensitivity was associated with faster extinction in the 25%75% condition (r= .33, p = .025) and slower extinction in the 75%25% condition (r = .37, p = .025). Punishment sensitivity was not associated with the extinction rate. The results provide evidence for extinction in trust learning and suggest that children’s sensitivity to reward (BAS) predicts the rate of extinction.

Authors

• Melisse Houbrechts, Katholieke Universiteit Leuven (KU Leuven)
  Presenting Author

• Lu Leng, KU Leuven

• Bram Vervliet, KU Leuven

• Rudi D'Hooge, KU Leuven

• Marinus H. van IJzendoorn, University College London (UCL)

• Marian J Bakermans-Kranenburg, ISPA Lisbon

• Chloë Finet, Katholieke Universiteit Leuven (KU Leuven)

• Guy Bosmans, Katholieke Universiteit Leuven (KU Leuven)