Validity of Importance Weighting in Subjective Well-Being Research

Abstract

The practice of giving certain life domains (such as health, family life) more weight in calculating an overall score, known as importance weighting, has been a subject of debate in subjective well-being (SWB) research for decades. In this paper, we present evidence by analyzing data from 513 Chinese adults to caution readers that findings of importance weighting in the SWB studies should be interpreted carefully. Given the many unsettled issues in assessing importance weighting, findings are often not clear-cut or definitive and overgeneralization can mislead our understanding of well-being. Future research on the topic of importance weighting should acknowledge study assumptions and limitations to avoid potential overgeneralization of findings.

Are all Life Domains Created Equal? Domain Importance Weighting in Subjective Well-Being Research

“…The main issue with the empirical findings used to argue against domain importance weighting in SWB studies is that the weighting approaches explored have been extremely limited (Hsieh & Kenagy, 2014;Hsieh & Li, 2020). Most studies against domain importance weighting presented results based on multiplicative scores or a proportional function of domain importance rating as if domain importance weighting is synonymous with multiplicative scores or a proportional function of domain importance rating (Hsieh, 2015).…”

“…For example, Trauer and Mackinnon's (2001) argued against domain importance weighting by presenting empirical evidence that was based only on multiplicative scores. Similarly, Rohrer and Schmukle (2018) and Wu and Yao (2006b) opposed domain importance weighting by presenting empirical evidence based on only a linear (or proportional) function of domain importance (Hsieh, 2013;Hsieh & Li, 2020).…”

“…In the SWB literature on the topic of domain importance weighting, there appears to be a tendency of treating evidence that failed to reject the null hypothesis as evidence supporting the null hypothesis (Hsieh, 2013(Hsieh, , 2018Hsieh & Li, 2020). Specifically, many studies explored domain importance weighting (using the principle of multiplicative scores) and did not find evidence supporting domain importance weighting (Philip et al, 2009;Russell & Hubley, 2005;Wu & Yao, 2006a, 2006b.…”

“…One direction that can help advance our understanding of the topic is the exploration of weighting methods. To date, the exploration of weighting has been limited to multiplicative scores or a linear or proportional function of domain importance (e.g., Rohrer & Schmukle, 2018;Hsieh, 2013;Hsieh & Li, 2020). Since the function of domain importance may not necessarily be linear (e.g., Campbell et al, 1976: Hsieh, 2004, Hsieh & Kenagy, 2014Hsieh & Li, 2020), other forms of domain importance weighting should be explored.…”

Testing importance weighting: Lessons from the quality of life literature

“…Although there is general consensus on the significance of domain importance (Hsieh, 2012), the debate regarding the integration of domain importance into measures of SWB, also known as domain importance weighting, persists (e.g. Campbell et al., 1976; Hsieh, 2003, 2004; Hsieh & Li, 2020; Hsieh et al., 2020; Mastekaasa, 1984; Russell et al., 2006; Wu, 2008a, 2008b). Although there are different approaches to domain importance weighting (Hsieh, 2012), the focus of the current study is on the approach that involves assigning distinct domain importance weights to different respondents, also known as individually weighted‐average models, or IWAMs (Marsh & Scalas, 2018).…”