Flotsam, Jetsam, and Forward-Moving Vessels on the Sea of Well-Being

Abstract

I dispute the characterization of psychological aspects of well-being as fragmented and a landscape of confusion in need of an organizing conceptual framework. After 30+ years of research in this area, I see progress toward a multi-faceted, ever more differentiated understanding of what well-being is. This richness reflects decades of painstaking empirical inquiry. I also challenge the view that emotion is the overarching theme of well-being research. Missing from the target article was much-needed emphasis on empirical assessment tools. A growing problem in contemporary research is the proliferation of thin, poorly validated measures, which should concern all of the newly funded Emotion Networks. I conclude with a call for greater emphasis on major historical challenges that are undermining the well-being and health of many.

The core assertion of the article is that work on psychological aspects of well-being is fragmented — a landscape of confusion in need of an organizing conceptual framework. I dispute this characterization on multiple levels. What I see, after 30+ years of scientific research in this field, is undeniable progress toward a multi-faceted, ever more differentiated understanding of what human well-being is. Who would ever have expected otherwise? Importantly, this richness has evolved over time via careful, painstaking empirical inquiry. This article, in contrast, fails to engage with the actual science of well-being, and instead, focuses on primarily on terminology, although few terms are defined and fewer still linked with assessment tools. Instead, pronouncements are issued about what well-being is, drawing on prior pronouncements by others. None of these constitute a conceptual framework. That task requires theory and guiding ideas, which are notably missing.

Stated otherwise, I applaud the targeted objectives of the U24 Emotion Networks, but question whether adopting a single, overarching formulation of well-being is necessary and may even be an impediment in fostering high-quality work in targeted areas. A further key point is that knowledge of well-being as an antecedent (e.g., to health), or consequent (e.g., of socioeconomic inequality), or moderator/mediator (e.g., between SES and physiological risk) must emerge from scientific investigations built on psychometrically sound measurement tools.

Resolving the Nature of Well-Being: Declarative Decrees vs. Scientific Inquiry

The article noted previous initiatives to define what well-being is, such as a 2013 National Academies Subjective Well-Being Panel, and a 2018 National Institute of Health Roundtable on the topic emotional well-being. Both stated that well-being consists of three components: evaluative aspects (life satisfaction), emotional aspects (also referred to as experiential or hedonic well-being), and eudiamonic aspects (often framed as meaning and purpose, thereby neglecting most of the dimensions in my widely used model). Also embraced was Feller and colleagues’ (2018) view that emotional well-being (EWB) is the appropriate umbrella term for all of these psychological concepts. That input was a short article published in Public Health Reports written by a group of physicians seeking to bring emotional well-being into public health initiatives. Their objective was worthy, but their claim that EWB is the appropriate umbrella term was mistaken. Decades of prior science, done by psychologists, and before them, social indicators researchers (e.g., Andrews & Withey, 1976) made clear that emotion is not the overarching theme of what well-being is. Evaluative aspects involve judgments about well-being, which are not appropriately characterized as emotional experience; similarly, eudaimonic aspects involve reflections about one’s self and one’s life that are also not properly framed as emotional experience. In short, the meetings that preceded creation of the U24 Networks plus Feller et al. (2018), and now this article, have collectively misrepresented 50+ years of well-being research by claiming it is centrally about emotion. That is simply not true.

It is worth noting that the RFA soliciting applications for these U24 Emotion Networks mandated three areas of focus: (1) ontology and measurement of emotional well-being, (2) mechanistic research on the role of emotional well-being in health, and (3) prevention research. The irony is that the call to attend to core components of well-being (i.e., its ontology) was at odds with the a priori assertion that emotion is the overarching theme. Readers of Affective Science will likely understand this contradiction, given longstanding distinctions in psychology between cognition, emotion, motivation, and perception. Well-being is not exclusively situated in any one of those domains — rather, it traverses them all.

The Measurement vs. Mismeasurement of Well-Being

Sadly, thin, poorly constructed measures of well-being are proliferating in contemporary science. This problem needs to concern everyone, including those in the U24 Emotion Networks. A recently edited book, Measuring Well-Being (Lee et al., 2021), drew attention to these matters. Some in the volume offered recommendations about which measures to use going forward (VanderWeele, Trudel-Fitzgerald, Allin et al., 2021). Others disagreed with such recommendations (Ryff et al., 2021a) arguing that ultra-short assessments ignore extensive prior science documenting the complex, multi-faceted nature of well-being. For longer assessments, the recommendation was to use the measure of thriving (Su et al., 2014), but we noted its multiple problems, including a questionable conceptual foundation, inadequate evidence of validity and reliability, and highly redundant items. Following the rules of debate, those making the recommendations then responded with further defense of single-item assessments, arguing that one is better than none (VanderWeele, Trudel-Fitzgerald, Kubzansky, 2021). We countered that single-item assessments perpetuate a simplistic view of well-being that is out of touch with how the field has progressed over past decades (Ryff et al., 2021b). We also questioned blanket advocacy for measures devoid of attention to the substantive questions of interest. Sound measurement choices require thoughtful engagement with the relevant prior literature.

Other notable examples of mismeasurement were included in my recent critique of the positive psychology movement (Ryff, 2022a), which began with a distillation of the historical myopia of this endeavor — i.e., its profound failure to recognize decades of prior psychological research dealing with positive aspects of human functioning. Seligman’s efforts to lead research on well-being exemplify poorly constructed and poorly validated measures, which are at odds with the claim that positive psychology rests on rigorous science. One study (Seligman et al., 2005) sought to validate different interventions by showing their impact on well-being. Happiness was described as “scientifically unwieldy” and was “dissolved” into three distinct components: (a) positive emotion and pleasure, (b) engagement in life, and (c) meaning in life. This tripartite formulation was operationalized with an unpublished 20-item inventory having no information on validity or reliability. All analyses nonetheless focused on a single unwieldy assessment of happiness.

Next came PERMA (Butler & Kern, 2016), which added to the prior components (positive emotion, engagement, meaning) two new components: relationships and accomplishments. Again, none were defined, or grounded theoretically in anything. To measure PERMA, items were taken from other instruments and then transformed into the PERMA-Profiler using online samples of mostly well-educated adults. Given use of items from prior instruments, problems of redundant content plagued tests of convergent validity. Subsequent work has shown that PERMA and subjective well-being are indistinguishable (Goodman et al., 2018).

These examples illustrate growing problem of poorly constructed, poorly validated measures of well-being, an issue that needs attention because it is undermining the entire field of well-being. The exclusive focus on definitional issues in the target article (see Table 1) was thus notably incomplete — issues of quality control in the measurement of well-being must be part of the discussion.

Where the Science Needs to Go: Well-Being Under Assault

Recently, I have advocated for emphasis on societal ills as research imperatives (Ryff, 2022b), and more pointedly, on the intersecting catastrophes of COVID-19 and the plague of ever widening inequality (Ryff, 2022b). These macro-level happenings are undermining the capacities of many to live good and fulfilling lives; hence, growing deaths of despair due to suicide, drug, and alcohol addictions. Prior to the pandemic, the MIDUS (Midlife in the U.S.) national longitudinal study had become a prominent forum for investigating health inequalities, given its rich psychosocial, behavioral, and biological assessments (Kirsch et al., 2019). Our findings (www.midus.wisc.edu) linked educational and income disadvantage to compromised well-being of many varieties, greater psychological distress, poorer health behaviors, higher stress exposures, elevated biological risk factors, and greater morbidity and earlier mortality.

A unique feature of MIDUS was recruitment of two national samples situated on either side of the Great Recession that began in 2008. Over the time period covered by the two samples, educational attainment in the USA improved, but the post-Recession (refresher) sample nonetheless reported less household income (after adjusting for inflation), lower financial stability, worse health (multiple indicators), and lower well-being (multiple indicators) than the pre-Recession baseline sample. Further work showed more compromised mental health among those of lower socioeconomic standing in the later refresher sample (Goldman et al., 2018).

Here I wish to commend the Emot-Econ U24 Network focused on the economic burden of disease and its impact on the emotional well-being of patients and their families. This work powerfully enriches prior research on emotional well-being by targeting the financial toxicity associated with health care — i.e., how out of pocket costs, missed work, lost income, medical debt or bankruptcy contribute to emotions of worry, distress, anxiety, and isolation. In the face of these very real life-challenges confronted by growing numbers of Americans, it makes no sense to legislate what measures of well-being (evaluative, hedonic, eudaimonic) researchers in the Emot-Econ Network should be using. Rather, they should be encouraged to select measures that make the most sense for the questions they are investigating.

I close by noting that MIDUS may be helpful in making measurement decisions across the six U24 Networks. Why? Because MIDUS is a public-use dataset that has tracked over 11,000 individuals across the decades of adult life, including augmented samples of African-Americans. Importantly, MIDUS includes the full panoply of well-being measures: life satisfaction (overall, domain-specific), positive and negative affect (assessed with multiple instruments), and my dimensions of psychological well-being. In addition, the study measures optimism, sense of control, personality traits, coping strategies, and social relational ties (parent/child, adult partners, friends). To date, more than 1,700 publications have been generated from MIDUS; 400+ of these have been about psychological well-being — how it is influenced by early life adversity, adult stressors, inequality and discrimination; how it is linked with biological risk factors, affective neuroscience, morbidity, and mortality; and how well-being and other individual-difference moderate or mediate various health outcomes. My point is this: as researchers across the U24 networks grapple with boundary conditions between different measures of well-being and other psychosocial constructs, MIDUS offers a user-friendly, well-documented source of data based on nationally representative samples that provide information on empirical associations among all of the above factors and more.