168 



PANEL DESIGN 



trying to drive an automobile in which the 

 clutch and break pedals were interchanged 

 in position — there would be many confusion 

 errors. In the same study, it was found 

 that 31% of the confusion errors were cases 

 of confusing the wheel and flap controls. 

 In many aircraft, both these controls are 

 switches, located near each other and used 

 together in landing and take-off operations. 

 They look and feel so much alike that a 

 small error in reaching can result in the 

 activation of the incorrect control. What 

 the pilot apparently needs is to have the 

 two switches sufficiently far apart so that 

 reaching errors will be rare, or to have the 

 two controls markedly different in appear- 

 ance so that he will know immediately when 

 he is about to make a mistake. 



Standardization and Coding of Controls 



Two major recommendations for the de- 

 sign of panels are obvious from this study. 

 In the first place, the position of all controls 

 in different models of the same type of 

 equipment should be standardized. Second- 

 ly, controls should be made as distinctive 

 as possible. Several specific suggestions 

 have been offered for making controls dis- 

 tinctive: (1) coding control handles by shape, 

 color, or size so that the operator learns to 

 associate one type of handle with each type 

 of control, (2) spatially separating controls, 

 that are apt to be confused, to avoid reach- 

 ing errors and to afford the operator dif- 

 ferential kinesthetic cues on which to base 

 his discriminations, (3) providing controls 

 with distinctive modes of operation such as 

 up and down, left and right, in and out, etc., 

 (4) as a last resort, using a system of warn- 

 ing lights or sounds to aid in the rapid rec- 

 tification of errors or using mechanical 

 guards to prevent accidental activation of 

 controls. The general principle underlying 

 these techniques for making controls dis- 

 tinctive is that the more sensory cues an 

 operator has to go on in making discrimina- 

 tions, the more accurate will be his perform- 

 ance. Therefore, the design engineer should 



be advised that the more distinctive visual, 

 tactual, or kinesthetic cues that can be built 

 into controls, the fewer will be the errors 

 of confusion. In fact, it has been suggested 

 further that the kinesthetic cues that are 

 already inherent in reaching for controls 

 in different positions could be used more 

 profitably if operators were trained in "blind 

 reaching" for controls; that is, were trained 

 to use kinesthetic cues dehberately. 



Only a start has been made in the in- 

 vestigation of the feasibiUty of these recom- 

 mendations. A study by Wietz has shown 

 that shape and color coding are very effec- 

 tive in reducing confusion errors (36). He 

 used a mock-up situation in which pilots 

 learned to associate each one of four identi- 

 cal levers with four different visual signals. 

 When the positions of the correct levers 

 were changed, performance in this situation 

 was greatly impaired and almost complete 

 retraining was required. If each lever had 

 a distinctive shape and color in the original 

 learning, however, confusion errors were at 

 a minimum when the positions of the correct 

 levers were changed as long as the distinc- 

 tive shape and color of each lever were still 

 associated with the same visual signals as 

 in original learning. Shape coding alone 

 served fairly well to reduce confusion errors, 

 but the effect was not as great as when both 

 shape and color were used. 



Although these results show clearly that 

 shape and color coding or shape coding alone 

 help to reduce confusion errors when the 

 position of controls are interchanged, it 

 should be pointed out that further work on 

 this problem is necessary. In the first place, 

 it has to be shoAATi that coding will work as 

 well in the operational situation as in the ex- 

 perimental mock-up. Secondly, we have to 

 be certain that the shapes used for coding do 

 not impair the efficiency of handling in- 

 dividual controls. Other experiments have 

 shown that the speed and accuracy with 

 which controls may be used and the amount 

 of strength one may apply to them depends 

 upon the handle or knob shape (22, 33). 



