Document m4awqODkowEBdr9aBJOYZxRZ

102 ROSS A. MC FARLAND in a discontinuous task carried on with the right hand when the left hand was. performing a fairly difficult but dissimilar and continuous task.80 "Expected" relationships probably originate in the previous experiences of individuals rather than in any anatomical organization. Even though the "unex pected" relationships may be learned quickly, in many cases the great amount: of previous contrary experience probably contributes toward the rapid forgetting, of. these learned relationships. Certainly, the use of other controls with confusing' similarities and differences between the particular "unexpected" display-control relationships would be likely to lead to errors in the operation of the latter. In addition, even though an operator has learned to make correct, "unexpected,1 display-control responses, he is very likely to respond in the way most natural to him when an emergency arises, and his response will therefore be in error. This, of course, will vary with the level of learning of the "unexpected" response. In one panel arrangement, subjects first turned a knob in order to center a light in a semicircular display, and then turned a second knob in order to center a, light in another semicircular display. This alternation between panels was con tinued for a series of responses. Time and error scores were not large when a clock wise control movement produced a clockwise indicator light movement on both < panels, nor were they large when a counterclockwise control movement produced: a counterclockwise indicator light movement on both. However, when a clockwiscclockwise relationship existed on one panel, and a clockwise-counterclockwise re-; lationship on the second, time and error'scores'incfeas6d_greatly'for"both:81-Sucili' data clearly indicate the importance of having standard display-control relation-' ships on all equipment in order to avoid errors in rapid responses. If the best arrangements and relationships are employed when controls are' designed fur situations in which visual displays must be responded to rapidly and) accurately, speed of response will be increased, learning of the control task Will probably be facilitated, and incorrect responses will be less likely to occur, par-.: ticularly when the operator is working, under conditions of boredom, fatigue, or' emergency. If compromises must be made in the direction-of-motion relationships) between display and controls, it is probably more important that all controls on) the panel have the same direction-of-mqtiqn relation to the displays even though) they may npjg^tbe tlaost "expected"^reiationsKipr'A:n6tlrer^sItei`ffative`is-to-make' thediafhayingVan."iinexpected" relationship to a control a different' type of s') play than the others (for example, a linear type if the others are circular), so that: its differentiation is obvious. In tSKi_wK'efe"'5Hly"'one control is used continuously,-the problem. oLlh " R. E. Jones, A survey of pilot preferences regarding knob shapes to be used in codingj aircraft controls. Memorandum Rept. No. MCREXD-894-4E, Air Materiel Command, TJSAI], October 1, 1947. "M. J. Warrick, Effect of motion relationships of speed of positioning visual indicators by rotary control knobs, Tech. Rept. No. BSlt, Air Materiel Command, TTSAF, October, 1949 HUMAN ENGINEERING AND INDUSTRIAL SAFETY 103 fp.rVcuon-of-motion relationship does not appear to be significantly important |jnleqSBthe task is of short duration, when it more nearly approaches the discrete y)p<:oflesponse situation. Even here, since these controls are likely to be operated Unijunction with others, it is better that "expected" relationships be used if fiblep fitc Shape. The most efficient shapes for hand controls were determined by fmgl&y.eral types of knobs subjectively and objectively on the basis of maxi- Jfb|ge.,applied, arc through which the knob turned, and time required for ngflnd releasing. A two-pointed (cigar) shape and a four-pointed star gave !fiiignes.t mechanical advantage and were also favored by the subjects. Of the Mp||^ur-Pointed knob is superior in that it is easier to obtain a new grip on jaormiwhen a partial turn has been made. Circular controls and three-, five-, "anlieight-pointed controls were much less satisfactory.82 Mpluse of different shapes has also been found to reduce the time required I;^Mcurate operation of a series of controls. For instance, if the positions of l - . exchanged after individuals have learned one set of operations, the time lucfiffid^perform accurately a series of tasks with the controls is increased.87 Ts^Jiife.the control handles are coded with regard to shape, the detrimental SsMfting the location of controls is minimized even when visual cues are *K|standardized, this type of coding would be of great value where new `^^^feiileaming to operate a machine with a large bank of controls, and it ^M|pa5ticulaE-value-when-a-trained-operator,-or-perhaps -an. older oper- S||||pm one machine to another where the controls are in a different posi|ffipP>verify his choice of controls and avoid errors even though he might ""Teh for and operate the controls through force of habit without the aid j!!y,`Tactual Displays" on page 92) .* Coding may also be achieved by Bfapbiflc colors for certain controls or types of controls. Color coding may ^ffi||fendently of, or in conjunction with, shape .coding. &|p|n .controls requiring the exertion of negligible force, such as the ||m|iator pointers, the diameter of the knob has little influence on the u||ed.5to make the setting provided only that it , can be conveniently ^^ffl'pbfof about 2a/4 inches in diameter appears to be a favorable , size, ^Scr!^afch_andle-does.not-aidrand.possibLy-hindersrthe-speed-of;adjust^ ^'jiifsioqfthis type.83 The same general finding has been reported for knobs a^S^fifor maximum signal strength, as where bracketing^ would be used SlI^^Migntrol approximately 30'to one side and then tp^the other, and ppSaLve-the distance bv feel). For a knob with the axis at.right angles flpiMorearm a diameter of 1 to 2 inches has been found superior. For *S^adi6hs and F. Hulverscheidt, Dre Bedienbarkeit von handgriffen_Vorrichtungen, Werkaleiter, 29, 193--197 (1935). `CTaJr,i'> tins and M. B. Connor, Some design factors in making settings on a linear 3,39HM09 (1949). * '!