Title Page

ABSTRACT

Contents

Acronyms 14

Chapter 1. Introduction 16

1.1. Research Background 16

1.2. Scope of dissertation and research gap 18

1.3. Research objective 20

1.4. Dissertation structure 21

Chapter 2. Literature Review 24

2.1. Augmented reality 26

2.2. User-centered design and usability 28

2.3. Augmented Reality vs Virtual reality 32

2.4. AR interface and usability 37

2.5. Visibility in AR 42

2.6. Motion sickness 45

2.6.1. Motion sickness questionnaire (MSQ) 45

2.6.2. Simulator sickness questionnaire (SSQ) 47

2.6.3. Virtual reality sickness questionnaire (VRSQ) 48

2.6.4. Need of motion sickness scale for AR environment 50

Chapter 3. User Experience in AR environment 52

3.1. Study- I : Effect of interaction method, size, and distance in AR environment for near interaction 52

3.1.1. Method 53

3.1.2. Results 61

3.1.3. Discussion 71

3.1.4. Limitations 78

3.2. Study- II : Visibility in AR environment due to Transparency levels and background condition 79

3.2.1. Method 79

3.2.2. Results 88

3.2.3. Discussion 96

3.2.4. Limitation 98

Chapter 4. Development of motion sickness questionnaire for augmented reality environments 100

4.1. Method 100

4.2. Results 101

4.2.1. Analysis (SSQ) 101

4.2.2. Factor analysis 105

4.2.3. EFA and CFA for SSQ 108

4.2.4. Analysis (ARSQ) 114

4.3. Discussion 121

4.3.1. Motion sickness index for AR environments 121

4.3.2. VRSQ and ARSQ 127

4.3.3. Validity of ARSQ 128

4.3.4. Effect of size and distance to object on sickness 129

4.3.5. Limitation 130

Chapter 5. Conclusion 132

References 138

국문 초록 157

Curriculum Vitae 162

Table 2.1. Some of the popular AR devices 36

Table 2.2. Previous studies and sickness measuring tools 49

Table 3.1. Independent and dependent variables. 57

Table 3.2. Summary of ANOVA main effects. 67

Table 3.3. Summary of ANOVA interaction effects. 68

Table 3.4. Independent and dependent variables. 84

Table 3.5. Summary of ANOVA main effects. 95

Table 4.1. ANOVA results for SSQ. 103

Table 4.2. Factor loadings of the SSQ items in EFA. 111

Table 4.3. Absolute fit indices for the confirmatory factor analysis. 112

Table 4.4. Augmented reality sickness questionnaire (ARSQ). 113

Table 4.5. Computational score of ARSQ. 113

Table 4.6. ANOVA results for ARSQ. 117

Table 4.7. Sickness items in MSQ, SSQ, and VRSQ. 123

Table 4.8. ARSQ items to measure scores. 126

Figure 2.1. Virtuality continuum. 34

Figure 2.2. A frame of usability issues in AR environment. 40

Figure 2.3. A demo of virtual object and background). 44

Figure 3.1. Example of target buttons (left: large buttons, right: small buttons). 58

Figure 3.2. Experimental area for each button size. 58

Figure 3.3. Demo of button selection task. 60

Figure 3.4. Sample of hand gesture and clicker interaction. 60

Figure 3.5. Task completion times for different (a) button sizes, (b) distances, and (c) interaction types. ***p 〈 0.001. 64

Figure 3.6. Task completion times based on (a) distance and interaction type, and (b) distance and button size. Different circles indicate significant... 65

Figure 3.7. Task completion times based on (a) interaction type and distance, and (b) interaction type and button size. Different circles indicate... 65

Figure 3.8. Number of errors for different (a) button sizes, (b) distances, and (c) interaction types. *p 〈 0.05. 66

Figure 3.9. User satisfaction scores according to (a) button size, (b) distance, and (c) interaction type. *p 〈 0.05, **p 〈 0.01, ***p 〈 0.001. 70

Figure 3.10. Interaction effect between the interaction type and distance for satisfaction. Different circles indicate significant differences. 74

Figure 3.11. Experimental environment 81

Figure 3.12. Prototype model with black and dynamic background 82

Figure 3.13. A button selection task set with interaction type. 85

Figure 3.14. A pictorial display of ten conditions according to transparency and background. 87

Figure 3.15. Task completion time according to transparency and background conditions. *p 〈 0.05, **p 〈 0.01, ***p 〈 0.001. 89

Figure 3.16. The number of errors according to transparency and background conditions. 90

Figure 3.17. Button visibility scores according to transparency and background conditions. *p 〈 0.05, **p 〈 0.01, ***p 〈 0.001. 92

Figure 3.18. Interaction effect between transparency and background for button visibility. Different circles represent a significant difference. 92

Figure 3.19. Background visibility scores according to transparency and background conditions. *p 〈 0.05, **p 〈 0.01, ***p 〈 0.001. 93

Figure 3.20. Disorientation scores according to transparency and background conditions. *p 〈 0.05, **p 〈 0.01. 94

Figure 4.1. Box-plot of SSQ score for the AR environment. 101

Figure 4.2. SSQ scores according to size and distance. *p 〈0.05; **p 〈0.01; ***p 〈0.001. 104

Figure 4.3. CFA diagram for ARSQ components. 110

Figure 4.4. Box-plot of ARSQ score for AR environment (Box is not showing for disorientation because of 0 values) 115

Figure 4.5. ARSQ scores according to size and distance. *p 〈0.05; **p 〈0.01; ***p 〈0.001. 118

Figure 4.6. ARSQ scores according to transparency and background condition. *p 〈0.05. 120

Figure 5.1. Two recommended transparency levels for AR interaction. 134