Measurement Equivalence of Subjective Well-Being Scales under the Presence of Acquiescent Response Style for the Racially and Ethnically Diverse Older Population in the United States

TitleMeasurement Equivalence of Subjective Well-Being Scales under the Presence of Acquiescent Response Style for the Racially and Ethnically Diverse Older Population in the United States
Publication TypeJournal Article
Year of Publication2020
AuthorsLee, S, Vasquez, E, Ryan, LH, Smith, J
JournalSurvey Research Methods
Volume14
Issue4
Pagination417-437
ISSN Number1864-3361
KeywordsCross-cultural Differences, Measurement Equivalence, Measurement Invariance, Minority and Diverse Populations, Subjective well-being
Abstract

With valid assessment of subjective well-being (SWB) at the forefront of social science research, comparability of measurement scales designed to capture SWB across
racial/ethnic groups has been questioned. This study examines measurement properties of
well-established SWB scales and their comparability between older Hispanics and Whites in
the U.S.. Methods. We analyzed the Health and Retirement Study data in order to examine
measurement invariance of the satisfaction with life (SWL), positive affect (PAF) and purpose
in life (PIL) scales across non-Hispanic Whites, Hispanics interviewed in English and Hispanics interviewed in Spanish through multigroup confirmatory factor analysis and examined their
validity by linking their latent scores with covariates of SWB. Examinations of measurement
properties further considered acquiescence response style. Results. Strict, scalar and metric invariance was observed for SWL, PAF and PIL, respectively. However, when latent scores estimated from these invariance models were regressed on the validation measures, the relationship
was weaker for Hispanics than Whites, suggesting a lower level of validity for Hispanics than
Whites. A lower level of invariance was observed for respondents who acquiesced than their
counterpart, while the validity was not necessarily hampered by acquiescent response style.
Discussion. Our analysis suggests that the traditional measurement invariance test may not be
effective under the presence of acquiescent response style for SWB instruments that use the
Likert-type response scales. Research into cross-cultural measurement of SWB that considers
systematic difference in conceptualization of SWB as well as response styles may improve our
ability to understand SWB of the increasingly diverse population.

DOI10.18148/srm/2020.v14i4.7413
Citation Key11145