May, 1960 EXPERIMENTAL PLOTTING 149 
of the pointing accuracy on the Wild Autograph A-7 in the NRC laboratories. The 
readings were recorded by another person so that purely psychological factors could 
be minimized. In addition, the readings were repeated every hour from 9:30 to 16:30. 
The room temperature was controlled to within 2°C. The points used in this experi- 
ment were triangulation point 5 and elevation points 56 and 57. The accuracy of 
consecutive pointing varied from 28 cm to 17 cm for the poorest operator and from 
13 cm to 8 cm for the best operator. The results are shown in Table V. The model 
scale was 1:16,666. The figures for Operators 1, 2, 3, and 4 are the mean values of 
mean square errors in consecutive pointing computed for all three points observed. 
Against expectation there was no drop in the pointing accuracy towards the end of 
the working day. 
  
  
  
0-03mm "OT m Tee rm) ELI o— -— o A 5 
7 ——9—-——-—77 
002 4 rte Den 
0-0/: 71 =X hE 
ue 
0 He 1M Time of 
"1099 1100 1200... 4300: 1400 4599 . 16% Observation 
-O.0| À NN "ett no A 
S Dd 057 
-0.02 4 oT 
-0.03mm= 
Pointing Accuracy 
JOM Top lip ug a et 
O Ns = 
  
  
  
Fig. 6. Fluctuations in elevation readings during the day. 
However, there is quite a fluctuation in the absolute elevation values during the 
day. As an example, variations of elevation of the above-mentioned three points are 
presented in Fig. 6. The readings were made by Operator 4, who has very good acuity 
in stereo perception. This fluctuation is astonishingly large. Its extreme values for 
each operator are as follows: 
  
Operator Extreme change 
1 + 71 cm 
2 + 83 cm 
3 +61 cm 
4 + 55 cm 
In order to express these changes in a practical way, a fluctuation index was calculated 
for each operator by dividing the maximum change by his accuracy figure in con- 
secutive pointing: 
Operator Fluctuation index 
1 3.1 
2 4.0 
3 4.7 
4 5.0 
The fluctuation index indicates the range in which the elevation readings at identical 
points may vary for the same operator. This index is larger for the operators with 
good pointing accuracy in spite of the fact that the range of fluctuation for good 
operators is smaller than for poorer operators.