From: owner-3dui@hitl.washington.edu on behalf of Shumin Zhai [zhai@almaden.ibm.com] Sent: Sunday, October 07, 2001 3:18 PM To: Ivan Poupyrev Cc: 3d-ui@hitl.washington.edu Subject: Re: text/number input in VEs I felt compelled to add two cents here. QWERTY was not designed to slow typists down. But it was not a human factors success through user testing either. The main rationale of the QWERTY design of Christopher L. Sholes, Carlos Glidden, and Samuel W. Soule in 1868 was to minimize mechanical jamming. Accidentally, this design also facilitates the frequent alternation of the left and right hand, which is a key factor to rapid touch-typing with two hands. There are other factors that are important to rapid touch typing, including the use of the home raw, finger assignments, and sequencing. Later layouts of typewriters such as that of Dvorak and colleagues took these factors into account. The actual amount of performance difference between Dvorak and QWERTY has never been settled, some researchers found the difference small, others found it quite large. It is worth noting that touch typing without looking at the keyboard was invented by L.V. Longley and F. E. McFurrin in the 1880s and it was not widely adopted by training schools until about 1915. Even if Sholes et al did user testing as some have suggested, it would not have been relevant to how people type today. For a few good sources of typwriting history, see H Yamada (1980) "A historical study of typewriters and typing methods: from the position of planning Japanese parallels". Journal of Information Processing, 2(4), 175-202. and also Cooper, W. E. (Ed.). (1983). Cognitive aspects of skilled typewriting. New York: Springer-Verlag. Dvorak, A., N.L. Merrick, W.L. Dealey, and G.C. Ford, Typewriting Behavior. 1936, New York: American Book Company. ___________________________________________ Shumin Zhai Research Staff Member, IBM Almaden Research Center, http://www.almaden.ibm.com/u/zhai Currently Visiting Professor at KTH (Royal Institute of Technology), NADA (Dept. Numerical Analysis and Computing Science), 10044 Stockholm, Sweden. Tel: +46-8-790-9278 (Office), +46-70-2924-332 (Mobile) "Ivan Poupyrev" @hitl.washington.edu on 10/04/2001 03:32:08 AM Sent by: owner-3dui@hitl.washington.edu To: <3d-ui@hitl.washington.edu> cc: Subject: Re: text/number input in VEs > From: "Boris Mansencal" > It is noteworthy that the QWERTY (or AZERTY in France) character layout > was "intentionally designed to slow down typist on the original mechanical > designs". This is actually a very persistent myth. I am attaching 2 messages from an ACM public mailing list that detail some of the issues of designing QWERTY keyboards (with references on literature) ivan ---------------------------------------------------------- From: "David Kieras" To: Sent: Thursday, July 22, 1999 12:49 PM Subject: Re: NYT Reporter seeks keyboard experts/designers > The actual history of how Sholes developed the QWERTY keyboard and the > first commercially viable typewriter is that he engaged in systematic > iterative user testing of prototype machines, with the design goal being to > support court recording. The "conventional wisdom" that the keyboard was > deliberately designed to slow people down couldn't be further from the > truth. It's remarkable how many human factors textbooks repeat this myth, > when the facts are an early human factors success story. > > At 9:49 AM -0400 7/20/99, Brian Shackel wrote: > >Dear Jenny Lee, > > > >Like so many you are convinced by hearsay and chitter-chatter "standard > >QWERTY inefficiency story". Oh dear! :-) Why not write up the facts? > > > >The latest good summary is by a former Doctoral student of mine Jan Noyes > >'Keyboards' Chapter 1 in 'Interface Technology - the Leading Edge' 1999, > >Research Studies Press, Baldock UK, and at 325 Chestnut Street, > >Philadelphia, PA 19106 (ISBN 0-86380-233-8). > > > >I quote from her chapter - > >"Despite these criticisms, the QWERTY layout has been shown to be nearly > >optimal in terms of typing speed; this is primarily because it allows > >almost all keystrokes to be made by alternate hands (Kinkead, 1975). > >Thomas (1972) and others reinforced this point." > > > >Also you might wish to refer to Norman & Fisher in Human Factors (journal) > >1982, vol 24, pp. 509-519, where they showed by extensive computer > >simulated tests that even the generally suggested best keyboard (the > >Dvorak) has never been shown, and could not by computer simulation be > >shown, to be more than 6% better than QWERTY - which is nowhere near enough > >to be worth learning to changeover to it. > > > >Yours, Brian Shackel.