An entropy-informed method to diagnose front-stage rationality = 엔트로피 기반의 프런트스테이지 합리성 진단 방법

다국어 초록 (Multilingual Abstract)

Customer behaviour in retailing environment often departs from the trajectories imagined by designers, yet there is no agreed method for systematically assessing how well the front stage of a service actually works in use. This thesis addresses that gap by developing and evaluating an entropy-based diagnostic method for front-stage rationality (FSR), defined as the extent to which the front-stage environment is coherent, legible and controllable from the customer’s point of view. A three-phase mixed-methods case study was conducted in a flagship luxury fashion store, linking perceived complexity (PC) and store chaos (SC). In Phase 1, customer movements and behaviours in 13 functional areas were mapped and converted into Shannon entropy values, which served as quantitative indicators of SC and produced an initial zoning of low-, medium- and high-chaos areas. In Phase 2, this zoning was combined with a PAD-based survey of 425 customers to examine how PC and SC jointly shape arousal, dominance and pleasure, and to test whether entropy-derived SC captures meaningful differences in front-stage experience. Phase 3 used a closed card-sorting study with 100 participants, split into high and low fashion involvement groups, to compare the entropy-based zoning with customers’ cognitive organisation of spatial and design cues. Across the three phases, PC and SC emerged as empirically distinct: some visually simple areas were behaviourally chaotic, while some visually rich areas supported orderly browsing; PC showed a double-edged pattern, increasing arousal while lowering dominance under certain conditions, and SC acted as a contextual factor that conditioned how complexity was experienced. The experimental procedures were consolidated into a standardised FSR diagnostic method, comprising a stepwise workflow and decision-tree style guidelines for interpreting entropy scores, emotional responses and card-sorting patterns together. The thesis contributes theoretically by sharpening the distinction between PC and SC and positioning SC as an entropy-based lens on front-stage environments; methodologically by offering a replicable workflow that integrates behavioural mapping, entropy analysis, PAD measures and cognitive structure validation; and practically by showing how designers and managers can diagnose where carefully planned experiences succeed or break down in retail settings. Looking forward, the thesis points to the potential for integrating the FSR diagnostic method with AI-based analytics to support real-time detection of emerging store chaos and adaptive adjustment of front-stage design.

국문 초록 (Abstract)

리테일 서비스 환경에서 고객 행동은 종종 디자이너가 상정한 동선과 사용 시나리오에서 벗어나지만, 서비스의 프런트 스테이지가 실제 이용 맥락에서 얼마나 잘 작동하는지를 체계적으로 평가하는 합의된 방법은 부족하다. 본 논문은 이러한 공백을 해소하기 위해 프런트 스테이지 합리성(front-stage rationality, FSR)을 진단하는 엔트로피 기반 방법을 개발·평가했으며, FSR을 고객 관점에서 프런트 스테이지 환경이 얼마나 일관적(coherent)이고 파악 가능(legible)하며 통제 가능(controllable)한지로 정의한다. 플래그십 럭셔리 패션 매장을 대상으로 3단계 혼합방법 사례연구를 수행하여 지각된 복잡성(perceived complexity, PC)과 매장 혼돈(store chaos, SC)의 연계를 검토하였다. 1단계에서는 13개 기능 구역에서의 고객 이동과 행동을 행동 맵으로 기록하고 이를 Shannon 엔트로피 값으로 변환하여 SC의 정량 지표로 삼았으며, 저·중·고 혼돈 구역으로의 1차 구역화를 도출하였다. 2단계에서는 이 구역화를 425명 PAD(pleasure–arousal–dominance) 설문과 결합해 PC와 SC가 각성(arousal)·지배감(dominance)·쾌(pleasure)를 어떻게 공동으로 형성하는지 분석하고, 엔트로피로 산출된 SC가 프런트 스테이지 경험의 의미 있는 차이를 포착하는지를 검증하였다. 3단계에서는 패션 관여도가 높은 집단과 낮은 집단으로 구분된 100명을 대상으로 폐쇄형 카드소팅을 실시하여, 엔트로피 기반 구역화와 고객의 공간·디자인 단서에 대한 인지적 조직을 비교하였다. 분석 결과 PC와 SC는 경험적으로 구별되었으며, 시각적으로 단순한 구역이 행동적으로는 혼돈이 높게 나타나는 반면 시각적으로 풍부한 구역이 질서 있는 탐색을 지원하는 경우도 확인되었다. 또한 PC는 조건에 따라 각성을 높이면서도 지배감을 낮추는 양날의 패턴을 보였고, SC는 복잡성이 어떻게 경험되는지를 조건화하는 맥락 요인으로 작동하였다. 본 연구는 절차를 표준화하여 엔트로피 점수, 정서 반응, 카드소팅 패턴을 통합적으로 해석하는 단계적 워크플로와 의사결정나무 형태의 가이드라인으로 구성된 FSR 진단 방법을 제시한다. 이로써 이론적으로는 PC와 SC의 구분을 정교화하고 SC를 프런트 스테이지 환경을 읽어내는 엔트로피 기반 렌즈로 위치시키며, 방법론적으로는 행동 매핑·엔트로피 분석·PAD 측정·인지구조 검증을 통합한 재현 가능한 진단 워크플로를 제공하고, 실무적으로는 리테일 현장에서 계획된 경험이 성공하거나 붕괴하는 지점을 디자이너와 매니저가 진단할 수 있는 근거를 제시한다. 향후에는 AI 기반 분석을 FSR 진단 방법과 결합하여 새롭게 발생하는 매장 혼돈을 실시간으로 탐지하고 프런트 스테이지 디자인을 적응적으로 조정하는 가능성을 제안한다.

목차 (Table of Contents)

• 1. Introduction 1
• 1.1 Background and Motivation 1
• 1.2 Research gap 1
• 1.3 Research Aim and Objectives 4
• 1.4 Research Questions 5
• 1.5 Scope and Contributions 6
• 1.5.1 Scope 6
• 1.5.2 Contributions 7
• 1.6 Structure of the Dissertation 8
• 2. Literature Review 11
• 2.1 Phenomenon-Based Theorising (PBT) 11
• 2.1.1 Why PBT fits design research? 11
• 2.1.2 Design Orientation 11
• 2.1.3 Contrast with Hypothesis-Driven Approaches 12
• 2.1.4 TEP framework 13
• 2.1.5 Application in this Study 13
• 2.2 Servicescape Front-Stage Foundations and Research Gap 14
• 2.2.1 Servicescape Theory 15
• 2.2.2 Service Blueprint and Touchpoint Architecture 17
• 2.2.3 Experience Guidance and Behavioural Logic 19
• 2.2.4 Existing Evaluation Approaches for Front-Stage Environments 22
• 2.2.4.1 Servicescape and Environmental Measurement Tools 23
• 2.2.4.2 Customer Journey and Touchpoint Assessment 24
• 2.2.4.3 User Experience and Usability-Oriented Evaluations 26
• 2.2.4.4 Spatial Observation and Behavioural Tracking 28
• 2.2.4.5 Experience and Satisfaction Measurement 29
• 2.2.5 Limitations in Current Approaches 30
• 2.3 Perceived Complexity & Emotional Response 34
• 2.3.1 Visual complexity + order 34
• 2.3.2 Cognitive Processing and Environmental Interpretation 35
• 2.3.3 Affective Responses and the PAD Model 36
• 2.4 Store Chaos & Behavioural Entropy 38
• 2.4.1 Chaos and Complexity in Environmental Contexts 38
• 2.4.2 Entropy and the Unpredictability of Human Behaviour 39
• 2.4.3 Positioning Store Chaos within Service Environments 40
• 2.5 From Perceived Complexity and Store Chaos to Front-Stage Rationality 41
• 2.5.1 Definition of Perceived Complexity 41
• 2.5.2 Definition of Store Chaos 44
• 2.5.3 Front-Stage Rationality 45
• 2.6 Literature Review Conclusion 46
• 3. Research Methodology 48
• 3.1 Methodological Orientation (PBT / TEP logic) 48
• 3.1.1 Philosophical Position 48
• 3.1.2 Applying the PBT Logic 49
• 3.1.3 The TEP Framework in Practice 49
• 3.1.4 Research Paradigm and Mixed-Method Rationale 50
• 3.1.5 Implications for Research Design 51
• 3.2 Research Design Overview (three-phase mixed method) 52
• 3.2.1 Purpose of the Research Design 52
• 3.2.2 Design Logic under the PBT / TEP Framework 52
• 3.2.3 Overview of the Mixed-Method Structure 54
• 3.2.4 Longitudinal and Contextual Design 55
• 3.2.5 Integration and Analytical Flow 56
• 3.2.6 Summary of the Design 57
• 3.3 Case Context: Luxury Retail Environment 58
• 3.3.1 Case Selection Rationale 58
• 3.3.2 Store Description 59
• 3.3.3 Service Front-Stage Characteristics 59
• 3.3.4 Research Access and Ethics 60
• 3.3.5 Case Significance 60
• 3.4 Phase 1: Behaviour Observation & Entropy Capture 61
• 3.4.1 Overview and Purpose 61
• 3.4.2 Spatial Layout and Zoning 61
• 3.4.3 Observation Procedure and Behavioural Recording 64
• 3.4.4 Results of Entropy Calculation 68
• 3.4.5 Interpretation and Relevance to Front-Stage Rationality 70
• 3.4.6 Summary 71
• 3.5 Phase 2: Perceived Complexity Survey (PAD) 71
• 3.5.1 Purpose and Position in the Study 72
• 3.5.2 Questionnaire Design and Measurement Constructs 72
• 3.5.3 Hypothesis 73
• 3.5.4 Sample Characteristics 74
• 3.5.5 Rationale for Sample Size 75
• 3.5.6 Reliability & Validity 75
• 3.5.7 Descriptive statistics 78
• 3.5.8 Hypothesis Testing 80
• 3.5.9 Discussion of phase 2 85
• 3.5.9.1 Distinguishing Complexity and Chaos in Retail Environments 85
• 3.5.9.2 Double-Edged Effects of Perceived Complexity 86
• 3.5.9.3 Moderating Effects of Store Chaos 87
• 3.5.10 Summary 88
• 3.6 Phase 3: Card Sorting Validation (Cognitive Structure) 89
• 3.6.1 Purpose and Rationale 89
• 3.6.2 Participants and Grouping 89
• 3.6.3 Rationale for Sample Size 90
• 3.6.4 Materials and Procedure 90
• 3.6.5 Analytical Approach 92
• 3.6.5.1 Hierarchical Clustering 95
• 3.6.5.2 Chi-Square Independence Test 97
• 3.6.5.3 Confusion Matrix 98
• 3.6.6 Discussion of phase 3 99
• 3.6.6.1 High Fashion Involvement Group 100
• 3.6.6.2 Low Fashion Involvement Group 100
• 3.6.6.3 Cross-Group Comparison and Quantitative Validation 101
• 3.6.7 Summary 101
• 3.7 General discussion 102
• 3.7.1 Integration Logic 102
• 3.7.2 Cross-Dimensional Relationships 103
• 3.7.3 Emergent Principles from Configuration Patterns 104
• 3.7.4 Methodological Validation through Triangulation 105
• 3.7.5 Summary 105
• 4. Development of the Entropy-Informed Method 107
• 4.1 Introduction to the Method 107
• 4.1.1 Purpose of Method Development 107
• 4.1.2 Link to Phenomenon and Research Questions 107
• 4.1.3 Overview of the Entropy-Informed Diagnostic Workflow 109
• 4.2 Conceptual Foundation 111
• 4.2.1 Behavioural Unpredictability and Information Entropy 111
• 4.2.2 PC and SC as Diagnostic Indicators of Front-Stage Rationality 112
• 4.2.3 Rationale for Parallel Behavioural and Perceptual Measurement 113
• 4.3 Core Workflow A: Store Chaos (SC) Strand 114
• 4.3.1 Preliminary Preparations 114
• 4.3.2 Observation and Behaviour Recording 114
• 4.3.3 Entropy Calculation and SC Classification 115
• 4.3.4 Visualisation of SC Outputs 116
• 4.4 Core Workflow B: Perceived Complexity (PC) & Emotions Strand 116
• 4.4.1 Preliminary Preparations 116
• 4.4.2 Data Collection and Processing (PC and PAD) 117
• 4.4.3 Visualisation of PC and PAD Outputs 118
• 4.5 Integration and Diagnostic Logic 118
• 4.5.1 Comprehensive Diagram of Normalised Group-Level Data 118
• 4.5.2 From Empirical Patterns to Configuration Principles 120
• 4.5.3 SC × PC Diagnostic Matrix 121
• 4.5.4 Decision Tree for Front-Stage Diagnosis 123
• 4.6 Optional Extension: Perceived Chaos by Customer Segment 126
• 4.6.1 Purpose and Scope of the Extension 127
• 4.6.2 Preparations and Data Collection 128
• 4.6.3 Clustering and Segment-Based Outputs 128
• 4.6.4 Using the Extension in Diagnosis 129
• 4.7 Standardisation and Quality Control 130
• 4.7.1 Standard Input and Output Definitions 130
• 4.7.2 Quality Control and Reliability 133
• 4.8 Chapter Summary 135
• 5. Theoretical Reflection and Experts Evaluation 137
• 5.1 Position and Purpose of this Chapter 137
• 5.2 Internal Theoretical Evaluation of the Method 138
• 5.2.1 Conceptual coherence and robustness 138
• 5.2.2 Empirical robustness and triangulation 139
• 5.3 Expert Evaluation 140
• 5.3.1 Background to the expert consultation 140
• 5.3.2 Service design expert opinions 141
• 5.3.3 Spatial designer expert opinions 146
• 5.3.4 Management expert opinions 147
• 5.4 Limitations and Extensions of the method 149
• 5.4.1 Conceptual scope: rationality and brand aesthetics 149
• 5.4.2 Temporal representation of behaviour 150
• 5.4.3 Context dependence of entropy thresholds 151
• 5.4.4 Standardisation of area division 151
• 5.4.5 Accessibility for designers 152
• 5.4.6 Resource demands and performance linkage 153
• 5.5 Chapter Summary 154
• 6. Conclusion and Implications 159
• 6.1 Recap of Research Aim and Approach 159
• 6.2 Summary of Key Findings 160
• 6.3 Theoretical Contributions 162
• 6.4 Practical Implications 165
• 6.4.1 Implications for designers and service planners 166
• 6.4.2 Implications for managers and retailers 166
• 6.4.3 Potential implications for education and professional development 167
• 6.5 Methodological Contributions 168
• 6.6 Research Limitations 169
• 6.7 Future Research Directions 170
• 6.8 Concluding Remarks 171