Diverse pathways to positive and negative affect in adulthood and later life: An integrative approach using recursive partitioning.

Abstract

Recursive partitioning is an analytic technique that is useful for identifying complex combinations of conditions that predict particular outcomes as well as for delineating multiple subgroup differences in how such factors work together. As such, the methodology is well suited to multidisciplinary, life course inquiry in which the goal is to integrate many interacting influences and understand subgroup variation. The authors conducted recursive partitioning analyses on a previously published study (D. K. Mroczek & C. M. Kolarz, 1998) that investigated life course profiles of positive and negative affect and incorporated various top-down (personality traits) and bottom-up (sociodemographic statuses, contextual influences) influences. The new analyses reveal multiway, nonlinear interactions among these variables in predicting affective experience and, importantly, life course differences in how these various factors combine. Included are details of how recursive partitioning trees are generated as well as descriptions of the software packages available for using such techniques. Overall, the methodology offers tractable strategies for discerning meaningful patterns in highly complex data sets.

Healthy aging is associated with increased neural processing of positive valence but attenuated processing of emotional arousal: an fMRI study

“…This is the first study to our knowledge to demonstrate reduced cortical reactivity to emotional arousal in healthy aging. Aging has been associated with the experience of fewer negative emotions (Gruenewald et al, 2008;Mroczek and Kolarz, 1998), as well as reduced neural processing of negative compared with positive emotional stimuli (Addis et al, 2010;Leclerc and Kensinger, 2008;Mather et al, 2004). Considering that negative emotional stimuli tend to be rated as more arousing than positive stimuli (Lang et al, 2008), and tend also to result in greater BOLD signal increases in emotional and visual brain regions than neutral or positive images (e.g., Lane et al, 1997;Mourão-Miranda et al, 2003), it is possible, based on our results, that an under-responsiveness in older adults to the effects of arousal may partly mediate the attenuation in their response to negative emotional stimuli.…”

Terminal decline in well-being: The role of multi-indicator constellations of physical health and psychosocial correlates.

“…We selected three sets of variables known to be associated with individual differences in mortality hazards and well-being (see Vaupel, 2010), including socio-demographics, health and burden factors, and psychosocial factors. We are building on and expanding two sets of earlier reports: Studies that used recursive partitioning to identify cross-sectional correlates of well-being in adulthood (Gruenewald et al, 2008;Wallace et al, 2002) and studies that used traditional growth models to identify correlates of terminal decline in well-being (Gerstorf et al, 2016). Specifically, we apply SEM Trees and SEM Forests to well-being data obtained in the last years of life so as to identify multi-way (nonlinear) constellations of socio-demographic, physical health and burden, and psychosocial factors associated with differences in terminal decline.…”

“…First, Wallace, Bergeman, and Maxwell (2002) applied recursive partitioning analyses to cross-sectional data so as to investigate associations of protective factors (perceived control, social support coping, and physical health) with well-being in later life. Second, Gruenewald, Mroczek, Ryff, and Singer (2008) applied decision trees to cross-sectional data from the MIDUS study to explore social, psychological, behavioral, and socio-demographic correlates of negative affect and positive affect in adulthood and later life.…”

Data mining for response shift patterns in multiple sclerosis patients using recursive partitioning tree analysis

“…These tools have been applied in a number of fields, including marketing research [1], economic prediction [2], epidemiologic research [3], and health services research [4], but their application to quality-of-life (QOL) research is only recent [5][6][7][8][9]. Such methods present intriguing tools for hypothesis generation in evaluating the role of qualitative processes in quantitative assessment.…”