Keywords: Midpoint Response Style, Survey Measurement, Item Validity

Response styles can pose validity threats to survey responses. Midpoint response style (MRS), respondents’ tendency to choose the midpoint category in ordinal response scales (e.g., Likert Scale), can introduce sources of bias. Studies have shown that response styles vary across racial/ethnic groups. Particularly, for some East Asian populations (e.g., Korean, Chinese, and Japanese), studies have hypothesized MRS, and few have shown this empirically. However, the response styles of Asian Americans have not been examined, limiting understanding about these populations.

This study aims to investigate response patterns of Korean Americans to two life satisfaction questions: 1) 5-item Diener’s satisfaction with life scale and 2) a single-item global life satisfaction. For this, we conducted a randomized experiment in the Health and Well-being of Korean Americans survey (n=772). Here, half of the respondents were assigned to response scales with a midpoint, and the other half without a midpoint. We will compare response distributions between the two response scales and examine the relationship between the reported life satisfaction and well-established correlates of life satisfaction (e.g., education, income) which will indicate the potential bias resulting from MRS. Additionally, we will analyze the time respondents spent answering each question to examine whether providing a midpoint response option results in a reduced response time which may be an indicator of satisficing. Our findings will provide insights into survey design practices that more accurately capture responses from Korean Americans, contributing to the validity of cross-cultural research. This study will also expand the literature on MRS by exploring whether response scales without a midpoint mitigate MRS and offer a valid measurement.