421 papers
This paper introduces an "inverse meta-analysis" method to recover pre-deletion measurement quality statistics from published studies, revealing that the widespread practice of deleting weak items in the Technology Readiness Index 2.0 artificially inflates reliability metrics and obscures the true weakness of its inhibitor dimensions.
This paper introduces Proof of Scientific Contribution (PoSC), a methodological framework that decouples eligibility-gated scoring from market pricing to allocate compensation for human biological data through mathematically verified, reserve-aware rules while explicitly delineating its scope as a computational specification rather than a validated market or biological solution.
This paper introduces new continuous distributions derived from the Median Based Unit Rayleigh (MBUR) distribution to model maximum extreme values, providing their mathematical properties, asymptotic behaviors (including Gumbel, Fréchet, and Weibull limits), and demonstrating their utility in real-world disaster-related applications.
This comment critiques Mahmood et al. (2025) for methodological errors in their comparison of working capital financing and firm performance, specifically highlighting the misclassification of principal component analysis, the incorrect reporting of linear regression statistics for tree-based models, and the presence of mutually inconsistent estimates that undermine the study's conclusion that machine learning models outperform fixed-effects models.
This preregistered analysis of linked 2020–2022 China Family Panel Studies data reveals that among adults who participated in fitness or leisure activities in 2020, those with 12 years of education were significantly less likely to report nonparticipation in 2022 compared to those with 6 years of education, suggesting that higher educational attainment is associated with sustained engagement in these activities despite the study's noncausal and non-population-representative limitations.
The paper introduces CoSReMA, a novel framework that synthesizes study-level regression coefficients and reconstructs synthetic datasets from summary statistics to accurately estimate both model parameters and overall fit for linear regression meta-analyses without requiring individual participant data.
This paper develops a reference-weighted decomposition method to measure how cited fields relate to authors' publication histories, revealing that citation diversity significantly predicts discovery and first confirmation outcomes while demonstrating that uncovered citation shares are lower than benchmarks and that AI-use associations were not confirmed in a prospective second validation.
This study evaluates position-specific allometric body-mass adjustments for NFL Combine tests across multiple years, finding that while such adjustments significantly improve prediction accuracy for body-transport tasks like sprinting and jumping—particularly for defensive linemen—their benefits are inconsistent across other tests and require careful calibration to account for cohort-specific score offsets.
This paper proposes a regression-based causal inference framework using ordinary least squares with probability-weighted treatment indicators to estimate person-specific immediate and lag-1 effects in within-subject randomized designs, demonstrating its robustness and applicability through simulation studies and empirical examples.
This study utilizes a state-dependent Detrended Cross-Correlation Analysis (DCCA) framework to demonstrate that the relationship between scoreboard pressure and wicket occurrence in IPL run chases is dynamic and non-linear, generally weakening as the required run rate increases.