94 



DESIGN AND USE OF INSTRUMENTS 



ments. Using either the 9, 12, or 3 o'clock 

 position for such alignment was very satis- 

 factory. Of these alternatives, however, the 

 3 o'clock position was the poorest from the 

 point of view of the accuracy with which 

 operators responded to the direction of in- 

 strument deviation when all were not in 

 check. In this respect the 9 o'clock position 

 is thought to have special advantage for 

 check reference because it is ''natural" to 

 interpret a high pointer position as meaning 

 too much and a low pointer position as 

 meaning too little of the indicated function. 



Perhaps the best way to summarize the 

 advantages of systematic instrument align- 

 ment is in terms of some of the specific data 

 reported by Warrick and Grether. If all an 

 operator has to do is indicate whether the 

 instruments do or do not check, he can make 

 this reaction to a bank of 16 aligned instru- 

 ments in about 0.75 seconds on the average. 

 When the instruments have differently ori- 

 ented check points, it takes him about 1.6 

 seconds for the same job. Similarly, if the 

 operator has to respond to the direction of a 

 deviation, not just to the fact that a devi- 

 ation exists, he saves about two seconds and 

 makes about 50% fewer errors if the check 

 points are aligned than if they are not. 



To answer the question whether scale form 

 is important for the speed and accuracy of 

 check reading an isolated or single instru- 

 ment, Grether and Connell (35) undertook a 

 reaction-time test to evaluate five forms of 

 data presentation. Subjects viewed an in- 

 strument in repeated exposures and were 

 required to respond to changes from previous 

 indicated values. The instrument forms 

 used were a circular scale with moving 

 pointer, a circular scale moving behind a 

 fixed pointer, two vertical scales, one with a 

 moving and one with a fixed pointer, and 

 finally, a direct reading counter. Unfortu- 

 nately, the report of this experiment leaves 

 unclear the extent to which the results were 

 influenced by the particular scale values 

 chosen as check points. The circular scale 

 with the moving pointer appeared to be 



clearly superior to the circular scale which 

 moved behind the fixed pointer, but the sig- 

 nificance of the differences between either of 

 these instrument forms and the others was 

 not reported. Until further test data be- 

 come available it seems appropriate to ob- 

 serve only that the observed differences in 

 check reading performance between the mov- 

 ing pointer circular instrument, the vertical 

 scales and the counter were all quite small. 



Problems in the Design of Specific 

 Instruments 



Typical of the research which may be 

 needed in order to discover satisfactory de- 

 signs for instruments which serve specific 

 purposes, is the work which has been done on 

 the class of orientation instruments and on 

 the design of a 2400 military clock. These 

 studies will be reviewed briefly. 



Interest in orientation instruments (30, 

 34, 38) stems particularly from the ease with 

 which operators often make reversed inter- 

 pretations of an indicated change in the 

 orientation of their vessel or aircraft. Such 

 reversed interpretations are very common 

 with the traditional form of aircraft gyro 

 horizon (11, 12, 54) on which the horizon 

 bar is stabilized and remains parallel to the 

 true horizon while the airplane figure on the 

 instrument remains in fixed position in the 

 airplane. The pilot, seeing the bar tilt with 

 reference to all other objects in his cockpit, 

 is inclined to interpret that tilt as the angle 

 of his \vings — which in fact tilt in the other 

 direction. Out of the study of this and sim- 

 ilar orientation devices, Fitts and Jones (27) 

 advance the hypothesis that in general the 

 "aircraft reference principle" rather than the 

 "external reference principle" should be fol- 

 lowed in the design of instruments for the 

 control of aircraft in situations where spHt- 

 second reactions are demanded. By the air- 

 craft reference principle they mean the prin- 

 ciple of having the moving element of an 

 indicator move in the same direction in rela- 

 tion to the pilot as the aircraft is moving in 

 relation to the ground. Reaction-time data 



