Dataset for: Digital Finance and Inclusive Green Growth: The Mediating Role of Green Innovation and the Moderating Role of Artificial Intelligence
Description
This repository provides a balanced panel dataset comprising 60 global economies observed annually from 2011 to 2025, yielding 900 total observations. It is structured to investigate the interrelationships among digital financial inclusion, green innovation, artificial intelligence readiness, and inclusive green growth for empirical publication in peer-reviewed journals like the International Review of Economics & Finance (IREF).File Structure (Digital_Finance_Inclusive_Green_Growth_Dataset_2011_2025.xlsx):The uploaded multi-tab Excel workbook contains four dedicated sheets:Panel_Dataset_2011_2025: The primary raw and structured panel matrix (900 rows, 11 columns) capturing the dependent variable ($\text{IGG}$), independent variable ($\text{DF}$), mediator ($\text{GI}$), moderator ($\text{AI}$), and macroeconomic controls ($\text{HC}$, $\text{TO}$, $\text{FDI}$, $\text{INF}$).Variable_Metadata: Full documentation dictionary detailing column codes, measurement units (indices, percentages, patents per capita), mathematical transformations (logarithmic scaling for patent data), and official international sources (World Bank WDI, IMF Financial Access Survey, WIPO, OECD, Oxford Insights, ITU, Penn World Table).Country_Sample_60: Complete inventory of all 60 cross-sectional sample countries categorized by geographical region and economic development status.Submission_Workflow_Steps: Procedural roadmap outlining the 5 consecutive research phases:Dataset auditing and cleaning.Mendeley Data repository deposit and DataCite DOI generation.Descriptive and diagnostic statistical testing.Advanced econometric estimation using Stata / R / Python (two-step System GMM dynamic panel models and moderated-mediation interaction regressions).Final IREF manuscript compilation and submission.
Files
Steps to reproduce
Step 1: Data Extraction and PreparationGather secondary annual panel data for the 60 sampled countries spanning from 2011 to 2025 (totaling 900 observations).Collect raw series from official international repositories: World Bank World Development Indicators (WDI) and World Inequality Database (WID) for Inclusive Green Growth ($\text{IGG}$), IMF Financial Access Survey for Digital Finance ($\text{DF}$), WIPO and OECD Patent Statistics for Green Innovation ($\text{GI}$), Oxford Insights and ITU for AI Readiness ($\text{AI}$), and the Penn World Table / WDI for Human Capital ($\text{HC}$), Trade Openness ($\text{TO}$), FDI Inflows ($\text{FDI}$), and Inflation ($\text{INF}$).Organize the data into the multi-tab Excel workbook (Digital_Finance_Inclusive_Green_Growth_Dataset_2011_2025.xlsx), structuring tabs for the raw panel data, variable metadata dictionary, 60-country classification sample, and submission workflow steps.Step 2: Repository Deposit and DOI GenerationLog into the Mendeley Data platform and create a new repository entry using the finalized title, abstract, and file upload.Attach the audit-ready multi-tab Excel workbook and submit it to the Mendeley moderation queue.Secure the permanent DataCite DOI upon approval for citation in the manuscript's Data Availability Statement.Step 3: Preliminary Statistical Audit & DiagnosticsImport the dataset into econometric software (such as Stata, R, or Python) to verify balanced panel integrity ($N = 60, T = 15$).Apply natural logarithmic transformations to skewed variables (e.g., Green Innovation patents per capita).Generate descriptive statistics (mean, median, standard deviation, min, max) and a pairwise correlation matrix to check for multicollinearity among regressors.Step 4: Econometric Estimation via System GMMExecute two-step System GMM dynamic panel regressions to address potential endogeneity, unobserved heterogeneity, and persistence in inclusive green growth.Estimate baseline models evaluating the direct and mediated effects of digital finance and green innovation.Run moderated-mediation interaction models incorporating multiplicative terms ($\text{DF} \times \text{AI}$ and $\ln(\text{GI}) \times \text{AI}$) to test the moderating role of artificial intelligence readiness.Perform post-estimation diagnostic checks, including the Arellano-Bond AR(2) test for serial autocorrelation and the Hansen test for over-identifying restrictions.Step 5: Manuscript Compilation and Journal SubmissionDraft the empirical research paper structured around standard sections (Introduction, Literature Review, Methodology, Results, and Policy Implications) for submission to peer-reviewed journals