Table 1. Investigation 1. Chi -values According to Bartlett’s Test. (Series 4—15).
Operator
no.
11.2.63
12.2.63
13.2.63
18.2.63
i
37. 1
highly sign
80. 1
highly sign
18.5
not sign
14.4
not sign
2
7.3
not sign
21.3
almost sign
50.2
highly sign
10. 8
not sign
3
65.1
highly sign
24.4
almost sign
103
highly sign
1.3
not sign
4
27
sign
39
highly sign
9.8
not sign
19. 15
not sign
Treatment of Data
Averages and standard deviations are computed from every series.
Grand averages and root mean square values of standard deviations
are computed for each operator everyday.
Primary data in these experiments are easily grouped for variance
analysis in order to investigate differences in the averages due to alco
hol. However, this analysis can be applied when the observations are
independent and normally distributed with the same variance (see [1]
16.2). Bartlett’s test has been used to determine if the computed stand
ard deviations are all estimates of the same theoretical value (see [1]
11.6). The test is applied to the coordinate of the horizontal parallax
only because we are most interested in the influences upon stereo
vision. According to the principles of the test the differences between
the standard deviations are said to be
not significant
if
chi 2
<
Chi 2 o.95
almost significant
if chi 2 0.95
^ chi 2
<
chi 2 o.99
significant
if chi 2 o 99
^ chi 2
<
chi 2 0.999
highly significant
if chi 2 0 999
^ chi 2
Table 2. Investigation 2. Chi 2 -values According to Bartlett’s Test. (All Series.)
Operator
no.
4.3.63
6.3.63
1
8.895
12.634'
2
7.358
0.834
3
14.838
6.763
4
9.087
11.085
4
12.158
17.175
No values is significant.
475