Facial emotion recognition and the development of delinquency: the mediating role of social bonds and self-control

ABSTRACT

Although facial emotion recognition has been identified as a potential factor influencing life-course-persistent delinquency, the mechanism underlying this relationship remains largely unknown. In this study, we explored two key variables—social bonds and self-control—as parallel mediators in the relationship between facial emotion recognition and general delinquency. An integrated design was employed using the models of social selection and social causation in criminology and the theory of attachment and conceptually parallel theories and research in psychology. The data came from a three-wave nationally representative longitudinal study of 1,423 children and adolescents carried out in the U.S. As expected, faces expressing anger and happiness were associated with general delinquency over 5 years, and both social bonds and self-control significantly mediated the associations. The accurate recognition of anger cues increased engagement in delinquent behavior by lowering both social bonds and self-control levels. In turn, the increased recognition of happy cues reduced delinquent behavior by enhancing both social bonds and self-control levels. Findings underscore the importance of identifying mediating pathways between facial emotion recognition and delinquency and suggest that these pathways are a potential target for behavioral intervention.

KEYWORDS:

• Facial emotion recognition
• social bonds
• self-control
• delinquency
• attachment theory
• social selection

Open Scholarship

This article has earned the Center for Open Science badge for Open Data. The data are openly accessible at https://www.icpsr.umich.edu/web/ICPSR/studies/8670.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We acknowledge that there are some potential conceptual differences between attachment constructs in control and attachment theories. Control theory emphasizes the impact of the social bond on crime and delinquent behavior, whereas attachment theory focuses on understanding healthy development and psychopathology, including delinquency, on the basis of parent-infant attachment relationships. Essential for control theory is the affectional bond that the child forms to his or her parents, which is thought to exercise strong control over the child’s behavior. However, an affectional bond is not the same as relationships in attachment theory given that an affectional bond is a stable characteristic of the child, whereas relationships are dyadic occurring through the reciprocal positive affectional sharing, which helps children to facilitate the optimal functioning of the attachment system. Consequently, an affectional bond is both a ‘necessary and sufficient’ condition for social bonds in control theory, whereas it is a ‘necessary but not a sufficient’ condition for attachment in attachment theory (e.g., Hoeve et al., 2012).

Despite these differences between attachment theory and control theory, we’d like to point out that both theories coalesce around three major interrelated assumptions which allow for integration: a) attachment is prosocial in nature, b) lack of attachment is associated with risk for crime and delinquent behavior, and c) attachment is generally depicted as a static representation of the development of delinquency over time (e.g., Giordano et al., 2010). Thus, it is safe to propose that both can be linked and integrated with each other to understand the underlying mechanisms of crime and delinquent behavior.

2 Although some scholars view social control and self-control as competing, even contradictory, theoretical perspectives, (Akers, 1991; Taylor, 2001; Gottfredson & Hirschi, 1990), they are generally considered distinct but nonetheless closely related to each other such that social control is conceived of as a normative socialization process affecting self-control development over the life-course (see Hay & Meldrum, 2016 for a review).

3 Given the age of the subjects (ages of 6-12) at the time of the study at Wave 1, there were no items suitable to determine an appropriate measure of general delinquency. The NSC contained several items used to measure general delinquency at Wave 3. However, the In-Home surveys asked in Waves 2 and 3 were not exactly the same. Thus, the questions used to measure the scale at each wave were not identical. Although using different instruments during different life stages may be appropriate to represent each developmental stage, doing so may raise questions about construct continuity. That is, if the researcher wants to document the change in the propensity to engage in delinquent behavior throughout the life-course, he or she must assume that the different items used to measure delinquent behavior at different ages do indeed measure the same underlying construct (Thornberry & Krohn, 2000). Given this, for the present study, a strategy of including a delinquent scale at Wave 2 into the model may be the best compromise.

4 The NSC employed the facial emotion task which was originally developed by Dashiell (1927) and adapted by both Ekman and Friesen (1976) and Ribordy et al. (1988).

5 A composite measure of the social bond contains items such as ‘children are very satisfied with, or interested in, their schoolwork.’ Although our delinquency measure is consistent with much delinquency research that employs a scale of delinquent acts containing both property and violent offenses (e.g., Matsueda, 1982; 1992; Elliott et al., 1985; Heimer & Coster, 1999), it includes items such as ‘getting into trouble with a teacher.’ Thus, if one is to study the relationship between a measure of the social bond and that of general delinquency, it may potentially lead them to be positively correlated through items overlapped (Drislane et al., 2014).

Although this study employs Hirschi’s (1969, 2004) view about the social bond in his social control theory, it may be that our single measure of attachment to parents precisely reflects the social bond in a manner comparable to the attachment questionnaire. As a result, in addition to a composite measure of the social bond, we chose to separately test a single measure of attachment to parents as a mediating variable and incorporate it into the model. Note that the results for this analysis were not shown in the tables.

6 Although not shown in the tables, the next model tested attachment to parents as a parallel mediator of self-control in the relationship between RFEE and general delinquency. This model also provided similar results. Both attachment to parents and self-control significantly mediated the association between RFEE (angry and happy faces only) and general delinquency. The indirect effect of angry faces through the two mediators increased general delinquency by 33 percent (indirect effect of .02/total effect of .06) from Wave 1 to Wave 2. The indirect effect of happy faces through attachment to parents (not self-control) decreased general delinquency by 40 percent across two-time points (indirect effect of −.02/total effect of −.05).