Assessing Parent-Adolescent Substance Use Discussions Using the Continuous Assessment of Interpersonal Dynamics

Abstract

Assessing parent-child interactions is critical for understanding family dynamics, however tools available for capturing these dynamics are limited. The current study sought to examine the validity of the Continuous Assessment of Interpersonal Dynamics (CAID) for understanding the dynamics of parent-adolescent substance use discussions. Specifically, we examined how CAID parameters were related to indicators of parenting and substance use. Sixty-one parent-adolescent dyads (M adolescent age = 14.02, 57% female; M parent age = 46.40; 98% female) completed three 9-minute video-taped conflict, alcohol, and cannabis discussions as well as self-report measures of parenting (e.g., monitoring, psychological control) and substance use behaviors (e.g., intentions, use with parental permission). Interactions were coded using the CAID which provides continuous assessments of parent and adolescent warmth and dominance. Parental warmth, adolescent warmth, and dominance complementarity CAID parameters were positively associated with adaptive parenting and negatively associated with maladaptive parenting factors. Parental warmth in the cannabis discussion was negatively associated with the substance use and intentions factor. These findings support CAID as a reliable and valid assessment of interpersonal dynamics that characterize parent-adolescent substance use discussions and suggest that substance use conversations may be most effective when parents and adolescents act warmly throughout the discussion and exhibit dominance complementarity.

Acknowledgments

Research reported in this publication was supported by the National Institute of Alcohol Abuse and Alcoholism of the National Institutes of Health under award number AA028414 for Samuel N. Meisel.

Disclosure statement

The authors have no conflicts of interest to declare.

Notes

1 An alternative to correlations would be estimating regression coefficients. This approach would have the benefit of providing information on unique associations between CAID parameters and the latent variables from the ESEM. However, a power analysis conducted in G*Power (Faul et al., 2007), suggested that the current study had power to detect only medium sized increments, and this has the disadvantage of perhaps obscuring informative univariate associations.

2 Cohen’s d values of 0.20, 0.40, 0.63, and 0.87 correspond to small, medium, large, and very large effect sizes based on Funder and Ozer (2019).

3 Average differences between the mean adolescent warmth across the three discussions in the current study and mean child warmth values in Klahr et al. (2013) and Nilsen et al. (2015).