Accountant Ethics in ESG Reporting: Institutional Challenges and Behavioral Determinants. Dataset

Accountant Ethics in ESG Reporting: Institutional Challenges and Behavioral Determinants. Dataset This repository contains the replication dataset, codebook, and structural equation modeling (PLS-SEM) protocol for the study: "Accountant Ethics in ESG Reporting: Institutional Challenges and Behavioral Determinants" published in Business Ethics and Leadership (2026, Vol. 10, Issue 2). Abstract The institutionalisation of ESG reporting has made accountants’ ethical judgement a core issue for business ethics, ethical leadership, corporate accountability and ethical corporate culture. Yet prior studies have not sufficiently quantified how professional competencies, digital verification, corporate ethical culture and institutional pressure jointly shape accountants’ ethical resilience and ESG reporting quality. This study aims to develop and empirically validate a structural model explaining the behavioural, institutional and technological drivers of reliable ESG disclosure. Methodologically, the paper utilizes a secondary quantitative and qualitative analysis of an extensive empirical dataset originally compiled by the Green Transition Office (NGO DiXi Group). The authors' independent analytical contribution consists of data restructuring, statistical processing, PLS-SEM modeling, and empirical interpretation. The modeled dataset comprises survey microdata from 420 Ukrainian enterprises from industry (32%), agriculture (24%), finance (18%), energy (14%) and services (12%), contextualized by transcripts from 12 institutional interviews and four expert focus groups. Repository Contents & File Descriptions 1. The processed metric survey microdata from n=420 valid responses of Ukrainian enterprises (Industry, Agriculture, Finance, Energy, and Services). Contains zero missing values. 2. SmartPLS 4 Project Configurations & Specification: Detailed settings and matrices used to run the Partial Least Squares Structural Equation Modelling (PLS-SEM) platform (v. 4.1.0). Methodological Overview & SmartPLS 4 Replication Steps - Stage 1 (Data Prep): Metric scaling applied. All indicator scores automatically standardized (mean = 0, variance = 1). - Stage 2 (Measurement Model): Path Weighting Scheme (max 300 iterations, stop criterion 10^-7). Internal consistency verified via Cronbach’s alpha (0.88 - 0.93) and Composite Reliability (0.91 - 0.95). Convergent validity met through AVE (0.68 - 0.78). Discriminant validity validated by HTMT matrix (< 0.85). Full collinearity issues rejected with VIF values between 1.69 and 2.61. - Stage 3 (Structural Model): Non-parametric bootstrapping executed with 5,000 resamples (Two-Tailed, alpha=0.05). Substantial predictive power achieved: R² = 0.76 for ESG Reporting Quality (SRQ), R² = 0.71 for Ethical Corporate Culture (ECC), and R² = 0.64 for Primary ESG Data Quality (IP). Latent Constructs & Codebook Structure - DV (Digital Data Verification): Indicators DV1–DV4 (Automation, digital internal QC, ERP/BI integration, Blockchain traceability). - PC (Professional Competencies): Indicators PC1–PC4 (ESG education, ethical reasoning, framework skills, continuous professional training). - IP (Institutional Pressure / Primary Data Quality): Indicators IP1–IP4 (Accuracy, completeness, timeliness, verifiability). - ECC (Ethical Corporate Culture / Stakeholder Trust): Indicators ECC1–ECC4 (Tone from the top, transparency, reliability, external trust). - ER (Accountant's Ethical Resilience): Indicators ER1–ER4 (Pressure resistance, judgment independence, objectivity, personal responsibility). - SRQ (ESG Reporting Quality): Indicators SRQ1–SRQ4 (GRI/ESRS alignment, materiality, comparability, greenwashing mitigation). Model Fit & Diagnostics Summary - SRMR: 0.064 (< 0.080 threshold by Henseler et al., 2016) - NFI: 0.908 (> 0.900 threshold) - Predictive Relevance: Q² values of 0.315 (IP), 0.442 (ECC), and 0.394 (SRQ), indicating robust predictive validity (Hair et al., 2022).