Executive Functioning and Aggression in Early Childhood: Using Directionality and Severity Techniques to Disentangle Effects

• In Event: 1-152 - Executive Functioning and Relational Aggression/Victimization: Longitudinal Findings from Childhood to Adolescence

Integrative Statement

There have been mixed results on how executive functioning (EF) is related to relational and physical aggression longitudinally (McQuade et al., 2017; Rohlf et al., 2018). The goal of the current study (N = 110, 52.6% boys, M age = 44.04 months, SD = 4.23; 23.9% ethnic/racial minority) was to disentangle these effects in early childhood by using severity and directionality statistical methods (e.g., Park et al., 2005) to examine how EF predicts changes in the quantity of aggressive behavior and the likelihood of using physical compared to relational aggression.

EF was measured at Time 1 (T1) in the spring using teacher reports of the Childhood Executive Function Inventory (CHEXI), which is comprised of planning, working memory, inhibition, and regulation subscales (Thorell & Nyberg, 2008) and were reliable in this sample (Cronbach’s α’s from .87 - .96). Physical and relational aggression were observed at T1 and Time 3 (T3), one year after T1, using focal child sampling with continuous recording procedures (Ostrov & Keating, 2004). Trained research assistants collected eight, ten-minute observations per child over a two-month period. Severity was calculated by averaging standardized scores of relational and physical aggression [(ZRA + ZPA)/2]. Directionality was calculated by subtracting the standardized half difference of relational aggression from physical aggression [(ZPA – ZRA)/2], which identifies what distinguishes physical aggression and relational aggression.

Models were tested with Path Analysis techniques using the ML estimator and full information maximum likelihood estimation was used to accommodate missing data within Mplus (Muthén & Muthén, 1998-2017). To test model fit, the likelihood ratio χ2 was used in conjunction with the fit indices, RMSEA, SRMR, and CFI. RMSEA and SRMR ≤ .08 and CFI > .95 indicated adequate fit. Gender and age were statistically controlled (see Table 1 for bivariate correlations).

First, a model was tested where directionality and severity at T3 were regressed on the CHEXI composite and directionality and severity at T1. The model was a good fit to the data [χ2 (4) = 1.28, p = .86, CFI = 1.00, SRMR = .02, RMSEA = .00]. The standardized factor loadings are shown in Table 2 for both models. EF impairment predicted an increase in overall aggression over time. Second, a model was tested where directionality and severity at T3 were regressed on the CHEXI subscales and directionality and severity at T1. The model was a good fit to the data [χ2 (16) = 9.20, p = .90, CFI = 1.00, SRMR = .03, RMSEA = .00]. Impairments in working memory and inhibition predicted a propensity for increased physical aggression relative to relational aggression over time, whereas impairments in regulation predicted a propensity for increased relational aggression relative to physical aggression over time.

Overall, consistent with findings from Rohlf et al. (2018), results suggest that overall EF impairment is related to high levels of general aggressive behavior, whereas specific facets of EF impairment predict whether children use relational or physical aggression. Ways in which this work reconciles the equivocal findings in the field will be discussed.

Authors

• Kristin Perry, University at Buffalo, SUNY
  Presenting Author

• Jamie Ostrov, University of Buffalo
  Non-Presenting Author