@article{fa6a9098be434fbdba2ebd332af00e7c,
title = "Exploring cultural variation in the emotional expressivity of online drawings",
abstract = "Extensive research points to cross-cultural differences in emotional expressivity and the use of context in communication. This study explored these ideas through digital, online, drawings produced using Google's Quick Draw (N = 4869). The selected pictures were of fish and had been drawn by individuals from across six nations: UK, USA, Australia (individualist), Japan, Saudi Arabia and the United Arab Emirates (collectivist). Participants from individualist societies produced images expressing emotion (e.g. smiling or frowning fish) more frequently than their collectivist counterparts. Similarly, participants from individualist nations were significantly more likely to include contextualising elements within their drawings (e.g. seaweed, bubbles etc.). The results support previous work on emotional expression across cultures and research in the area of high and low context communication. This study extends these ideas into the area of computer-based drawing, suggesting Google's Quick Draw represents a useful resource for exploring emotional and cultural variation through the medium of online drawings.",
keywords = "Culture, Digital drawing, Emotion, Online, Quick Draw",
author = "Justin Thomas and Aamna Al-Shehhi and Ian Grey and Tai Broach",
note = "Funding Information: There were several significant differences between the IND and COL nations. Firstly, drawings from IND nations more frequently depicted emotionally expressive fish. This finding supports the study's emotional expressivity hypothesis and is in line with previous cross-cultural studies exploring the expression of emotion across cultures (Matsumoto, 1991; Keltner, 1995; Stephan et al., 1998; Tsai, Louie, et al., 2007; Tsai, Miao, et al., 2007; Cordaro et al., 2017). These findings also fit well with the “affective preference hypothesis”, which suggests that Western (IND) and Eastern(COL) nations differ in their preferences for high (exited, happy) and low (content, calm) arousal positive emotions, respectively (Tsai, Miao, et al., 2007). It is arguably easier to draw and detect high arousal emotions relative to low arousal ones, and this might also explain the apparent preponderance of IND fish deemed to be expressing emotion. This finding is further supported by the observation that IND and COL nations did not differ regarding the frequency with which they drew mouths. Mouths curving up (happy) or down (sad) are commonly used to depict the expression of emotion in drawings (see Fig. 1: Fishes 1 and 4). Interestingly, participants from COL nations were more likely to include the fish's eye in their depictions. This emphasis on eyes may reflect the importance of eyes in societies (KSA and UAE) where female veiling (burka, lithma, hijab) has been and remains a relatively common practice. In a study of face recognition, comparing the US and Emirati citizens, the Arab/Emirati participants were far superior in recognising faces based on eyes. The authors attribute this finding to the “Hijab effect”, that is the widespread use of veiling and head covering results in better non-peripheral facial recognition(Wang et al., 2015). Eyes are also often used to provide non-verbal cues during communication, and the higher frequency of their inclusion may reflect the HC communication style of collectivist nations (Hall, 1976; Okabe, 1983).The present study has several important limitations. Firstly, the use of only one image type (Fish) precludes ruling out the specificity of the present findings. For example, would images of cats or houses show similar cross-national patterns of affective expressivity and context inclusion? A similar limitation is our reliance on a relatively small number of nations, the observed findings could be an artefact of other national idiosyncrasies beyond the hypothesized cultural factors. Future studies should include a more varied range of pictures and expand the number of nations included. However, the present study was a preliminary attempt at exploring a novel data source and drawing medium. As such, including multiple categories of images (cat, dog, camel) was beyond the study's scope and resources. Furthermore, given that previous studies have supported the “ideal affect” and “affective expressivity” hypotheses in drawings of humans (Tsai, Louie, et al., 2007; Gernhardt et al., 2013, 2015) it seems highly unlikely that the present findings are restricted to fish. Future studies could indeed greatly increase the sample size, target image type and the international scope of the analysis. Beyond that, training computer algorithms to detect emotional expression in images (deep learning) would then permit automated analyses, allowing large amounts of data to be processed quickly. Publisher Copyright: {\textcopyright} 2020 The Authors",
year = "2020",
month = aug,
day = "1",
doi = "10.1016/j.chbr.2020.100002",
language = "British English",
volume = "2",
journal = "Computers in Human Behavior Reports",
issn = "2451-9588",
publisher = "Elsevier B.V.",
}