Public comprehension of blockchain technology, especially concerning its societal implications, is a relatively underexplored area. This study seeks to fill this void by investigating how such technological advancements might influence consumer purchasing decisions. Specifically, it examines whether individuals are inclined to buy “social” craft beers (produced by social farms) that offer transparent details regarding their origin, social aspects (including the involvement of women who have experienced domestic violence in production), and processing techniques. To achieve these findings, a survey based on a modified Technology Acceptance Model within the Italian context was carried out. The results demonstrate that information on the origin of craft beer, its social impacts and processing methods should provide reciprocal advantages among Italian consumers and social farms. In this sense, blockchain would be a useful tool to generate knowledge both to support the economic inclusion of female victims of domestic violence, and to improve social farms contexts. In other words, blockchain technology should offer transparent information regarding the origin of beer, its social aspects (supporting women economic inclusion) and processing techniques, thereby providing the end consumer with a high-quality product.
This study examines the roles of FinTech development and green finance in inclusive green growth and whether government effectiveness mediates these relationships. The study uses fixed effects with Driscoll–Kraay standard errors, cross-sectionally augmented autoregressive distributed lag, augmented mean group, system GMM, quantile regression, and panel threshold models to address cross-sectional dependence, heterogeneity, persistence, endogeneity, and nonlinear effects. The measurements of inclusive green growth include the Inclusive Green Growth Index, which quantifies digital financial innovation, sustainable capital placement, and institutional capability. The results show that both FinTech and green finance have positive and significant effects on inclusive green growth in both static and dynamic specifications. Government effectiveness increases the direct effects of inclusive green growth and enhances the impact of both FinTech and green finance. The quantile and threshold results also indicate that these effects are stronger in countries with higher performance and governance above the thresholds. This study endorses a conditional leapfrogging process propelled by digital and green finance and competent institutions in developing countries.
This study examines corporate investment and dividend payout responses to dividend tax abolition and whether investment responses vary with firms’ policy exposure and internal financing capacity in an emerging-market setting. Using an unbalanced panel of 452 Indonesian listed non-financial firms over 2017–2024, we first employ a before–after specification to examine post-reform changes in investment and dividend payout. We then use a continuous ownership-exposure difference-in-differences (DID) design to assess whether investment responses vary with firms’ pre-reform domestic ownership exposure, followed by a continuous difference-in-difference-in-differences (DDD) specification to examine whether these responses depend on pre-reform internal financing capacity. The results provide no evidence of higher corporate investment in the post-reform period, whereas dividend payouts are significantly higher. Moreover, firms with greater pre-reform domestic ownership exposure do not exhibit stronger investment increases, and there is no detectable evidence that the ownership-related investment response is stronger among firms with more limited internal financing capacity. These findings remain broadly unchanged across alternative reform timing, ownership-exposure measures, and financing-capacity measures. Overall, the evidence is more consistent with the new-view prediction and suggests that Indonesia’s dividend tax abolition is reflected more clearly in corporate payout than in investment or capital reallocation.
Recently, image construction is a salient topic in discourse analysis, but scholars seldom investigated the relationship between lexis and the images constructed by it. Accordingly, this study employed corpus-based methods to examine the news on Guangdong, China, in American newspapers, in order to delve into the semantic categories of lexis utilized to construct Guangdong’s images and the categories of the images. The research results showed American newspapers exploited 18 semantic categories of lexis to construct 51 categories of discursive images (6 macro images) for Guangdong. At the surface level, the lexis utilized to portray Guangdong exhibited diversity in semantic categories, while at the deep level, the discursive images converged primarily on economic images. American newspapers constructed Guangdong’s images primarily through an economic lens, as economy might constitute the most critical dimension of Guangdong society. This study introduced the analysis of lexical semantic categories into the image construction research, thereby extending the conceptual scope and application boundaries of image construction research.
International organizations need to monitor large amounts of economic and financial data to prevent or uncover potential problems in policies implementation. The analysis of such time series cannot ignore the potential presence of anomalies and structural changes. In this paper, we elaborate on a robust framework for time series analysis based on Least Trimmed Squares able to treat outliers and points where a change in level takes place. We enhance the flexibility of the model by introducing new terms in its definition and extend its applicability to cases with missing observations and multiple level shifts. Moreover, we study its properties, propose a variable selection procedure and introduce instruments for its use in operationally intensive environments requiring accurate and stable outcomes. We demonstrate its potentialities in simulation studies and applications to concrete cases related to trade policies of major relevance for the European Union, such as circumvention of sanction by rerouting of trade flows.
Zakat distribution remains predominantly based on unconditional cash transfers, which often provide short-term financial relief but have limited capacity to promote sustained behavioural change and human capital development among recipients. This study aims to develop and validate behavioural conditionalities for Zakat Conditional Cash Transfer (EZCCT) programmes as a mechanism to strengthen the effectiveness and transformative impact of zakat distribution. A quantitative instrument development approach was employed to assess the content and face validity of a 27-item behavioural conditionality instrument covering education, health, and employment dimensions. Thirteen experts from academia, policy, and practice evaluated the instrument using the Content Validity Ratio (CVR), Content Validity Index (CVI), Randolph’s kappa, and Face Validity Index (FVI). The findings indicate strong validity and expert consensus across all measures. CVR values ranged from 0.85 to 1.00, I-CVI values from 0.92 to 1.00, S-CVI/Ave reached 0.99, Randolph’s kappa recorded 0.954, and S-FVI/Ave achieved 0.98. All 27 items were retained in the final instrument. Theoretically, the study extends Conditional Cash Transfer (CCT) scholarship into the domain of Islamic social finance by integrating Maqasid al-Shariah and behavioural conditionalities within zakat distribution. Practically, the validated instrument provides policymakers and zakat institutions with an evidence-based framework to implement conditional zakat transfers aimed at strengthening education, health, employment participation, and long-term self-reliance among asnaf. The originality of this study lies in developing and validating a behavioural conditionality instrument specifically designed for EZCCT programmes, thereby offering a novel pathway for transforming zakat distribution from assistance-oriented support towards sustainable human development.
This paper investigates the pro-cyclical relationship between business cycles and Lifespan Inequality, specifically focusing on the causal effect of the 2008 Great Recession on this measure across 26 European countries. To the best of our knowledge, the research is the first to prioritize economic crises as a key determinant in assessing life expectancy disparities. The study employed a Difference in Differences methodology, categorizing countries into treatment and control groups based on their exposure to the Great Recession, measured through variations in unemployment rates. This analysis revealed a significant, beneficial effect of the Great Recession on Lifespan Inequality for approximately three years, signifying a reduction in lifespan disparities and suggesting potential health benefits in affected countries (in line with the so-called Thomas Effect). The manifestation of this effect aligns congruently with the intensity of the crisis endured by individual countries.
Household waste management practices represent a key behavioral entry point to the waste-to-food nexus, yet the social, digital, and community factors associated with these practices — and their perceived household-level outcomes — remain underexamined at the behavioral level. This study examines the structural associations among social influence (SOC), digital exposure (DIG), perceived behavioral control (CAP), community programme participation (CPP), household waste management behavior (BEH), and perceived food-related and economic benefits (PFEB) in urban and rural West Java, Indonesia. All constructs were measured through self-reported Likert-scale items; findings reflect perceived associations rather than objectively measured waste volumes, compost yields, food production quantities, or experienced food insecurity. A cross-sectional survey was conducted among 125 households (63 urban, 62 rural) across Kota Bogor, Kota Bekasi, Kabupaten Bogor, and Kabupaten Bekasi, selected to represent contrasting BPS-classified urban and rural administrative units. Data were analyzed using PLS-SEM and Multi-Group Analysis (MGA) with prior statistical power verification (G*Power, f2 = 0.15, α = 0.05, power = 0.80). Household waste management behavior shows a strong positive association with PFEB (β = 0.663, p < 0.001), positioning it as the central structural pathway. Community programme participation is the strongest predictor of behavior within this sample (β = 0.463, p < 0.001), followed by digital exposure (β = 0.211, p = 0.031) and perceived behavioral control (β = 0.202, p = 0.020). Social influence shows no significant unique association with behavior (β= −0.083, p = 0.381), consistent with a possible suppression pattern involving shared variance with community programme participation (r = 0.685); it should not be interpreted as evidence of behavioral irrelevance. Indirect association analysis confirms that waste management behavior functions as a structural pathway from community programme participation, digital exposure, and perceived behavioral control to PFEB; no significant indirect association is observed for social influence. Multi-group analysis detected no statistically significant structural differences between urban and rural groups; given the limited per-group sample sizes (n = 63 and 62), this result is exploratory and does not constitute formal evidence of structural equivalence. The findings support a behavior-centered framework for the perceived waste-to-food nexus in which community programme participation shows the strongest unique association with household waste practices. Community-based programs and digital engagement show complementary associations with household waste-related practices. These findings should be interpreted within the constraints of a cross-sectional, self-report design with a purposively selected sample from a single province.
Variable selection is a critical step in building robust and interpretable linear regression models. With the increasing complexity and dimensionality of datasets, identifying the optimal set of explanatory variables is paramount. Including irrelevant predictors leads to overfitting and reduced interpretability, while omitting crucial ones undermines estimation accuracy and inferential validity. This article provides a comprehensive guide to the theory, methods, and applications of variable selection in linear regression. It aims to equip researchers and data analysts with the knowledge to select and implement the most appropriate technique for their specific dataset, whether it involves a few covariates or high-dimensional data, thereby enhancing model predictive power and interpretability. The methodological framework of this tutorial encompasses a systematic examination of four principal classes of variable selection techniques. These include: (1) subset selection algorithms—best subset, forward selection, backward elimination, and stepwise selection—valued for their interpretability in low-dimensional contexts; (2) shrinkage/regularization approaches—Ridge regression (L2 penalty), Lasso regression (L1 penalty), and the hybrid Elastic Net—which are robust to multicollinearity and scale effectively to high-dimensional problems; (3) derived input direction methods, such as Principal Component Regression (PCR) and Partial Least Squares (PLS), which transform predictors to address multicollinearity; and (4) multivariate extensions, including Group Lasso and Multivariate Shrinkage Regression, designed for scenarios with multiple, correlated response variables. The choice of a variable selection strategy profoundly impacts model validity, coherence, and generalizability. There is no one-size-fits-all solution; the optimal method depends on the data structure, research goals, and computational resources. Subset methods offer transparency, shrinkage methods provide stability and selection in high dimensions, and dimension reduction techniques uncover latent structures. This guide concludes that mastering this portfolio of techniques, supported by modern software in SPSS, R, Stata, and Python, empowers analysts to build models that are not only statistically sound but also meaningfully actionable, blending scientific rigor with domain expertise.
This paper addresses the critical structural anomalies that comprehensive emerging universities encounter when navigating global ranking methodologies (e.g., QS and THE) to improve their global standings. This paper formalizes and validates the Structural Dilution Phenomenon—a baseline institutional distortion distinct from disciplinary field normalization—wherein the statistical aggregation of highly cited Science, Technology, Engineering, and Mathematics (STEM) departments with teaching-intensive Humanities and Social Sciences (HSS) faculties lowers the core institutional metric of Citations per Faculty (CpF). Utilizing longitudinal bibliometric profiles from King Saud University (Scopus 2019–2024; p < 0.001, Cohen”s d = 1.84) and Nanyang Technological University (2010–2025), a prescriptive mathematical framework is constructed. Sensitivity testing demonstrates a modeled inflection point at 30
Commodity price projections have long been a major source of reliance for both the government and investors. The difficult problem of predicting the daily regional steel price index in the central south Chinese market from January 1, 2010 to April 15, 2021 is examined in this paper. The literature has not given adequate attention to the forecast of this important commodity price indicator. Gaussian process regressions are adopted as the forecasting model and they are trained based upon the combination of Bayesian optimizations and cross validation. The models that were constructed had an out-of-sample relative root mean square error of 0.5370
Private pension plans have been structured as a retirement supplement and have received fiscal incentives in most developed countries. Investor preferences regarding different tax treatments within the Spanish pension system play a significant role in shaping long-term investment behavior. In this context, the aim of this study is to analyze investor preferences in private pension plans and how socioeconomic variables affect the distribution of these preferences. To this end, a choice experiment was conducted with a sample of 1,287 Spanish pension plan investors. The econometric model applied is the mixed logit, which allows for the estimation of individual preference parameters and the analysis of their distribution. The main findings reveal a generally low preference for tax-related attributes in pension plans, with marked differences when variables such as age, gender, income, household size, and educational level are considered. These results provide a foundation for the development of appropriate strategies both for financial planners and advisors, as well as for the design of economic policy in terms of taxation.
Few studies examine financial globalisation and terrorism, and fewer distinguish between its de facto (actual cross-border flows) and de jure (regulatory frameworks) dimensions in Nigeria; a distinction that matters for Nigeria, where Boko Haram and ISWAP exploit informal and digital financial channels. This study examines the associations between financial globalisation, internet usage, unemployment, economic growth, and terrorism incidents in Nigeria over 1991–2021, using an ARDL–ECM framework complemented by Dynamic OLS and frequency-domain causality tests. The results point to a differentiated pattern: overall and de jure financial globalisation exert a long-run terrorism-mitigating effect, whereas de facto financial flows show no statistically significant association with terrorism incidents; internet usage is associated with higher terrorism over longer horizons but lower terrorism in the short run; unemployment exhibits a temporally contingent relationship—reducing terrorism in the long run but increasing it in the short run; and economic growth is associated with lower terrorism through long-run structural channels. In sum, the findings suggest that the security implications of global integration in Nigeria depend less on the scale of openness than on its regulatory form and timing, with sustained growth and formal financial oversight playing a central role in constraining terrorist activity.
Digital transformation has become a central agenda in public administration, yet the literature remains fragmented across e-government, digital governance, public-sector innovation, and AI-related studies. This study examines research on digital transformation in public administration during 2018–2025 using a hybrid design that combines bibliometric analysis and a nested systematic literature review. Using Scopus data and a PRISMA-guided selection process, 149 records were included in the bibliometric corpus and 45 studies in the nested review. The findings show substantial post-2020 expansion of the field, with publication output, citation visibility, and international collaboration unevenly distributed across the retrieved sources, countries, and networks. The thematic structure broadened from e-government and service-delivery concerns toward governance transformation, organizational change, interoperability, AI-enabled administration, resilience, and citizen-centered reform. Methodologically, the field is theoretically plural and analytically multi-scalar, while longitudinal, comparative, multi-level, and outcome-oriented designs remain comparatively less developed. Recurrent constraints extend beyond technology availability and include governance capacity, organizational and workforce readiness, fragmented data systems, interoperability, digital inclusion, and AI-governance capability. The literature also reports potential or documented opportunities for more accessible, integrated, resilient, data-informed, and responsive public administration under supportive institutional and organizational conditions. Overall, digital transformation is increasingly understood as a multidimensional administrative and governance process rather than a narrow technological upgrade. These conclusions apply to the Scopus-bounded corpus examined and indicate continuing research needs concerning governance capacity, inclusion, interoperability, and trustworthy AI.
Smoking remains a major public health concern due to its persistent spread through social influence and its long-term consequence in the development of life-threatening diseases such as cancer. This study seeks to suggest a fractional-order mathematical model to investigate the coupled dynamics of smoking behaviour and smoking-induced cancer, incorporating nonlinear peer influence, relapse, and multiple recovery pathways. Smoking-related cases were obtained from Norway between 2010 and 2024 and were compared to the model’s solution. Through a quantitative analysis, the model is proven to be positive and bounded. By further employing the fixed point theory, Lipschitz criterion, Hyers-Ulam, and Hyers- Ulam-Rassias conditions, the model is shown to be well-posed. The smoking threshold number (ℛ_0) was calculated and further shown that whenever ℛ_0 < 1 , smoking behavior is minimized over time, whereas the smoking behaviour persists when ℛ_0 > 1 . Local and global sensitivity analysis were performed by employing three-dimensional plots, contour plots, and partial rank correlation coefficient (PRCC) plots to ascertain how the model’s parameters influence the smoking threshold number. The sensitivity analysis explicitly shows that without any control approach implemented, parameters like α _1, α _2 and ω elevate the active smoker population, which results in an increase in the smoking-induced cancer compartment over time. On the other hand, the quitting rate ρ when enhanced will help mitigate the smoking behavior persistence in the Norwegian population. Finally, it has been clearly shown through numerical simulations that the fractional-order operator exerts enormous effects on the smoking behaviour dynamics in the population. Specifically, strong memory effects (lower fractional order values) lead to a slower accumulation of active smokers and a delayed but persistent increase in smoking-induced cancer cases, whereas weak memory effects (fractional values close to unity) exhibit an acceleration of both smoking progression and smoking-induced cancer disease persistence.
Recent advances in Generative Artificial Intelligence have increased interest in using Large Language Models (LLMs) to generate synthetic populations for computational social science and decision research. However, many existing applications rely on loosely specified personas, limiting theoretical interpretability and reproducibility. This study proposes and validates the Grounded Synthetic Generation (GSG) strategy combined with a novel Digital Habitus theoretical framework, demonstrating how structural demographic variables inherently constrain the outputs of LLM-simulated personas. We simulated an ambiguous corporate bonus scenario under severe outcome uncertainty across 500 distinct Italian socio-economic profiles using Gemini 3.1 Pro Preview. The generative system was constrained by 12 distinct demographic vectors (including Class, Personality, Dreams, and Fears). Advanced statistical analysis reveals a strong association between socioeconomic class and simulated decision criteria (Cramér’s V = 0.484, p < 0.001), alongside significant associations for primary fears (V = 0.224), education (V = 0.183), and personality traits (V = 0.161). The study maps these findings back to classic Decision Theory criteria (e.g., Laplace, Wald, Savage) via discrete output choices. This study details practical protocol recommendations for full reproducibility and delineates clear policy implications for addressing algorithmic bias in generative agent architectures.
Understanding the structural organization of conflict is essential for designing effective regional conflict strategies. While previous studies have predominantly identified conflict hotspots using incident frequency or spatial density, these approaches often overlook the structural relationships among geographically connected regions. This study proposes an integrated graph-based framework for identifying structural conflict hotspots in Papua by combining spatially weighted network construction, graph-theoretic feature extraction, unsupervised clustering, and comprehensive statistical validation. Verified conflict data from Human Rights Monitoring (HRM) and Komnas HAM, covering the period 2018–2024, were aggregated at the district level. A spatial weighted undirected network was constructed using administrative adjacency and conflict intensity, from which degree centrality and betweenness centrality were extracted as structural descriptors. Districts were subsequently clustered using K-Means based on three standardized features: total verified conflict incidents, normalized degree centrality, and normalized betweenness centrality. The number of clusters was selected using the Gap Statistic one-standard-error rule, while the Silhouette Coefficient (SC), Davies-Bouldin Index (DBI), Calinski-Harabasz Index (CHI), bootstrap Jaccard similarity, Kruskal-Wallis tests, and Moran’s I were used as complementary validation procedures. DBSCAN was additionally applied as a complementary density-based analysis rather than as a criterion for selecting or confirming the number of K-Means clusters. The resulting network comprised 40 districts connected by 53 spatial relationships, with a density of 0.0679 and a modularity of 0.5689. The Gap Statistic one-standard-error rule retained K = 2 as the parsimonious K-Means solution, yielding an SC of 0.5978, a DBI of 1.0189, a CHI of 31.6455, and a bootstrap Jaccard similarity of 0.8184. Kruskal-Wallis tests further showed significant between-cluster differences in degree centrality (H(1) = 15.1714, p < 0.001) and betweenness centrality (H(1) = 24.3471, p < 0.001). Significant positive spatial autocorrelation (Moran’s I = 0.5629, p = 0.0005) indicated that similar conflict-structural profiles exhibited a non-random spatial organization. Unlike conventional hotspot mapping based solely on conflict intensity, the proposed framework integrates observed conflict magnitude with local network connectivity and brokerage position, thereby characterizing districts according to both conflict intensity and their structural roles within the regional conflict network. This integrated methodology provides a transparent and reproducible approach for structural conflict analysis and offers a robust analytical foundation for future spatial network modelling and conflict early-warning research. The proposed workflow is fully reproducible and can be adapted to other regional conflict systems where only district-level conflict records are available.
This study explores the mechanisms of culinary co–creation in Muslim–friendly tourism experiences in Hoi An, a UNESCO World Heritage City. It analyses the interactions among three primary stakeholders: Muslim tourists, service providers, and local communities, focussing on their collaborative value creation process. The study utilises a mixed-methods design that includes: (1) in-depth interviews with 50 participants from six stakeholder groups, (2) four focus group discussions consisting of 6–8 participants each, (3) participant observation at 12 representative sites, and (4) a systematic analysis of 87 scholarly articles sourced from the Scopus database covering the years 2019 to 2025. The analysis of data was performed utilising NVivo 14 software, adhering to a thematic analysis methodology. The results identify three main co–creation channels: direct interactions, which account for 67
This comprehensive guide addresses a gap in methodological literature by synthesizing quantitative, qualitative, and mixed research approaches within a structured framework. It bridges theoretical foundations, practical applications, and tool selection—a resource often lacking in traditional texts. The article presents comparative tables and a decision-support tree outlining core objectives, analytical strategies, and compatible software tools. Implementation is illustrated through methodological case studies focused on administrative and organizational contexts. Designed as both an accessible roadmap for novice researchers and a consolidated reference for advanced users, this guide supports rigorous and informed methodological decisions across research settings. While comprehensive in scope, it serves as a starting point rather than an exhaustive manual, encouraging readers to adapt and expand the framework to their specific research needs.
Early dropout remains a persistent challenge in higher education and is particularly salient in fully online learning environments, which are expanding across national systems. Despite this growth, empirical evidence on early withdrawal in fully online universities remains limited. Drawing on administrative data from a national Italian online university, this study examines early dropout among first-year students enrolled in five bachelor’s programs during the 2024/2025 academic year (N = 652), contributing to the international literature on student persistence in digitally mediated higher education. A two-part modelling strategy is adopted: logistic regression is used to estimate the probability of early withdrawal, while a Cox proportional hazards model analyses the timing of dropout among students who exited (n = 146). Results indicate that academic engagement, operationalized through exam completion and participation in synchronous lessons, is a strong negative predictor of dropout, reducing the likelihood of withdrawal, consistent with findings from both online and campus-based higher education systems internationally. By contrast, among students who withdraw, engagement variables lose significance, and age emerges as the primary predictor of withdrawal timing, with older students exiting earlier. These findings underscore the importance of distinguishing between the determinants of dropout occurrence and dropout timing, and they highlight the need for engagement-oriented instructional design and differentiated support strategies for mature learners in fully online higher education. The study offers empirically grounded insights with relevance for retention research and policy across diverse higher education systems undergoing digital transformation.
