"Statistical Evidence for the IRT-Based Maturity Model Validation Framework (IMVF) Applied to Digital Transformation Assessment "

"This document compiles comprehensive statistical evidence supporting the IRT-Based Maturity Model Validation Framework (IMVF) applied to Digital Transformation Assessment (DTA). Drawing on Measurement Theory principles analogous to sensor calibration in engineering systems, it demonstrates how Item Response Theory provides a probabilistic measurement foundation for assessing organisational digital maturity with precision and reproducibility. The materials document the statistical refinement of a 47-item pool to a validated 33-item instrument calibrated across 2,008 organisations spanning 103 countries and 21 industry sectors. Complete parameter estimates, model-data fit statistics, precision profiling, and R-code enable full reproducibility of the validation process and support practitioners in interpreting DTA scores with appropriate confidence intervals for strategic decision-making.Document Structure OverviewSection I \u2014 DTA Instrument Description Development pathway, conceptual foundations, target audience specification, comparative analysis against existing maturity models, and complete 47-item pool documentation with domain mapping.Section II \u2014 Sample Characterisation Report Comprehensive demographic profiling: geographic distribution (103 countries), industry sectors (21 NACE classifications), organisational size (turnover and headcount), company age, respondent management levels, and field of activity. Comparative analysis between full sample (n=2,008) and cleaned calibration sample (n=1,284).Section III \u2014 Data Collection & Preparation Deliverables Missing data protocols, exclusion criteria, sample size adequacy verification (KMO=0.965, Bartlett's \u03c7\u00b2(1,081)=34,847.2, p<0.001), response category consolidation procedures, and dataset readiness assessment.Section IV \u2014 Dimensional Analysis Two-iteration refinement process: Parallel Analysis results, Exploratory Factor Analysis (EFA) with full factor loading matrices, Principal Component Analysis (PCA) biplot classification, and Quality Gate 1 assessment confirming essential unidimensionality of the 33-item pool.Section V \u2014 IRT Model Selection & Parameter Estimation Graded Response Model (GRM) justification and implementation. Complete item discrimination parameters (a), threshold parameters (b\u2081\u2013b\u2084), Item Category Response Curves (ICRCs), model-data fit statistics (SRMSR=0.0289, S-X\u00b2 item-level assessment), latent trait estimates (\u03b8) with three scale transformations (logit, S\u2080,\u2081, S\u2080\u208b\u2081\u2080\u2080), Test Information Function (TIF) precision profiling, and marginal reliability (\u03c1\u2098=0.9721).Section VI \u2014 Final Validated Instrument Documentation Executive statistical validation report with quality verdicts, complete instrument calibration and quality records (11 structured sections), scoring specifications, and computational environment documentation.Section VII \u2014 R-Algorithm Reproducible code snippets for all analytical procedures: dimensional analysis, IRT parameter estimation, scale transformations, precision profiling, and visualisation routines."