Scientific Research on Keyboard Ergonomics
Survey of Research on Keyboard Ergonomics
Great article, survey and summarize research on keyboard ergonomics.
[The Split Keyboard: An Ergonomics Success Story by David Rempel, University of California, Berkeley.]
Biomechanics of the Upper Limbs
Here is a different article by another ergonomic researcher.
The standard keyboard creates several biomechanical problems for the operator. First, the hands tend to be ulnarly deviated up to 40° (mean values of 25°; Smutz et al., 1994) placing additional loading on the carpal tunnel and increasing the pressure within the tunnel as much as 13% (Werner et al., 1997). Second, to obtain a flat palm, the forearm tends to be pronated close to the anatomical limit (mean values of 76°), which requires the activation of the forearm muscles (mainly pronator teres and pronator quadratus). Such tension over extended time periods can also lead to muscular fatigue. Third, to compensate for this tension, there is a tendency for operators to lift the upper arms laterally and forward, which requires the activation of the shoulder muscles (primarily the deltoid and teres minor). Again, static tension may lead to fatigue. Fourth, depending on the height and slope of the keyboard, there is a tendency for the wrists to be extended up to 50° (mean values of 23°; Serina et al., 1999). Of all the possible wrist deviations, this wrist extension may be the most critical with carpal tunnel pressures increasing to 63 mmHg (for fingertip forces of 6 N), considerably above 30 mmHg, the threshold level for potential injury (Rempel et al., 1997).
Such problems at a typewriter keyboard were noticed as early as 1926 by Klockenberg, who proposed that the keyboard be split into two halves, each angled 15° from the center line (Figure 10.21, included angle is 30°), as well as tilted laterally down (sometimes termed tented). Furthermore, Klockenberg (1926) suggested an arching of the key rows for each half of the keyboard to better configure with the natural layout of the fingers. The lateral tilting was more specifically examined by Creamer and Trumbo (1960) with a mechanical typewriter cut into two halves and tilted at five different angles. Keying at the middle position of 44° was 5% significantly faster than at the extremes of 0° (flat) or 88° (nearly vertical). Kroemer (1964, 1965, 1972) performed a more detailed analysis by varying also the upper arm position and found that the subjects preferred a similar hand orientation of 40° for the upper arms hanging down naturally. Although the subjects preferred typing on a split and tilted keyboard over a standard keyboard, typing speed did not show any differences. Error rates, however, decreased by 39%.
Further experimentation by Zipp et al. (1981, 1983) using EMG measurements of the shoulder, arm, and hand muscle indicated optimal ranges of 0 to 60° for pronation and 0 to 15° for ulnar deviation, with the standard position for keyboards of 90° pronation and 20 to 25° ulnar deviation clearly beyond the optimal range. A 13° angulation from the centerline (26° included angle) showed lower EMG than a 26° angulation. In addition, preferred lateral tilt angles of 10 to 20° were smaller than the 44° found by Kroemer (1964, 1965, 1972). Because only three subjects had been utilized in the above experiments, Nakaseko et al. (1985) performed further testing on 20 experienced typists and found similar results with subjective preferences, which led to the first commercial split model standardized at a 25° split (internal angle), a 10° lateral tilt, and a 10° horizontal tilt (far edge higher) (Buesen, 1984). Since then, several other split or tilted models have been introduced and evaluated scientifically to provide better hand and wrist postures (Gerard et al., 1994; Tittiranonda et al., 1999; Zecevic et al., 2000).
from page 506 to 507 of [Biomechanics of the Upper Limbs: Mechanics, Modeling and Musculoskeletal Injuries By Andris Freivalds. At Buy at amazon ]