Page 69 - Azerbaijan State University of Economics

Swaty Sharma, Munish Gupta: Does the Rise of Emerging Technologies Transform Digital
Entrepreneurial Activity? Evidence from OECD Nations

4.3.3 Unit Root Test:
A unit root test by Pesaran (2007) establishes the stationarity of the variables in the
developed models. The test utilizes Pesaran’s (2004) approach to assess cross-
sectional dependence (CSD) together with Pesaran & Yamagata’s (2008) slope
heterogeneity (SH) analysis to verify proper stationarity of the variables. The analysis
depends on stationarity as this condition helps researchers prevent spurious outcomes
while maintaining analysis validity.

4.3.4. Co-integration Analysis:

This study applies Westerlund’s (2007) cointegration test to determine the existence
of long-term equilibrium relations among digital entrepreneurship (TEN) and digital
transformation variables (ICT, Internet, RR&D, and R&D), along with control
variables (EST_BUS, GDP, and Financial Risk). The analysis evaluates the existence
of persistent, enduring relationships among essential variables across the OECD
nations included in the study.

4.3.5. Granger Causality Test

Measure directional effects amongst variables, especially how the independent
variables lead to changes in TEN.
The study employs a combination of advanced econometric techniques to explore the
relationship between digital transformation and digital entrepreneurship thoroughly.
It utilises descriptive statistics, unit root and co-integration tests, along with quantile
regression and Granger causality analysis. Through these methods, the research seeks
to uncover both short-term and long-term dynamics, thus enhancing the robustness
and reliability of its findings.

5. Results and Discussion
The initial analysis in this study begins with a descriptive analysis, as presented in
Table 1. The sample average, standard deviation, maximum, minimum, skewness,
kurtosis, and Jarque-Bera (JB) normality test are provided. Such statistics will provide
a basic idea of the distributional characteristics of data across OECD countries during
2004-2025.

Based on the Jarque-Bera test, the majority of variables, including ICT exports,
Internet penetration, R&D expenditure, and the number of researchers, are generally
not distributed (p < 0.01). Such confirmation of the normality of variables is
imperative for empirical analysis, since non-normal observations can skew regression
estimates and hypothesis tests (Gujarati & Porter, 2009).

69