Methodological implications of the affect revolution: A 35-year review of emotion regulation assessment in children

Abstract

This investigation analyzed the methods used over the past 35 years to study emotion regulation (ER) in children. Articles published from 1975 through 2010 were identified in 42 child clinical, developmental, and emotion psychology journals. Overall, 61.1% of published ER articles relied on one method and 23.6% used two methods. Analyses revealed (a) 82.8% of published ER research occurring within the past decade; (b) higher rates of observational methods with infant and toddler/preschool samples, but more use of self-report methodology with middle childhood and adolescent samples; (c) a longer history of published ER research with samples of infants to 5-year-olds, including the use of more longitudinal design, compared with older samples; and (d) a positive association between journal impact ratings and the use of physiological and observational measurement. Review of the measurement tools used to capture ER revealed great diversity in how emotion processes are understood and evaluated.

Introduction

The regulation of emotion involves the management of diverse systems, including physiological arousal, facial and behavioral expressions, motivation, inter- and intrapersonal goals, and cognitive evaluations (Thompson, 1994). Due to this diversity in systems, it has been a challenge to find a common definition for emotion regulation (ER) and then develop and implement optimal assessment methods (Cole, Martin, & Dennis, 2004; Langlois, 2004). Measurement of ER is directly linked to theoretical issues regarding the definition and nature of the construct. Furthermore, debate has ensued regarding the conceptual and methodological dissociation of emotion from ER (see Child Development, Vol. 75, 2004), with some purporting a two-factor solution (emotion as preceding ER) (e.g., Cole et al., 2004) and others arguing for the inseparability of emotion from ER (e.g., Campos, Frankel, & Camras, 2004). As such, for this review, the following commonly accepted definition of ER is adopted: “the extrinsic and intrinsic processes responsible for monitoring, evaluating, and modifying emotional reactions, especially their intensive and temporal features, to accomplish one’s goals” (Thompson, 1994, pp. 27–28).

The study of ER in children is a relatively new field, with the past two decades described as an affect revolution (Fischer & Tangney, 1995) that has transformed how emotion processes have been conceptualized. However, the progress in assessment of these constructs lags behind and, as a result, limits the questions that can be addressed and makes the interpretation of mixed findings a daunting task. The overarching aim of this research was to provide documentation of how ER has been studied in children over the past 35 years because this information has the potential to provide insights concerning the strengths and limitations of the field and to offer guidance for future directions.

Even with a cleverly devised methodological system, the challenging nature of scientific inquiry into ER must be acknowledged. As Larsen and Prizmic-Larsen (2006) stated, “Emotion researchers need to keep clearly in mind that constructs are never purely measured. Rather, all measures are construct–method composites. … The theoretical meaning of a construct is given, in part, by the methods used to measure it” (p. 342). Because of the complexity of the constructs involved in ER and equifinality of dynamic systems, multilevel and multimethod approaches to the scientific inquiry of ER in both normative and atypical child samples have been advocated. Emotion is best defined as a multifaceted construct inferred from multiple indicators, and by focusing and isolating one aspect of emotion, an incomplete picture of the emotion system is likely to emerge (Larsen and Prizmic-Larsen, 2006). However, measurement in this system is complicated because the converging multiple response systems are loosely and imperfectly coupled, complexly interactive, and mutually influential. Thus, ER may be understood through accessing a variety of different processes from central and peripheral nervous system processes, sociocultural influences, and biological and behavioral contexts (Thompson, Lewis, & Calkins, 2008). The current study sought to systematically analyze the ways in which ER researchers have grappled with investigating this construct.

Researchers interested in studying children’s ER measure this construct primarily in four ways: self-report, other informant (parent, teacher, or peer), observational, and physiological–biological indicators (for reviews, see Morris, Robinson, & Eisenberg, 2006; Zeman, Klimes-Dougan, Cassano, & Adrian, 2007). Self-report methodologies provide an important assessment of ER even when the reporter may be a young child (Durbin, 2010). Individuals are the only ones able to assess and integrate a variety of levels of information about their own emotions (e.g., meld information about cognitions and physiological symptoms such as heart rate). However, self-report methods are limited by the demands they place on children such that they require children to be aware of emotions, have the ability to monitor emotions, recall their emotional experience retrospectively, and then communicate this information effectively to others (Zeman et al., 2007). The recent invention and use of electronic communication sources (e.g., electronic diaries) offers a partial solution to some of these issues, yet their incremental validity over other approaches is still questionable (Suveg, Payne, Thomassin, & Jacob, 2010).

Other reporters (e.g., parents, teachers) are also often enlisted to measure components of emotion because they provide an opportunity to understand children’s ER in diverse social settings. Although this information is considered to have better reliability and validity than self-reports (Morris et al., 2006), research has revealed that there are biases and other processes (e.g., parental psychopathology) that can affect ratings of children’s behavior (Fergusson, Lynskey, & Horwood, 1993). Although rarely explored, peer methods of nominations and emotion-based sociometrics could be a valuable method to explore the connection between ER and social processes (Perry-Parrish and Zeman, 2011).

Observational methods are often considered the “gold standard” in developmental research (Cummings, Davies, & Campbell, 2000). Body gestures, tones of voice, and facial expressions relay important information about emotional experiences, particularly emotional expression. In addition, observational coding systems can be developed in a flexible manner to capture the precipitants, reactions, and consequences of emotional expressions. However, an observational approach captures only some of the components of ER and does not tap into the associated internal unobservable processes. Furthermore, as children develop, so does their ability to dissemble emotional displays, thereby affecting the validity of this method (Saarni, 1984).

Neurophysiological systems are critically involved in the experience, production, and regulation of emotion (Curtis and Cicchetti, 2007). Recent evidence indicates that key neurocircuitry relevant to emotion processing undergoes several periods of reorganization well into adolescence (Giedd et al., 1999). Due to recent technological advances, a number of methods have evaluated different biological emotion systems. To investigate the activity of the cortico-limbic circuitry that is involved in emotion processing and regulation (LeDoux, 1996), electroencephalogram and neuroimaging techniques have been used to provide insight into these processes. The hypothalamic–pituitary–adrenocortical (HPA) axis is implicated in ER efforts, and studies with children have used cortisol assays to understand the impact of HPA disruption on ER efforts. Activation of the parasympathetic system, which is associated with ER abilities, has been measured by the vagal tone index (Porges, 1991).

Along with the increasing advances in the study of children’s ER, an acute awareness of the complexity of the construct has arisen; as such, achieving clarity in what constitutes ER and understanding the dynamic interplay between the developmentally important constituents has been slow to emerge (Thompson et al., 2008). As such, the overarching goal of this review was to evaluate the methodological state of the field of ER research using a multipronged approach. First, we sought to provide a description of the field with respect to validating the claim that we are indeed in the middle of an affect revolution. Second, we examined the number and type of measurement strategies used in the study of ER in children. Third, we investigated whether the use of certain methodologies to study ER differed as a function of developmental status. Fourth, we explored whether inquiries into typical versus atypical ER processes used different methodologies. Fifth, we investigated whether specific methodologies provided differential impacts on the understanding of ER in children. Sixth, we compiled a summary of the measures used to evaluate ER to provide researchers with a guide to the available instruments and methods as well as to illuminate gaps in our assessment toolbox. To meet these goals, 42 child clinical, developmental, and emotion psychology journals were reviewed to evaluate the methods used and their contributions to the study of childhood ER throughout a 35-year period.