The last experiment focused on VR sickness. It is critical for walking interactions to prevent VR sickness in VR applications and to provide a sense of immersion to users. If VR devices are designed only for immersion, with no consideration for user VR sickness, only short sessions can be safely experienced. Thus, we conducted experiments to analyze the effects of the proposed walking interaction methods on VR sickness. We collected data using a simulator sickness questionnaire (SSQ) [38
]. The SSQ consists of 16 questions and allows the analysis and measurement of simulator sickness from various perspectives. Each question was derived from the results of various experiments on the effects of simulators and interaction methods used in VR applications, and the symptoms of users. Participants chose one of four answers (none, slight, moderate, and severe) for each of the 16 questions. None, slight, moderate, and severe are converted to 0, 1, 2, and 3 points, respectively. The total score of each participant is then calculated after applying a weight value. The scores of all participants are statistically analyzed to determine the level of VR sickness. In particular, the SSQ can analyze sickness in detail by classifying the questions. Items 1, 6, 7, 8, 9, 15, and 16 indicate nausea; items 1, 2, 3, 4, 5, 9, and 11 indicate oculomotor sickness; and items 5, 8, 10, 11, 12, 13, and 14 indicate disorientation sickness. In general, sickness is analyzed based on the sum of the weighted (nausea: 9.54, oculomotor: 7.58, disorientation: 13.92) detail items (nausea, oculomotor, and disorientation). For objective analysis of symptoms, the raw data excluding weights can also be used [40
]. In this study, both methods were applied to derive results. Table 4
displays the general analysis results for SSQ. The VR sickness results for each interaction based on the raw data were 6.45, 6.20, and 5.50 for gamepad control, hand control, and march-in-place control, respectively. When these values are divided by the total number of items (16) in the questionnaire, the results are 0.4, 0.39, and 0.34. When 0 is interpreted as no VR sickness and 1 is interpreted as slight VR sickness, these values suggest that the three-step walking interaction causes very little VR sickness on average. Furthermore, the walking interaction with march-in-place control caused the lowest levels of VR sickness (28.80, 5.50 < 31.98, 6.45 (G), 31.04, 6.20 (H)). The reasons for these results were analyzed by using the detailed items. In the case of the gamepad or hand interfaces, when there is an input, the camera moves toward the gaze direction without swaying. In the case of the walking simulator, however, when the user marches in place, the reaction is conveyed directly to the head and the user sees a reflection of this motion in the virtual scenes. This helps users feel like they are walking and helps them to accurately perceive the direction of movement. This had the greatest influence on the disorientation items. Furthermore, it was found that a longer period of walking resulted in a greater chance of VR sickness. In the case of the gamepad, 6.45 became 7.15 after 180 s and 8.05 after 300 s. Thus, the difference in value increased from 0.7 to 0.9. In the case of the hand interface, the value change ranged from 0.55 to 0.85, which is lower than that of the gamepad. In the case of the march-in-place interaction, the value change was the lowest, ranging from 0.4 to 0.5. This result suggests that user walking with their own legs experienced lower VR sickness when participating in long sessions in a VR environment. Finally, the raw data for all participants was plotted on graphs, which are shown in Figure 9
. There was a wide range in the levels of VR sickness indicated by users, from a minimum of 0 to a maximum of 22. However, participants whose total score was 16 or higher (mean score is 1.0 (slight) or higher) only accounted for approximately 10% of the participants. In other words, 90% of the participants in this experiment felt small levels VR sickness while experiencing the proposed walking interactions.