Bowing raised attractiveness ratings only for Japanese raters in a six-country experiment

The second experiment settled the confound. Models appeared in four racial appearances, East Asian, European, South Asian and African, and two groups rated them, one Japanese and one from the United States ^[7]. Eight conditions in all ^[13]. The Japanese group rated bowing models as more attractive whatever the model looked like, and the US group showed little to no bowing effect ^[8].

The first round of experiments left an alternative open. Every model in it was East Asian in appearance ^[5], so a Japanese rater's higher score could have come from the bow, or from the pairing of a bow with a familiar-looking person. "The observed effect may have been driven by the culturally familiar combination within an East Asian group," the authors wrote ^[6].

In that first round, raters came from Japan, the United States, Brazil, India, China and Taiwan ^[5], and the increase appeared among the Japanese raters while the others showed little to no effect ^[4]. Chinese and Taiwanese raters therefore sit in the no-effect group, so whatever produces the response does not extend across East Asia as a region ^[14].

The hypothesis the team went in with was the universal one. "A downward head tilt increases perceived submissiveness, which could enhance perceived attractiveness across diverse cultures," they wrote, setting out the idea they aimed to test ^[3]. The six-country result did not support it ^[4].

A third set of experiments checked whether the two groups were using the rating scale in the same way. With East Asian models whose physical attractiveness was varied, both the Japanese and the US participants responded to those differences, and the bowing increase was substantially larger among the Japanese participants ^[9]. "These results suggest that bowing motions are interpreted consistently within each cultural population, regardless of racial and physical appearance," the researchers wrote ^[10].

The study does not identify which part of Japanese experience produces the response. Daily exposure to bowing as etiquette, its specific meaning in greeting and apology, and an association learned early in life would all predict the pattern reported here, because the only thing the design varies on the observer side is nationality ^[7][8]. The earlier work that started this line used portraits tilted to simulate a bow, and Japanese viewers rated those as more attractive too ^[1].

The phys.org account did not report sample sizes or numerical effect sizes; it calls the Japanese increase substantial and the other groups little to no effect ^[15]. The paper is by Takayuki Osugi and colleagues and appears in Japanese Psychological Research ^[12]; Jun Kawahara of Hokkaido University ran the study with Osugi, of Yamagata University ^[2].