Model Specification Problem
It should be pointed out that associations between parental warmth and children’s adjustment may also be due to other factors such as common genetic endowment, neighborhood, socioeconomic status (SES), and the like (Cumberland-Li, Eisenberg, Champion, Gershoff, & Fabes, 2003). Warm parents may have more resources, may have better lives, and live in better places, and these related factors may affect children’s behavior. Parental rejection is also likely to be associated with a host of factors that are associated with antisocial behavior including abuse, other negative parenting practices (such as low supervision or inconsistent discipline), and low parent education. In our evaluation of studies, we will highlight findings from multivariate studies which control for these variables.
The Child Effects Problem
One of the most important classes of potentially confounding factors in studies of parental warmth is the “child effect.” We discussed child effects generally in Chapter 2. In this chapter, we raise the concern that a child’s temperament or behavior may elicit greater warmth from parents or their rejection, introducing the possibility that an estimated association between parental warmth and child antisociality is not causal (Rowe, 1986). If a child characteristic, such as callousness, evokes low levels of warmth in parents and also has an independent effect on violent behavior, estimates of associations between parental warmth and violent behavior in children may be spurious. Numerous authors have questioned the causal direction of the association between parenting and various outcomes in children due to this possibility (e.g., Andersson, 2005). To complicate matters, if the child’s character elicits bad parenting, but bad parenting really does affect the child’s subsequent behavior, the effect of parenting is not spurious but is at least partially indirect. Interpreting the many designs that address “child effects” is complicated, and we refer the reader to Savage (2014b) for a complete treatment of that issue. To protect against biased estimations, indicators of the child’s personality, behavior, and genetic endowments should be considered as control variables to distill the relationship between parental warmth/rejection and child antisocial behavior. Path analysis can also be used to estimate indirect effects.
Though some authors conclude that the “predominant causal flow is from parental rejection to delinquency” rather than the reverse (Simons et al., 1989, p. 306), others disagree. Many studies have demonstrated child effects on parental warmth and rejection (Rowe, 1986). For example, several studies have linked attention defi- cit/hyperactivity disorder (ADHD) to reduced parental warmth. Lifford, Harold, and Thapar (2008) report that ADHD symptoms affect mother-child rejection. Some authors, using genetically- informed designs, have concluded that child behavior evokes maternal criticism (Narusyte et al., 2011) and that children’s genetic composition substantially influences many aspects of parenting, including parental warmth (Reiss, 2008). For older children, the potential for “child effects” exerted by child misbehavior seems quite obvious. Delinquent and oppositional teenagers are certainly likely to find their parents less warm and accepting. Research confirms what we would all expect based on common sense, that “A self-perpetuating cycle of acting- out, punishment, and rejecting resulting in more acting-out has often been firmly established” (Landy & Peters, 2007, p. 17). The dynamic may be particularly potent when the police become involved (Stewart, Simons, Conger, & Scaramella, 2002).