Individual and culture-level components of survey response styles: A multi-level analysis using cultural models of selfhood

Variations in acquiescence and extremity pose substantial threats to the validity of cross-cultural research that relies on survey methods. Individual and cultural correlates of response styles when using 2 contrasting types of response mode were investigated, drawing on data from 55 cultural groups across 33 nations. Using 7 dimensions of self-other relatedness that have often been confounded within the broader distinction between independence and interdependence, our analysis yields more specific understandings of both individual- and culture-level variations in response style. When using a Likert-scale response format, acquiescence is strongest among individuals seeing themselves as similar to others, and where cultural models of selfhood favour harmony, similarity with others and receptiveness to influence. However, when using Schwartz's (2007) portrait-comparison response procedure, acquiescence is strongest among individuals seeing themselves as self-reliant but also connected to others, and where cultural models of selfhood favour self-reliance and self-consistency. Extreme responding varies less between the two types of response modes, and is most prevalent among individuals seeing themselves as self-reliant, and in cultures favouring self-reliance. As both types of response mode elicit distinctive styles of response, it remains important to estimate and control for style effects to ensure valid comparisons.