Raw Data Table
Case Display Tot Err Err Time Tot Time % Err
1 M 1 18 62 29.032
2 M 1 17 85 20.000
3 M 1 9 72 12.500
4 M 0 0 51 0.000
5 M 0 0 93 0.000
6 M 1 14 91 15.385
7 M 0 0 40 0.000
8 M 0 0 36 0.000
9 M 2 36 91 39.560
10 M 1 12 45 26.667
11 M 0 0 88 0.000
12 M 0 0 44 0.000
13 H 0 0 77 0.000
14 H 13 108 140 77.143
15 H 2 37 86 43.023
16 H 2 13 57 22.807
17 H 0 0 43 0.000
18 H 0 0 56 0.000
19 H 0 0 40 0.000
20 H 3 105 161 65.217
21 H 2 17 54 31.481
22 H 0 0 41 0.000
23 H 0 0 42 0.000
24 H 2 10 45 22.222
Variable N Min Max Mean Std.Err Std.Dev Skewness Kurtosis Total Err 24 0 13 1.292 0.543 2.662 3.988 17.716 Err Time 24 0 108 16.500 6.081 29.791 2.548 6.075 Tot Time 24 36 161 68.333 6.548 32.077 1.487 2.234 % Err 24 0 77.1 16.877 4.492 22.005 1.379 1.487
MONITOR MONITOR MONITOR
Variable t df 2-Tailed p N Mean Std.Dev
Total Err -1.323 22 0.199 12 0.583 0.669
Err Time -1.278 22 0.215 12 8.833 11.256
Tot Time -0.274 22 0.786 12 66.500 22.561
% Err -1.107 22 0.280 12 11.929 14.156
HMD HMD HMD
Variable t df 2-Tailed p N Mean Std.Dev
Total Err -1.325 22 0.199 12 2.000 3.643
Err Time -1.278 22 0.215 12 24.167 40.009
Tot Time -0.274 22 0.786 12 70.167 40.436
% Err -1.107 22 0.280 12 21.825 27.543
Effect Wilk's Lambda Rao's R df 1 df 2 p-level 1 0.8399 0.8580 4 18 0.5075Three graphs are given below comparing the monitor and HMD performance data. The center dash is the mean, with extending bars showing the range of standard deviations.
From the data and graphs no conclusive evidence exists to support either visual mode. Subjects using the monitor system slightly out performed the HMD subject, but not to an significant degree.
Only a few bias were discovered. Due to the nature of the experiment and the subject selection process, the stratification is on the most obvious variable, the subject's declaration of right or left hand manipulation preference. The apparatus used in the experiment was not modified for a preference, and was configured for right-hand operation (refer to the Appendix on figures). Therefore, this limitation presented a known bias for those subjects requesting left-hand operation. However, in actuality the left handed subjects performed superior to right handed. All five left handed subject accomplished the task error free, while over half of right handed subjects committed at least one error. An explanation may be because the left handed subject knew of the greater difficulty and concentrated more on the experiment.
Men seemed to have higher performance than women. Four of five women committed an error compared to less than forty percent of men, which committed error.
Although age, in general, did not seem like a factor. All test subjects over the age of thirty made at least one error during their test.
Other than the above observations, no concise conclusions could be made. Most of the biases had no affect on the experiment. In the areas of gender and age, an equal number of subjects performed on the different visual modes. Therefore any advantages, due to gender or age, were equally spread and did not change the relative performance ratings of the monitor and HMD.
However, the handedness may have affect this result. Most of the left handed subjects were tested on the monitor. Therefore a correlation may exist. The left handed subjects may have been error free because they used a superior viewing system, the monitor. Or perhaps the explanation for the slight better performance ratings of the monitor is due to the adept left handed subjects.