Artificial intelligence (AI) is increasingly recognised as a key driver of business innovation, yet its adoption among small and medium-sized enterprises (SMEs) varies considerably. This study examines whether AI Capital, defined as AI-related knowledge, skills and capabilities, is associated with business innovation among SMEs in England. Using a two-wave longitudinal panel dataset comprising 504 observations from SMEs collected in 2024 and 2025, the study develops and validates a 45-item AI Capital of Business scale. Business innovation is measured across five dimensions: product and service innovation, process innovation, technology adoption, market and customer engagement, and organisational culture and strategy. Regression models, including pooled OLS, Random Effects, and Fixed Effects specifications, are employed. The findings reveal a robust positive association between AI Capital and business innovation across all model specifications. This association holds across all business innovation dimensions and remains consistent for SMEs with differing levels of financial performance, size, and operational maturity. Each component of AI Capital independently exhibits a positive association with business innovation outcomes. The results highlight the central role of AI Capital in enabling SMEs to translate AI adoption into tangible business innovation. From a policy perspective, the findings indicate the value of targeted interventions that prioritise AI upskilling, organisational capability development, and accessible support mechanisms to promote inclusive and sustainable AI-driven business innovation among SMEs. • Developed and validated a 45-item AI Capital for Business scale. • Longitudinal SME study links AI Capital to stronger innovation outcomes. • AI Capital boosts product, process, market, and strategic innovation. • Robust panel data confirms AI Capital as a driver of SME competitiveness. • Policy insights support targeted AI training and inclusive SME innovation.

Experiences of nature are often appraised as interesting, novel or perspective-changing, with both experimental and qualitative work invoking these constructs, particularly regarding biodiversity. However, these psychological outcomes of interactions with nature are seldom investigated quantitatively, perhaps due to the lack of a coherent conceptualisation. Drawing on ‘psychological richness’ – an account of subjective well-being from positive psychology which unifies these constructs – we explored the predictors of psychologically rich nature visits. Using data from English adults taken from 10 years of the Monitor of Engagement with the Natural Environment survey (n=9,827), we used exploratory and confirmatory factor analysis to identify a set of survey items pertaining to the psychological richness experienced after recent nature visits and subsequently constructed linear regression models which explored sociodemographic and visit-related variables that predicted this. To demonstrate the discriminant validity of the construct, we further examined comparable associations with survey items concerning enjoyment and perceived restorativeness. We found that ratings of psychological richness were higher on nature visits taken by people who were female, older/retired, lower socioeconomic status, non-white-British, non-disabled, car-free respondents, and on visits taken to community gardens, farmland, woodlands, uplands, inland waters, country parks, and designated sites, involving wildlife watching or fishing, and taken further afield or while on holiday. These findings were largely robust to two alternative model specifications. Associations differed in strength, sign, or statistical significance when considering enjoyment/restorativeness. We argue that these findings demonstrate that psychological richness overlaps with, but is distinct from, related psychological outcomes of nature experience. We discuss the results with regards to socioeconomic inequalities and biodiversity, and present research avenues for the continued exploration of the psychological richness of nature experience. • We posit ‘psychological richness’ as an understudied appraisal of nature experience • We studied its predictors in a large sample of voluntary nature visits from England • Higher ratings of psychological richness were given by marginalised populations • Higher ratings were also associated with biodiverse environments and activities • Differing links with enjoyment/restorativeness indicate discriminant validity

A promising avenue for improving physical health and reducing health risks in older adulthood is intervening on transdiagnostic protective factors-those that form the mechanisms of several diseases. A potential candidate for a transdiagnostic protective factor is subjective well-being (SWB). Despite the accumulating evidence favoring SWB as a protective factor, individual studies tend to neglect the cumulative effect of publication bias, explicitly specifying a causal model, and the possibility of alternative causal models that guide the selection of confounders. Using three waves of longitudinal data from the Canadian Longitudinal Study of Aging with a sample of 51 338 older adults, we tested whether the relationship between life satisfaction and 23 health outcomes (eg, cancer, arthritis, asthma) were robust to alternative causal model specifications. A one standard deviation increase in life satisfaction, under the preferred causal model, predicted a 3-31% reduction in risk of diagnosis for several health outcomes three years later, including osteoarthritis, movement disorders, and respiratory tract diseases and predicted greater self-rated general health. Importantly, the effects persisted under different causal model specifications. These results suggest that life satisfaction is prospectively linked to multiple positive health outcomes and demonstrate the value of an analytical framework that acknowledges the uncertainty inherent in confounder selection.

Most firearms recovered from crime scenes in Mexico originate in legal U.S. retail markets, yet little is known about whether federal enforcement constrains this diversion. This article examines how compliance citations issued by the Bureau of Alcohol, Tobacco, Firearms, and Explosives (ATF) affect subsequent firearms trafficking from U.S. Federal Firearms Licensees (FFLs) to Mexico. ATF inspection-violation records are merged with more than 12,000 firearms traced from Mexican seizures and U.S. trafficking court cases, exploiting within-dealer variation over time. Across multiple model specifications, ATF citations are strongly associated with reductions in trafficking—each additional citation corresponds to a 20–44% decline in trafficked firearms, and cited dealers contribute substantially fewer trafficked guns than comparable uncited retailers. These results suggest that even limited regulatory enforcement can meaningfully disrupt illicit firearms supply chains upstream of cross-border smuggling.

This article provides a roadmap to explore the relationship between school quality and median single-family town-level home prices, school, and neighborhood characteristics using a unique dataset consisting of 152 towns in Greater Boston. Before the session, we ask students to read a well-known paper exploring the relationship between school quality and home prices. We discuss the paper in class and provide students with a new dataset containing similar information. As part of the class session, students consider how to use what they learned from a published paper to practice hypothesis formulation and testing with new data. We conduct simple and multiple regression analysis, interpret results, and estimate alternative model specifications. A goal of the exercise is to encourage students to create theory-driven regressions, rather than looking for a “right” answer. The topic and the dataset are appropriate for students who have taken introductory statistics and can be used in courses that include content related to educational achievement and financing, differences in socioeconomic status, and economic opportunities. Based on the topic’s broad appeal and accessibility, instructors could use the dataset to practice interpreting results of regression analyses without additional prework or context. We share the new dataset and accompanying material online.

Abstract It is increasingly common to study mobility and migration of individuals between social and physical locations as networks in which locations are nodes connected by mobile people. This conceptualization as mobility networks facilitates the analysis of how individuals influence one another in their mobility destinations. Technically, this amounts to analysing interdependence between individuals’ mobility paths. A recently proposed framework—the endogenous log-linear model (ELMo)—allows the statistical modelling of these social processes and, therefore, dependence in mobility, combining insights from exponential random graph models (ERGMs) and log-linear models. However, little attention was paid to how such models should be specified in a principled, theoretically informed way. In this study, we apply statistical theory to propose model specifications that can be used to analyse emergent structures in mobility. We first reformulate the model under analysis as a conditional multinomial logit with dependent observations. Subsequently, we show how to specify models that (i) are based on clear dependence assumptions on the individual level, that (ii) have a clear individual level interpretation, and that (iii) avoid (near-)degeneracy, a common problem for models with dependent observations. We end with an example application pertaining to the mobility of computer science faculty between university departments.

Abstract We propose a statistical model of heterogeneous beliefs wherein investors are represented as different machine learning model specifications. Investors form return forecasts from their individual models using common data inputs. We measure disagreement as forecast dispersion across investor-models (MFD). Our measure aligns with analyst forecast disagreement but more powerfully predicts returns. We document a large and robust association between belief disagreement and future returns. A decile spread portfolio that sells stocks with high disagreement and buys stocks with low disagreement earns a value-weighted return of 13% per year. Further analyses suggest MFD-alpha is mispricing induced by short-sale costs and limits-to-arbitrage.

This study empirically investigates whether AI-based technology valuation produces different outcomes compared to conventional expert-led evaluation methods. The Korea Technology Finance Corporation (KOTEC), a public institution that supports SME financing through credit guarantees, has implemented an AI-driven valuation system known as KPAS to assess firms’ intellectual property. To evaluate the effectiveness of this AI-based approach relative to traditional expert appraisals, we examine changes in two key performance indicators: firm sales, which reflect current business performance, and the Tech Index, which serves as a proxy for innovation capacity. Using a difference-in-differences (DID) framework combined with propensity score matching (PSM), we compare the post-guarantee performance of firms evaluated by each method. The results show that guarantees based on technology valuation improve firm performance overall, and that KPAS substantially reduces the time and cost of evaluation compared to expert appraisal. Despite its lower resource requirements, the AI-based method delivers performance outcomes that are comparable to, or even better than, those of traditional evaluations. These findings remain largely robust across alternative model specifications and offer empirical support for the broader adoption of AI-assisted valuation systems in technology finance.

The study aims to analyze the impact of exchange rate fluctuations and other macroeconomic factors on Pakistan’s sovereign credit ratings from 2014 to 2014. While exchange rate fluctuations are generally considered a key fluctuation are generally considered a key determinant of sovereign risk in emerging markets, their direct impact on credit ratings remains ambiguous. To address this issue, this study employs time-series econometrics methods based on ordinary least squares (OLS) and generalized method of moments (GMM) to account for dynamic relationships and potential endogeneity issues. The empirical analysis has considered a wide range of macroeconomic variables, such as exchange rate variations, exchange rate volatility, inflation, interest rate differentials, GDP growth rate, foreign exchange reserves, and external balance indicators. The empirical findings of the present study reveal that inflation is the most critical and significant determinant of sovereign credit ratings; however, its impact is negative and statistically significant across all model specifications. Moreover, the present study also reveals that the impact of the lagged dependent variable is also statistically significant across all model specifications, indicating the presence of persistence in sovereign credit ratings over time. On the other hand, the impact of exchange rate fluctuations and exchange rate volatility is not statistically significant across all model specifications; therefore, these variables may be having an indirect impact on sovereign credit ratings through other macroeconomic factors. Diagnostic tests of the empirical model of the present study confirm the validity and stability of the model; therefore, the empirical findings of the present study are reliable and valid for drawing conclusions regarding the impact of various macroeconomic factors on sovereign credit ratings for the period of 2014-2024 in the context of Pakistan’s economy. The empirical findings of the present study highlight the relative importance of inflation over exchange rate fluctuations in determining sovereign credit ratings; therefore, the present study is important for policymakers and practitioners as well as for academicians and researchers interested in the subject of sovereign credit ratings and their determination through various macroeconomic factors.

The relevance of this study arises from Mongolia’s emerging retail investment landscape, where public participation in the stock market remains limited. The insufficient understanding of individual investment behavior complicates efforts to enhance financial inclusion. Examining socio-demographic and behavioral characteristics helps identify the incentives and constraints shaping investment decisions and strengthens the foundation for long-term financial culture. The purpose of the study is to provide a comprehensive assessment of how socio-demographic and behavioral factors influence both the likelihood of participating in the stock market and the amount invested, as well as to clarify the mechanisms through which differences in investment behavior emerge. Objectives. The objectives include evaluating the effects of gender, age, household size, and income on the probability of purchasing stocks; assessing how investment horizon, education, prior experience, and specialized training affect investment amounts; and comparing model specifications to identify stable determinants of investment activity. Methodology. The methodology relies on a dataset of 868 respondents collected over sixteen days, of whom 519 were identified as active investors. A probit model is applied to estimate participation probabilities, while a Tobit model examines the determinants of investment amounts. Results. The findings show that gender, age, household size, and income significantly influence participation decisions, whereas income, gender, investment horizon, education, experience, and training strongly increase investment volume. Conclusions. The study highlights the need to expand financial education, strengthen practice-oriented training, and promote long-term planning as key measures to increase public engagement and support the development of a more robust investment environment in Mongolia.

On the associations between adolescent social media use and health outcomes: An exploratory specification curve analysis and comparison with other concurrent predictors.

Warm, dry environments create living conditions that challenge plant growth, reproduction and survival. Plants in these environments have evolved adaptive strategies to enhance water-use efficiency and ensure reproductive success, two of which are annuality and C4 photosynthesis. However, life history variation is rarely included in large-scale studies of plant diversity, and the extent to which these traits coevolve and are jointly selected for is not known. To address this, we used Pagel's models of independent and correlated evolution for over 4000 species of grasses (Poaceae), while accounting for evolutionary rate heterogeneity and potential type I statistical errors. We found that there are more C4 than C3 annuals and that C4 origins predate evolution of annuality, but no support for correlated evolution between the two traits. Our results indicate that any habitat or trait similarities (e.g. small seeds, fast growth) between annuals and C4 species reflect independent adaptations to similar environmental conditions or are contingent on the two traits themselves, rather than the result of evolutionary or functional links between them. Our results further highlight the importance of appropriate null model specification for testing evolutionary hypotheses across large, old clades.

Likelihood-free parameter inference for spatiotemporal stochastic biological models using neural posterior estimation.

This paper aims to provide evidence of the impact on the minimum wage to employment in Greece over the period 2016 to 2024. The main contribution of this paper is the examination of the effects of the minimum wage during a period characterized by many difficulties and research interest not only nationwide but also across regions with high heterogeneity. The case of Greece is particularly interesting to study during this period as it provides a unique context to explore the effects of minimum wage increases on employment. Greece constitutes a distinctly singular case within the European context due to the exceptional structural characteristics of its labor market. Following a protracted economic crisis, successive waves of labor market reforms, and the additional disruptions generated by the COVID-19 pandemic, Greece provides an illustrative, and in many respects unique, example of how extensive policy interventions interact with a gradually recovering economy and persistently elevated unemployment levels. Overall, the results strongly indicate that there is little to no impact of the minimum wage on employment and the findings vary considerably across the different regional contexts. Finally, the DiD methodology used supports the credibility of the findings and suggests that the lack of impact of the minimum wage is not due to model specification or timing bias.

The growing emphasis on sustainability reporting around the world has heightened interest in Environmental, Social, and Governance (ESG) among researchers and regulators. Disclosure as a determinant of business performance. In accordance with current sustainability reporting advancements led by the International Sustainability Standards Board, firms in emerging economies are under rising pressure to expand transparency beyond standard financial indicators. However, there is still insufficient and conflicting empirical data regarding the relationship between ESG performances in Sub-Saharan Africa. The study examined the impact of ESG disclosure on the financial performance of Nigerian listed manufacturing firms. The study uses balanced panel data from the annual and sustainability reports of manufacturing companies listed on the Nigerian Exchange Group between 2016 and 2025, using an ex-post facto research design. ESG disclosure is measured through a content analysis–based index constructed in accordance with Global Reporting Initiative (GRI) standards, while firm performance is peroxide by Return on Assets (ROA), Return on Equity (ROE), and Tobin’s Q. To account for Heteroskedasticity and unobserved firm-specific effects, the study uses panel regression techniques, such as fixed and random effects models with robust standard errors. Control variables include firm size, leverage, and firm age. The research findings show that overall ESG disclosure has a statistically significant and positive impact on a company's financial performance, indicating that increased sustainability transparency raises market value and profitability. Disaggregated data show that while environmental and social disclosures have moderate but positive benefits, governance disclosure has the strongest correlation with performance indicators. The outcomes hold up well when using different model specifications and performance proxies. The research enhances the sustainability accounting literature by offering evidence from a growing African market, therefore broadening stakeholder and legitimacy theory perspectives within the ESG performance. This provides useful insights for regulators, investors, and corporate executives, urging the enhancement of ESG reporting frameworks and alignment with global sustainability standards to foster long-term wealth generation and sustainable industrial advancement in Nigeria.

ABSTRACT This study examines whether Asian companies pay higher premiums in cross‐border mergers and acquisitions (M&A) and identifies the institutional factors driving this behavior. Grounded in the concept of Asian institutional logic—characterized by state coordination, relational governance, and long‐term strategic orientation—we argue that these features shape distinctive acquisition patterns compared to Western market logics. Using a large sample of cross‐border M&A during the period 2003–2021, we first uniquely compare whether the geographical origin of the acquirer firm is a relevant determinant of the premium paid, namely for cross‐border operations targeting Asia, Europe, and the United States. We find that Asian acquirers pay significantly higher premiums compared to their European and U.S. counterparts. Employing a moderating effect approach, we find that this relationship is amplified by four mechanisms aligned with the Asian institutional logic. Specifically, we analyze the role of Chinese state‐owned enterprises (SOEs) and find that they contribute significantly to the higher premiums paid. Our results are robust across different model specifications and subsample analyses, shedding light on the distinct dynamics of cross‐border M&A involving Asian firms.

To investigate the association between the C-reactive protein-Albumin-lymphocyte (CALLY) index and obesity in adolescents, and evaluate its predictive performance compared to traditional inflammatory markers. Data from NHANES 2015-2018 were analyzed, including 2188 adolescents aged 12-20years. The CALLY index was calculated as (albumin × lymphocytes)/CRP. Multivariate weighted logistic regression was used to assess associations with obesity. Receiver Operating Characteristic (ROC) curves and Area Under the Curve (AUC) were computed to evaluate the predictive performance of the CALLY index compared to traditional inflammatory markers (NLR, PLR, SII, SIRI). Survey-weighted logistic regression assessed associations of CALLY with obesity; model fit (Nagelkerke R2) and discrimination were evaluated, with sensitivity analyses for model specifications. The CALLY index was inversely associated with obesity (OR = 0.968, 95% CI 0.958-0.977, p < 0.001), with each 1-unit increase reducing obesity odds by 3.20%. This association remained significant after multivariate adjustment and varied by age, race, and physical activity. The standalone AUC of CALLY (0.743) significantly outperformed traditional markers (PLR: 0.515, NLR: 0.546, SII: 0.593, SIRI: 0.598). Combined with covariates, AUC increased to 0.774 (DeLong test p < 0.001). At the cutoff of 15.10, CALLY achieved > 80% sensitivity with a positive likelihood ratio near 2.0. The CALLY index is a robust and independent predictor of obesity in adolescents, outperforming traditional inflammatory markers. Its integration with traditional covariates further enhances predictive accuracy, suggesting that the CALLY index could be a valuable tool for early screening and intervention in adolescent obesity.

The volatility of cryptocurrency markets has increased substantially in recent years, particularly for Ethereum (ETH), which exhibits fat-tailed distributions and persistent volatility clustering that traditional linear models are unable to capture. This study aims to analyze and model the volatility of ETH/USD using high-frequency hourly data to determine the most appropriate volatility model for describing Ethereum’s intraday market dynamics. The dataset consists of 8,760 hourly closing prices from October 31, 2024 to October 31, 2025, obtained through the CryptoCompare API. The methodological framework includes data preprocessing, log-return transformation, stationarity analysis using the Augmented Dickey–Fuller test, detection of heteroskedasticity via the ARCH–LM test, and estimation of several ARCH and GARCH model specifications. The results show that ETH/USD returns are stationary, non-normally distributed, and exhibit clear volatility clustering. Among the ARCH models, only ARCH(1) adequately captures short-term fluctuations, while ARCH(2) provides no additional benefit. In contrast, GARCH models demonstrate superior performance in capturing both short-term shocks and long-term persistence. Based on AIC, BIC, and log-likelihood values, GARCH(1,2) emerges as the best-performing model, offering the highest flexibility in representing Ethereum’s persistent and reactive volatility patterns. These findings confirm that ETH/USD volatility is predictable and can be modeled statistically. Future research may incorporate asymmetric GARCH extensions or external explanatory variables to improve predictive performance.

Isolated non-albumin proteinuria (iNAP) has been linked to kidney function decline in diabetes, but its prognostic role in hypertensive patients with chronic kidney disease (CKD) and normal albuminuria remains uncertain. We evaluated whether baseline iNAP predicts longitudinal estimated glomerular filtration rate (eGFR) decline and kidney events in this setting. In a retrospective cohort of 166 hypertensive CKD outpatients with normal albuminuria, iNAP was defined as total 24-hour proteinuria ≥150 mg/day with albuminuria <30 mg/day. Kidney function change was assessed as (i) eGFR trajectory over 6 years using linear mixed-effects models and (ii) a composite kidney endpoint (≥30% eGFR decline or incident end-stage kidney disease) using cumulative incidence and Fine-Gray competing-risk regression (death as competing event). Confounding control was guided by a directed acyclic graph (DAG), yielding a minimal adjustment set (age, sex, diabetes, on-treatment mean arterial pressure, baseline eGFR). Sensitivity analyses examined alternative model specifications and endpoint definitions. Median follow-up was 5.4 years (interquartile range 4.5-6.0). iNAP was present in 41/166 participants; 32 reached the composite kidney endpoint and 25 died. Baseline eGFR was similar in iNAP versus normal proteinuria (57±24 vs 61±24 mL/min/1.73 m²; p=0.318). In mixed-effects models, eGFR declined by -0.61 mL/min/1.73 m²/year (95% CI -0.96 to -0.25) in normal proteinuria and by -1.53 mL/min/1.73 m²/year (95% CI -2.15 to -0.91) in iNAP, with a between-group slope difference of -0.92 mL/min/1.73 m²/year (95% CI -1.63 to -0.22; time×iNAP interaction p=0.011). In competing-risk analyses, iNAP was associated with a higher risk of the composite kidney endpoint (Fine-Gray subdistribution hazard ratio 3.15, 95% CI 1.58-6.31; p=0.001), while cumulative incidence of death without prior kidney endpoint (the competing event) did not differ between groups (Gray's test p=0.920). Findings were consistent after DAG-minimal adjustment and across sensitivity analyses. In hypertensive CKD patients with normal albuminuria, baseline iNAP is associated with faster eGFR decline and a higher risk of kidney events, independent of key baseline risk factors in DAG-guided analyses.

This study analyzes the dynamic relationship between rice prices and selected economic variables using a Vector Autoregression (VAR) model. The analysis utilizes daily data from January 2022 to December 2023, encompassing rice prices, chicken meat prices, chicken egg prices, the Rupiah-to-USD exchange rate, inflation, and crude oil prices. The estimated VAR model is stable, as all eigenvalues lie within the unit circle. Residual diagnostics based on the Portmanteau (Ljung–Box) test indicate no residual autocorrelation across all equations (LB statistics with df = 1, p-values &gt; 0.05), confirming the adequacy of the model specification. The model demonstrates good predictive performance for the rice-price series, achieving a Mean Absolute Percentage Error (MAPE) of 0.42% over the out-of-sample testing period (the last 20% of observations). Empirical results suggest that rice prices are influenced by dynamic interactions within the system, particularly through their relationships with chicken meat prices and the Rupiah–USD exchange rate. These findings offer valuable policy insights for maintaining rice price stability, a crucial component of national food security.