Extending Model for Enterprise AI Adoption
| Researcher | Fabio Correa |
| Institution | Touro University Worldwide |
| Program | Doctor of Business Administration (DBA) |
| Updated | December 30, 2025 |
📝 View Current Thesis Draft (PDF) | Defense Presentation | Chair approved - Defense scheduled
| Component | Status | Progress |
|---|---|---|
| Analysis Pipeline | ✅ Complete | 11/11 notebooks verified |
| Thesis Chapter 1 | ✅ Complete | Introduction (~2,800 words) |
| Thesis Chapter 2 | ✅ Complete | Literature Review (~5,200 words) |
| Thesis Chapter 3 | ✅ Complete | Methodology (~3,800 words) |
| Thesis Chapter 4 | ✅ Complete | Results (~3,200 words) |
| Thesis Chapter 5 | ✅ Complete | Analysis & Discussion (~4,100 words) |
| Thesis Chapter 6 | ✅ Complete | Conclusions (~2,500 words) |
| Bibliography | ✅ Complete | 93 references verified |
| Tables & Figures | ✅ Complete | 61 tables + 15 figures, all LaTeX formatted |
| IRB Approval | ✅ Complete | Letter embedded |
| Chair Review | ✅ Approved | Defense scheduled |
This research bridges academic rigor with business impact:
| Academic Contribution | Business Contribution |
|---|---|
| Extends UTAUT2 with AI-specific constructs | 16-item diagnostic tool for AI readiness |
| Rigorous EFA→CFA→SEM validation | Intervention targeting for 4 user segments |
| Novel moderator discovery (Experience) | Tailored training by experience level |
| Cross-population validation | Leaders as AI champions (d=0.74-1.14) |
| Empirical gap closure | Links to McKinsey/BCG adoption-value gap |
This research develops and validates the AI Readiness Scale (AIRS), a psychometrically sound instrument extending UTAUT2 for enterprise AI tool adoption. Through rigorous split-sample validation (N=523), we demonstrate:
| Finding | Evidence |
|---|---|
| Validated 8-factor model | 16 items, CFI=.975, TLI=.960, RMSEA=.065 |
| Price Value dominates | β=.505 (p<.001), strongest predictor |
| Experience moderates adoption | HM×Exp (p=.007) significant |
| Four user typologies identified | Enthusiasts 16%, Cautious 30%, Moderate 37%, Anxious 17% |
| Leaders dominate tool usage | d=0.74-1.14 across all AI tools |
| Model explains 85.2% variance | R²=.852 in Behavioral Intention (8-factor diagnostic model) |
Organizations now have a validated 16-item diagnostic instrument for assessing AI adoption readiness with:
- Theoretical grounding in UTAUT2 framework
- Empirical validation across student and professional populations
- Four distinct user segments for targeted intervention design
| Parameter | Value |
|---|---|
| Full Sample | N = 523 |
| EFA Development | n = 261 (50%) |
| CFA Holdout | n = 262 (50%) |
| Random Seed | 67 |
| Role | n | % |
|---|---|---|
| Student | 216 | 41.3% |
| Professional | 184 | 35.2% |
| Leader | 123 | 23.5% |
flowchart LR
subgraph UTAUT2["UTAUT2 Core"]
PE[Performance Expectancy]
EE[Effort Expectancy]
SI[Social Influence]
FC[Facilitating Conditions]
HM[Hedonic Motivation]
PV[Price Value]
HB[Habit]
end
subgraph Extensions["AI Extension"]
TR[AI Trust]
end
subgraph Outcome["Outcome"]
BI[Behavioral Intention]
end
PE -->|β=-.028| BI
EE -->|β=-.008| BI
SI -->|β=.136*| BI
FC -->|β=.059| BI
HM -->|β=.217*| BI
PV -->|β=.505***| BI
HB -->|β=.023| BI
TR -->|β=.106†| BI
| Index | Value | Threshold | Interpretation |
|---|---|---|---|
| CFI | .975 | ≥.95 | Excellent |
| TLI | .960 | ≥.95 | Excellent |
| RMSEA | .065 | ≤.08 | Good |
| SRMR | .048 | ≤.08 | Excellent |
| Factor | α | CR | AVE |
|---|---|---|---|
| Performance Expectancy | .803 | .804 | .673 |
| Effort Expectancy | .859 | .861 | .756 |
| Social Influence | .752 | .763 | .621 |
| Facilitating Conditions | .743 | .750 | .601 |
| Hedonic Motivation | .864 | .865 | .763 |
| Price Value | .883 | .883 | .790 |
| Habit | .909 | .909 | .833 |
| AI Trust | .891 | .891 | .804 |
| ID | Hypothesis | Result | Evidence |
|---|---|---|---|
| H1a | PE → BI (+) | ❌ Not Supported | β=-.028, p=.791 |
| H1b | EE → BI (+) | ❌ Not Supported | β=-.008, p=.875 |
| H1c | SI → BI (+) | ✅ Supported | β=.136, p=.024 |
| H1d | FC → BI (+) | ❌ Not Supported | β=.059, p=.338 |
| H1e | HM → BI (+) | ✅ Supported | β=.217, p=.014 |
| H1f | PV → BI (+) | ✅ Supported (Strongest) | β=.505, p<.001 |
| H1g | HB → BI (+) | ❌ Not Supported | β=.023, p=.631 |
| H2 | TR → BI (+) | ❌ Marginal | β=.106, p=.064 |
| H3 | Experience moderates | HM×Exp p=.007 | |
| H4 | BI predicts usage | ✅ Supported | ρ=.69, p<.001 |
| H5 | Role differs in usage | ✅ Supported | Leaders > Others |
All notebooks in airs_experiment/ are complete and verified:
| Phase | Notebook | Purpose | Status |
|---|---|---|---|
| 0 | 00_Create_Split_Samples |
Sample preparation | ✅ |
| 1 | 01_EFA_Experiment |
Factor exploration | ✅ |
| 2 | 02_CFA_Experiment |
Model validation | ✅ |
| 3 | 03_Measurement_Invariance |
Group equivalence | ✅ |
| 4 | 04_Structural_Model |
Hypothesis testing | ✅ |
| 5 | 05_Mediation_Analysis |
Indirect effects | ✅ |
| 6 | 06_Moderation_Analysis |
Interaction effects | ✅ |
| 7 | 07_Tool_Usage_Patterns |
Behavioral validation | ✅ |
| 8 | 08_Qualitative_Feedback |
Thematic analysis | ✅ |
| 9 | 09_Comprehensive_Review |
Gap analysis | ✅ |
| 10 | 10_Final_Synthesis |
Integration | ✅ |
AIRS_Data_Analysis/
├── airs_experiment/ # Analysis notebooks and outputs
│ ├── data/ # Processed datasets
│ ├── plots/ # Generated visualizations
│ ├── results/ # JSON output files
│ └── tables/ # CSV summary tables
├── thesis/ # DBA thesis documentation
│ ├── chapters/ # Chapter files (1-6)
│ ├── tables/ # Publication-ready tables (9)
│ ├── figures/ # Exported figures
│ ├── references/ # Bibliography (93 refs)
│ ├── EXECUTIVE_SUMMARY.md # Standalone executive summary
│ ├── PRACTITIONER_BRIEF.md # Standalone practitioner brief
│ ├── THESIS_SUMMARY.md # 5-page thesis summary
│ ├── DEFENSE_PRESENTATION.md # Committee defense (28 slides + backup)
│ ├── PRESENTATION_SLIDES.md # Committee presentation draft
│ └── DRAFT 07.pdf # Current compiled thesis
├── research/ # Literature and industry sources
│ ├── AIRS literatyre review.md # Theoretical foundation
│ ├── Bridging the AI Adoption Gap... # Industry benchmarks (585 lines)
│ ├── Meta-Analysis of UTAUT... # Blut et al. 2022 (84 pages)
│ └── Venkatesh_AOR_2021.md # AI-UTAUT research agenda
├── plan/ # Analysis planning documents
├── data/ # Raw data files
├── docs/ # Additional documentation
└── scripts/ # Utility scripts
| Document | Location | Words |
|---|---|---|
| Chapter 1: Introduction | thesis/chapters/01_introduction.md |
~2,800 |
| Chapter 2: Literature Review | thesis/chapters/02_literature_review.md |
~5,200 |
| Chapter 3: Methodology | thesis/chapters/03_methodology.md |
~3,800 |
| Chapter 4: Results | thesis/chapters/04_results.md |
~3,200 |
| Chapter 5: Analysis & Discussion | thesis/chapters/05_analysis_discussion.md |
~4,100 |
| Chapter 6: Conclusions | thesis/chapters/06_conclusions.md |
~2,500 |
| Appendices A-H | thesis/appendices.md |
Complete |
| Tables 4.1-4.9 | thesis/tables/ |
9 tables |
| Bibliography | thesis/references/bibliography.bib |
93 refs |
See PDF_GENERATION.md for complete documentation on:
- Full thesis build (
thesis/build-thesis.ps1) - Standalone document conversion (
thesis/convert-to-pdf.ps1) - Prerequisites and troubleshooting
| Decision | Rationale | Location |
|---|---|---|
| 4 constructs dropped (VO, EX, ER, AX-orig) | Poor reliability (α = .301–.582) | Ch 3 §3.4.4, Ch 4 §4.2.1 |
| AIRS vs UTAUT2 comparison | ΔR² = .016 (modest improvement) | Ch 4 §4.3.5 |
| H5 mediation untestable | EX/ER excluded from model | Ch 5 §5.6.2, §5.7.1 |
pip install -r requirements.txtKey dependencies: pandas, numpy, scipy, factor_analyzer, semopy, pingouin, scikit-learn, matplotlib, seaborn
# Navigate to experiment folder
cd airs_experiment
# Run notebooks sequentially
# 00 → 01 → 02 → ... → 10The validated 16-item scale measures 8 predictor constructs plus Behavioral Intention:
# Calculate construct scores (mean of 2 items each)
# 8 Predictor Constructs (16 items)
PE = mean(PE1, PE2) # Performance Expectancy
EE = mean(EE1, EE2) # Effort Expectancy
SI = mean(SI1, SI2) # Social Influence
FC = mean(FC1, FC2) # Facilitating Conditions
HM = mean(HM1, HM2) # Hedonic Motivation
PV = mean(PV1, PV2) # Price Value (STRONGEST predictor β=.505)
HB = mean(HB1, HB2) # Habit
TR = mean(TR1, TR2) # AI Trust
# Outcome Variable (4 items)
BI = mean(BI1, BI2, BI3, BI4) # Behavioral Intention- Venkatesh, V., et al. (2003). User acceptance of information technology. MIS Quarterly, 27(3), 425-478.
- Venkatesh, V., et al. (2012). Consumer acceptance and use of information technology: Extending UTAUT. MIS Quarterly, 36(1), 157-178.
- DeVellis, R. F., & Thorpe, C. T. (2021). Scale development: Theory and applications (5th ed.). Sage.
- Hair, J. F., et al. (2019). Multivariate data analysis (8th ed.). Cengage.
- Shin, D. (2021). The effects of explainability and causability on perception, trust, and acceptance. IJHCS, 146, 102551.
Full bibliography: thesis/references/bibliography.bib
@phdthesis{correa2025airs,
author = {Correa, Fabio},
title = {Artificial Intelligence Readiness Scale: Extending Model for Enterprise AI Adoption},
school = {Touro University Worldwide},
year = {2025},
type = {Doctoral dissertation}
}Code: MIT License Documentation: CC BY 4.0
Last Updated: December 30, 2025 Version: 7.1 (Defense presentation prepared)