Is Typing in Roman Letters the Reason for the Low Proficiency of Japanese Language Typists?

Professor SHIMADA Hiroyuki and Research Fellow ASHITAKA Yuki from the Kobe University Graduate School of Maritime Sciences, as a result of researching the habits of skilled typists, have identified the reason that Japanese language typists have a lower proficiency level than English language typists. The need to type Roman letters requires typists to decompose hiragana characters into combinations of consonants and vowels. The paper was published on June 2, 2014, in the online version of the international journal Attention, Perception, & Psychophysics. The same journal published the duo’s paper. titled What Skilled Typists Don’t Know About the QWERTY Keyboard in February 2014, and it attracted wide public attention. In June’s online version, the February paper was the Editor’s Pick.

In Japan, the touch-typing rate even among college students is 15 per cent. This rate is much lower than the 85 percent of American college students. Professor Shimada and Research Fellow Ashitaka hypothesized that the reason for Japanese college students’ lower proficiency is not only the lack of official typing training but also because it is necessary that they type using Roman letters. Under this hypothesis, the duo conducted a typing training study using only a few keys. Unlike alphabets, typing hiragana does not allow typists to memorize a specific alphabet letter on a single key location on a one-to-one basis as each hiragana character is associated with a combination of a consonant and a vowel. As a result, when typing Roman letters corresponding to hiragana, used as visual stimuli, the typing difficulty varied based on this combination of consonants and vowels. For example, when typing the hiragana ka in Japanese, typists have to convert it into a pair of letters (i.e., k and a). In such cases, the study showed that vowels were easier to memorize while consonants were harder to remember. It also revealed that when typing words consisting of a pair of hiragana characters, requiring four keystrokes in Roman letters, typists had a hard time memorizing all four keystrokes. Instead, they memorized two keystrokes per hiragana character, or one consonant and vowel per combination, as a chunk, and then merge the two chunks. The study used a combination of mappings: half of the keys corresponding to the alphabet letters on a one-to-one basis (nonoverlapped mapping), while the remaining half of the keys did not (overlapped mapping). When typists decompose each hiragana character in Roman letters under these mappings, the study discovered that the development of typing skills is interfered with when nonoverlapped mapping conditions exist.

(Office of the Public Relations)