ISASE2016 Differences in Ratings of Different The Japanese Calligraphic Kanji Characters and The Japanese Font Shioko MUKAI *, Haruo HIBINO * and Shinichi KOYAMA * * Chiba University Yayoi-cho 1-33, Inage-ku, Chiba 263-8522, Japan Abstract: This study aimed to investigate the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font. We used 11 pictures consisting of 10 calligraphic styles by a master of calligraphy and a Japanese font. Participants (N = 316) were asked to provide their impressions of each picture using a 20-point Likert scale, which was based on the semantic differential method. Exploratory factor analysis showed there were two common factors in the ratings of the different Japanese calligraphic kanji characters and a font, namely Comfortableness and Expandability. The results indicated that the font was rated the highest on Comfortableness and the Japanese calligraphic kanji character was rated the highest on Expandability. A detailed exploration on the rating differences between each style should be investigated in the near future. Keywords: Japanese calligraphic kanji character, Font, Rating 1. INTRODUCTION Currently, there are 1650 types of Japanese fonts [1]. Japanese fonts are widely utilized in various fields (e.g., marketing) because of its simplicity and legibility. On the other hand, the recent ratings of Japanese calligraphic characters in logo designs have improved the advertising field [2]. This fact could imply that Japanese calligraphic characters might have different benefits from Japanese fonts. A recent study has reported differences in the shapes between Japanese calligraphic characters and fonts [3]. However, rating the impressions between them has not yet been firmly investigated. Because of this, it is unclear what kind of differences of impressions between them. Therefore, we investigated the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font in this study. In selecting writing styles for the present study, we referred to five kinds of the traditional Japanese calligraphic styles: Kaisho (regular script), (semi-cursive), (cursive), (clerical script), and Tensho (seal script). We then added applied them above, which have been used in advertising. However, it has been noted that Kaisho and Tensho are hardly used in packaging [2]. Thus, we adopted the applied styles instead of the traditional Kaisho and Tensho. As was the most frequently used font [2], we selected the DFP regular (Dyna font, DynaComware Corporation) as our Japanese font to compare with the Japanese calligraphic characters (Figure 1). In choosing the characters for this study, we were aware that many kanji s ideogrammatic nature may trigger personal meanings and feelings. Thus, we selected kanji characters that were not adjectives. When selecting a master of calligraphy to write in our selected styles, a recent has shown there were differences in characters form or shape between masters of calligraphy and ordinary people [3]. Characters used in this study were thus written by a calligrapher and not by ordinary people. In deciding which writing instrument should be used in the present study, in accordance with a fact that all styles or writing described before have been usually written with a brush in both classic calligraphic materials and nowadays. In choosing the pair of adjectives for this study, we sourced them from recent studies which related to research regarding the rating of advertising or products. In order to explore users' impressions of the characters' form or shape in detail, it was suggested that the semantic differential (SD) method [4] is one of the most effective procedure. Many researchers have frequently used this method to select rating scale descriptors and to perform exploratory factor analysis (EFA) [5-8]. We conducted a survey as a pilot study in order to investigate the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font, by using characters of the traditional style, its applied one, and a font. In this survey, we asked participants to rate each character s form or shape using the SD method. The data was then compared for each participant s response and analyzed using EFA to investigate the differences in impressions regarding
the characters' form or shape of different Japanese calligraphy and a font. We subsequently considered whether (and how) factor structure differences were present. 2. METHODS 2.1 Materials We prepared 11 pictures for this study, which comprised of 10 kinds of Japanese calligraphic characters and a Japanese font (i.e., DFP regular). Figure 1 shows all 10 styles and a font used in this study. Of the 10 Japanese calligraphic characters, three kinds were in the traditional style and seven kinds were its applied version. The former included,, and. The latter included applied styles of the following: the (based on Kaisho), the Blurred and s (based on ), the Horizontally-long style (based on ), the Vertically-long style (based on Tensho, as well as the Rounded and s (both based on ) These styles were named for this study. Each character was written by a master of calligraphy in reference to classic calligraphy examples [9]. Characters of the were written with gray ink, while all others were written with black ink. All characters were written with a brush. Each picture included 48 kanji characters. We adjusted all character sizes to appear as 62-point to ensure identical character sizes. Each character was placed in six rows consisting of eight characters horizontally in each column, and written at regular intervals. An 8 mm space was placed between the lines, (semi-cursive) (cursive) (clerical script) (based on Kaisho) (based on ) (based on ) (based on ) (based ) Horizontally-long style (based on ) Vertically-long style (based on Tensho) DFP regular (Dyna font) Figure 1: Schematics of the 11 Japanese calligraphic Kanji styles and font used in this study. and a 7 mm space was placed between the columns. These 11 pictures were edited as full-sized A4 images (1229 x 1653 pixels) with white backgrounds for individual presentation. The generated pictures were projected onto a 17-inch monitor in a randomized order to account for possible order effects. We chose 12 pairs of adjectives for this research from recent studies relating to the rating of advertising or products the following: two items relating to advertising, Pleasant/Unpleasant and Familiar/Unfamiliar [10]; nine items relating to products, Like/Dislike [11], Beautiful/Ugly, New/Old, Relaxed/Tense, and Clear/Muddy [12], Reliable/Unreliable, Comfortable/Uncomfortable, Stable/Unstable, Creative/Imitative, and Gorgeous/Plain [13]; and additionally, Legible/Illegible. The selected pictures and adjectives were projected in separate windows on a 17-inch monitor; they were also shown in a randomized order to account for possible order effects. 2.2 Participants and Procedures The participants were 316 adults (156 men, 160 women, average age = 49.78, SD = 10.87, aged 30-69). The survey was conducted using the Internet survey company Macromill (Macromill Inc, Tokyo, Japan). Participants were asked to rate each picture in terms of their impressions (see Table 1) on a 20-point Likert scale using the SD method. 3. RESULTS 3.1 Comparing each participant s response in each item The survey in this study reported the Japanese calligraphic kanji characters and font which were rated the highest in each adjective by participants as follows: DFP regular in 10 kinds of adjectives (i.e., Comfortable/Uncomfortable, Like/Dislike, Beautiful/Ugly, Stable/Unstable, Legible/Illegible, Familiar/Unfamiliar, Reliable/Unreliable, Clear/Muddy, Pleasant/Unpleasant, and Gorgeous/Plain); in two kinds of adjectives (i.e., New/Old and Creative/Imitative); and in an adjective Relaxed/Tense. We conducted Friedman tests on each participant s responses to compare raw scores between the various Japanese calligraphic kanji characters and a font. Friedman test for each of the 13 adjectives showed that there was a significant difference in all of the adjectives between the 11 Japanese calligraphic kanji
styles and font in each adjective (p <.001). Then we performed Wilcoxon test for each Japanese calligraphic kanji character and font which was rated the highest by participants (i.e., DFP regular,, and ) with all others. The results revealed that DFP regular was significantly rated the characters in seven adjectives: Comfortable/Uncomfortable (p <.001), Like/Dislike (p <.001), Beautiful/Ugly (p <.001), Stable/Unstable (p <.001), Legible/Illegible (p <.001), Familiar/Unfamiliar (p <.001), Reliable/Unreliable (p <.001). With respect to Gorgeous/Plain, DFP regular was significantly rated higher than (p <.05), (p <.001), (p <.001), Angular style (p <.001), s (p <.01), s (p <.01), Horizontally-long style (p <.001), Vertically-long style (p <.001), Rounded style (p <.001), s (p <.001). As for Clear/Muddy, DFP regular was significantly Table 1: Factor loadings for exprolatory factor analysis on the ratings for various Japanese calligraphic characters and a font. Item Comfortable /Uncomfortable Like/Dislike Beautiful/Ugly Stable/Unstable Legible/Illegible Familiar/Unfamiliar Reliable/Unreliable Clear/Muddy Pleasant/Unpleasant Gorgeous/Plain New/Old Creative/Imitative Relaxed/Tense Cumulative proportion Correlations between response dimentions Comfortableness Expandability Cronbach s alpha.91.91.89.85.81.80.76.72.68.53.16 -.20.03 1 -.36 *.93 Factor Comfortableness Expandability.03.09 -.05 -.11 -.04.13 -.27 -.24.10.16.61.54.45 48.90 57.16 1.53 Communality.83.83.79.73.65.66.65.58.47.31.40.34.20 Notes. Factor loadings >.40 are in boldface. We performed the EFA using major factor method with varimax rotation. Comfortableness and Expandability explained 57.16% of the variance in the evaluation of 10 kinds of Japanese calligraphic characters and a kind of font. * p <.05 rated higher than (p <.01), (p <.001), (p <.001), (p <.001), Blurred styles (p <.001), s (p <.001), Horizontally-long style (p <.001), Vertically-long style (p <.001), (p <.001), s (p <.001). Regarding Pleasant/Unpleasant, DFP regular was significantly rated higher than (p <.001), (p <.001), (p <.001), Angular style (p <.001), s (p <.001), s (p <.001), Horizontally-long style (p <.001), Vertically-long style (p <.001), Rounded style (p <.05), s (p <.001). With regard to New/Old, was significantly rated the characters and font (p <.001). As for Relaxed/Tense, was significantly rated the highest than all of the Japanese calligraphic kanji characters and font (p <.001). Finally, regarding Creative/Imitative, Angular style was significantly rated higher than (p <.001), (p <.001), (p <.001), Blurred styles (p <.001), s (p <.001), Horizontally-long style (p <.001), Vertically-long style (p <.001), (p <.001), but there was no significant difference from s (ns). 3.2 Factor structure First, we conducted EFA the principal factor solution without a rotation to investigate whether the 13 adjectives could explain differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font or not, and in order to find how many factors could be appropriate. Results showed that all values of the initial communality regarding the 13 items exceed.17. The resulting scree plot indicated that two factors might be appropriate in this study. Therefore, we assumed common two-factor rating different the Japanese calligraphic kanji characters and a font using all of the 13 adjectives. Next, we performed EFA using the principal factor solution with varimax rotation assuming two-factor to investigate the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font. The results indicated that all values of communality regarding the 13 items exceed.20. As shown in Table 1, EFA revealed a two-factor solution that accounted for 57.16% of the variance. We named each factor as the following: Factor 1 Comfortableness (α =.93) including Comfortable/Uncomfortable, Like/Dislike,
Average score of each facor 16 14 12 10 8 6 4 Factor 1: Comfortableness Factor 2: Expandability * 2 0 DFP regular Horizontally-long style Vertically-long style DFP regular Vertically-long style Horizontally-long style Figure 2: Average score of each factor of ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font. p <.001, * p <.05, Error bars denote the standard error of the mean (N = 316) Beautiful/Ugly, Stable/Unstable, Legible/Illegible, Familiar/Unfamiliar, Reliable/Unreliable, Clear/Muddy, Pleasant/Unpleasant, and Gorgeous/Plain; Factor 2 Expandability (α =.53) including New/Old, Creative/Imitative, and Relaxed/Tense. There was a significant negative positive correlation between Comfortableness and Expandability was found (r = -.36, p <.05). Finally, we conducted Friedman and Wilcoxon tests on each participant s responses to compare factor scores between the various Japanese calligraphic kanji characters and a font. As shown in Figure 2, the results revealed that DFP regular has the highest score in Comfortableness (p <.001). With respect to Expandability, has higher score than (p <.05) and all others of the Japanese calligraphic kanji characters and a font (p <.001). 4. DISCUSSION This study investigated the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and a Japanese font. We conducted a pilot survey using characters of three traditional styles, seven applied versions of traditional styles, and a Japanese font. In this survey, we ask participants to rate each character s form or shape using the SD method, compared for each participant s response and analyzed the data using EFA. We revealed that: the Japanese font was significantly rated the highest for 10 kinds of adjectives (i.e., Comfortable/Uncomfortable, Like/Dislike, Beautiful/Ugly, Stable/Unstable, Legible/Illegible, Familiar/Unfamiliar, Reliable/Unreliable, Clear/Muddy, Pleasant/Unpleasant, and Gorgeous/Plain); for an adjective New/Old and for Relaxed/Tense were significantly rated the highest. With regard Creative/Imitative, there was no Japanese calligraphic kanji character or font which was significantly rated the highest. We found two common factors in our ratings for various Japanese calligraphic kanji characters and a font, namely Comfortableness and Expandability (Table 1). We also revealed that the Japanese font was rated the character for the Comfortableness. With regard to the Expandability, of Japanese calligraphic kanji characters was rated the highest than all others including font (Figure 2). In conclusion, the present study s findings suggest that: the font used in this study might have generated a more comfortable impression than the Japanese calligraphic kanji characters; Rounded style of Japanese calligraphic kanji characters used in this study might have generated a more expansive impression than the font. As for this study s limitations, we could only use a
font to reduce the burden of the participants in the research. Future studies should compare other traditional fonts. Moreover, as a recent study has shown [3], there was a difference in the character height-width ratio between calligraphers characters and the font in our study. This may imply that differences in character height-width ratio could impact people s impression of various characters. A detailed study exploring the differences in ratings for the characters' form or shape of various Japanese calligraphic kanji characters and fonts should be conducted in the near future. ACKNOWLEDGMENTS This work was supported by the Grants for Research of Japan Academy of Advertising, the Kanto Region Committee 2014. REFFERENCES 1. H. Komiyama; Classification of Japanese Fonts. In; Editorial Department for Idea (Ed.) Sample Collection of Basic Japanese Typefaces: OpenType version. Seibundo shinkosha Inc, 2008. 2. Japan Package Design Association; The Japanese package design s yearbook. Tokyo: Rokuyosha, 1985-2013. 3. T. Saiki, A. Hayashi, and K. Sato; Comparisons of circumscribed quadrangle of handwritten and word-processed characters in vertically Japanese sentences: To realize new word-processed printing with feature of handwriting, Journal of Japan Ergonomics Society; 43(3), pp.111-116, 2007. 4. C. E. Osgood, G. J. Suci, and P. H. Tannebaum; The measurement of meaning. Urbana: University of Illusion Press, 1957. 5. M. Hirata; Aesthetic Impression of Hook Strokes in the Block Style Characters of Calligraphy, Bulletin of Japanese Society for the Science of Design; 61(4), pp.1-8, 2014. 6. I. M. Al-Harkan, and M. Z. Ramadan; Effects of pixel shape and color, and matrix pixel density of Arabic digital typeface on characters legibility, International Journal of Industrial Ergonomics; 35(7), pp.652-664, 2005. 7. P. W. Henderson, J. L. Giese, and J. A. Cote; Impression Management Using Typeface Design, Journal of Marketing; 68(4), pp.60-72, 2004. 8. S. H. Hsu, M. C. Chuang, and C. C. Chang; A semantic differential study of designers and users product form perception, International Journal of Industrial Ergonomics; 25(4), pp.375-391, 2000. 9. C. Fushimi; Kadokawa Chinese and Japanese Calligraphy Character Dictionary (219th ed.); Kadokawa Corporation, 1977. 10.Z. J. C. Sanabria; The role of familiarity and creativity in the generation of affective responses to advertising : proposal and evaluation of a pairing-congruity method, University of Tsukuba; 2012. 11.S. Mukai, and G. Miyazaki; Factorial Invariance In Evaluation of Impressions Between Fonts and Package Designs Using Structural Equation Modeling, Bulletin of Japanese Society for the Science of Design; 61(5), pp.107-110, 2015. 12. P. Dalton, C. Maute, A. Oshida, S. Hikichi, and Y. Izumi; The Use of Semantic Differential Scaling To Define The Multidimensional Representation of Odors, Journal of Sensory Studies; 23(4), pp.485-497, 2008. 13. K. A. Hsiao, and L. L. Chen; Fundamental dimensions of affective responses to product shapes, International Journal of Industrial Ergonomics; 36(6), pp.553-564, 2006.