Comparing Approaches to Conceptualizing and Analyzing Childhood Adversity

• In Event: Conceptualizing & Analyzing Childhood Adversity

Abstract

Research on childhood adversity and/or adverse childhood experiences has been steadily accumulating since the landmark CDC Kaiser ACE Study (Felitti et al., 1998). Surprisingly, across the more than 100,000 published articles containing the terms “childhood adversity” or “adverse childhood experiences,” there remains little consensus regarding how best to conceptualize and analyze childhood adversity.
Regarding conceptualization, a variety of dimension reduction techniques (e.g., factor analysis, latent class analysis), both data- and theory- driven, have yielded discrepant results. Of these diverse conceptualizations, threat and deprivation have emerged as distinct influences on biological and emotional functioning (Milojevich et al., 2019; Sumner et al., 2019). More recently, unpredictability has been posited as a third category of adverse childhood experiences (Ellis et al., 2022). Building on this literature, the first aim of this study was to evaluate a novel typology based on relational, environmental, and biological types of childhood adversities, as well as more traditional threat and deprivation conceptualizations (McLaughlin et al., 2014).
Paralleling the abundance of childhood adversity conceptualizations, researchers have posited numerous approaches to analyzing childhood adversity data. The second aim of this study was to evaluate a range of analytic approaches, including cumulative risk models, other-rated severity indices, and various weighting schemes. Drawing on a longitudinal study of child-caregiver dyads (N = 250; 50% female, 46% Latin-origin), this paper evaluated relations between childhood adversity and subsequent academic achievement and mental health outcomes across distinct conceptual and analytic models of childhood adversity.
To achieve our study goals, we first analyzed bivariate correlations of three novel typologies (i.e., relational, environmental, biological), threat, deprivation, and overall adversity (regardless of typology) exposure from birth to age 6 with later child academic achievement (Woodcock et al., 2001) and psychopathology (Achenbach & Rescorla, 2001) at age 8. We also compared five scoring approaches for each adversity variable including a simple sum of dichotomous adversities, a subjective (other-rated) overall score, a standardized composite (i.e., each adversity risk variable was standardized and then an overall average of these variables was calculated), a PCA-weighted composite, and a severity-weighted composite (i.e., weighted by the average level of risk for each adversity). Path models evaluated the predictive ability of each adversity variable as related to typology and analytic approach.
Our results indicated that adversity typologies (both our novel typologies as well as threat and deprivation) are differentially related and predictive of later child functioning. However, the different analytic approaches to scoring adversity were largely equivalent. That is, if one scored version of an adversity variable was related to a particular outcome, then the other scored versions were as well. Taken together, our results suggest that childhood adversity research is in need of more discussions and investigations focused on the utility of different conceptualizations of adversity rather than on different analytic approaches.

Authors

• Jessie Bridgewater, University of California - Riverside
  Presenting Author

• Sara R. Berzenski, California State University - Northridge

• AnnaMaria Boullion, University of California - Riverside

• Alexandra Sara Aringer, University of California - Riverside

• Tuppett M. Yates, University of California - Riverside