Sharpened Romberg data from three subjects were not analyzed due to
excessive difficulty with the task, particularly in the occluded condition.
However, their data were not exceptional on other variables and were
therefore included in the rest of the analysis![[*]](/host/stout4/perseant/latex2html/icons.gif/foot_motif.gif){kind=icon}
. One
outlier was removed from the visibility data (the subject rated the visual
background as quite visible in both the see-through and occluded
conditions). The three pre-exposure ataxia trials were averaged, as were
the two post-exposure ataxia trials (after determining that the
post-exposure trials were not significantly different). Data on the CI
visual task (calling out the colored signs) were not formally analyzed as
all subjects performed perfectly or near-perfectly in both conditions.
The data are summarized in Table 6.3.4.
| Condition | ||
| Measure | See-through | Occluded |
| SSQ | 15.0(16.2) | 18.7(15.0) |
| Total Stance Breaks * | 1.6(1.9) | 2.6(2.7) |
| SB1 | 0.6(0.8) | 1.0(1.2) |
| SB3 | 1.0(1.6) | 1.6(1.9) |
| Vection | 3.0/7.0(1.7) | 4.0/7.0(1.9) |
| Background Visibility | 2.5/7.0(0.9) | N/A |
| Pre-Exposure Ataxia | 6.2(1.9) | 6.8(2.3) |
| Post-Exposure Ataxia | 7.4(3.8) | 8.3(3.9) |
``SSQ'' is a reported simulator sickness value [55]. ``SB1'' and ``SB3'' are the total number of stance breaks in the first and third minute, respectively. For the skewed data (SSQ, vection and visibility ratings), medians are reported. Other values are means. Standard deviations are given in parentheses. ``*'' implies that the difference between conditions is significant at p < .05.
Total stance breaks, and pre- and post-exposure ataxia data were normal and therefore no transformations were required. These data were analyzed with a 2-tail paired t-test. The questionnaire data, SSQ and vection ratings, were analyzed using a non-parametric, 2-tail paired Wilcoxon, as these data were markedly skewed.
Total stance breaks were significantly lower in the see-through than occluded conditions (p < .03).
No difference was found across conditions for SSQ, vection or post-exposure ataxia scores. The SSQ scores after both conditions were significantly higher than the pre-test SSQ scores (p < .03), indicating that the stimulus did have an effect, simply not a condition-specific effect. After finding no difference between conditions, pre- and post-exposure ataxia data were each pooled across conditions, resulting in significantly greater ataxia for post-exposure than pre-exposure (p < .04). This indicated the existence of increased ataxia after only a 4.5 minute exposure to a vection stimulus. Pooling across conditions, no difference was found between SB1 and SB3.