Considering the limitations of monotherapies due to chemoresistance and side effects, this research aimed to determine whether low doses of sulforaphane (SFN) combined with docetaxel (DCT) could enhance therapeutic efficacy. Prostate cancer cell lines LNCaP and PC-3 were treated with individual IC50 doses of SFN and DCT and half-reduced IC50 values for the SFN:DCT combination. Metabolic markers, including glucose consumption, lactate production, reactive oxygen species (ROS), mitochondrial mass, and caspase activity, were assessed. In LNCaP cells, the SFN:DCT combination reduced cell viability to 50%, comparable to DCT monotherapy (48%). Caspase 3 activation was also higher with SFN:DCT (2.4 ± 0.75 RFU) than DCT alone (2.1 ± 0.47 RFU), while caspase 8 activation remained comparable, indicating equivalent effectiveness at lower concentrations. In PC-3 cells, the combination induced caspase 3 activation (1.16 ± 0.0484 RFU) at levels slightly lower than DCT (1.51 ± 0.2062 RFU) but achieved greater reductions in mitochondrial mass, reflecting its ability to target metabolic vulnerabilities in aggressive phenotypes. Our findings suggest that the SFN:DCT combination is a promising strategy for early-stage prostate cancer. By achieving comparable efficacy to DCT monotherapy at low doses, the SFN:DCT combination maintains the therapeutic impact, mitigating the adverse effects of conventional DCT treatment.

Student attrition remains a persistent challenge in higher education and is shaped by interacting socioeconomic, academic, institutional, and wellbeing-related mechanisms. Although learning analytics and educational data mining increasingly support early-warning and intervention workflows, dataset reuse is often limited by incomplete documentation and inconsistent variable definitions. This Data Descriptor presents a structured cross-sectional survey dataset on factors influencing student persistence at a Colombian public university campus (La Paz). Data were collected between August and December 2025 through an online questionnaire and subsequently cleaned to remove duplicate entries and personally identifiable information. The released dataset contains 333 student records and 33 variables covering demographics (e.g., age, gender, first-generation status), socioeconomic conditions (e.g., residential stratum, housing, financial aid), academic experience and satisfaction (multiple 1–5 Likert items), perceived dropout intention across personal/socioeconomic/academic domains, thematically coded open-ended items describing challenges and motives, and a self-allocation of 0–100 weights across three dropout-factor domains. We provide a machine-readable codebook, a transparent preprocessing description, and technical validation checks (value ranges, category consistency, and composite-score integrity). The dataset is intended to support reproducible retention research, equity-oriented analyses, and benchmarking of predictive models, while encouraging responsible reuse through privacy-preserving release practices and FAIR-aligned metadata, repository deposition, and versioning.

In data analysis, validating the normality assumption is crucial for determining the suitability of applying parametric methods. The objective of this research was to compare the power and sensitivity of sixteen normality tests, classified according to various aspects. The methodology involved simulating data using the Fleishman contamination system. This approach allowed us to evaluate the tests under non-normality conditions across ten distributions with varying degrees of deviation from normality. The results obtained showed that tests based on correlation and regression, such as Shapiro-Wilk and Shapiro-Francia, outperform the others in power, especially for large samples and substantial deviations from normality. For moderate deviations, the D’Agostino-Pearson and skewness tests performed well, while for low deviations, the Robust Jarque-Bera and Jarque-Bera tests were the most effective. Additionally, some tests exhibited high power across multiple distribution types, such as Snedecor-Cochran and Chen-Ye, which performed well for both symmetric platykurtic and asymmetric leptokurtic distributions. These findings offer valuable insights for selecting appropriate normality tests based on sample characteristics, which improves the reliability of statistical inference. Finally, it is concluded that this research demonstrates scenarios in which the most commonly used statistical tests are not always the most effective.

Student dropout represents a significant challenge for the sustainability of higher education systems, particularly in regional public contexts where academic trajectories are heavily shaped by socioeconomic conditions. While prior research typically examines whether students consider leaving, less attention has been given to how they prioritize the factors that may lead to dropout. This study analyzes students’ perceived factors associated with dropout intention at a regional campus of the Universidad Nacional de Colombia using survey data from 287 valid constant-sum responses. The empirical approach combines descriptive statistics, compositional analysis, and $k-$means clustering on centered log-ratio coordinates. On average, students assigned the greatest weight to socioeconomic factors (42.4 points), followed by personal (32.9) and academic factors (24.6), a pattern reinforced by the compositional center (51.1%, 31.6%, and 17.3%, respectively). Students living in rented housing placed greater emphasis on socioeconomic constraints, while differences across other characteristics were modest. Cluster analysis identified one dominant mixed profile (81.2% of the sample) and several smaller edge profiles, none primarily defined by academic factors alone. These findings indicate that enrolled students perceive potential dropout as a multidimensional set of pressures rather than a single dominant cause, highlighting the need for integrated financial, psychosocial, pedagogical, and academic support strategies to strengthen educational sustainability.

Generative artificial intelligence in higher education refers here to the use of computational systems that produce text, code, explanations, feedback-like responses, images, and other outputs from user prompts in university learning, coursework, assessment, and student study practices. This entry focuses on how students use generative AI while studying, preparing assignments, seeking explanations, revising work, programming, brainstorming, or responding to assessment tasks. It defines such use as a situated educational practice shaped by disciplinary expectations, assessment design, AI literacy, study habits, and academic integrity norms. From this perspective, the same AI-supported action may be acceptable as learning support in one course, ambiguous in another, and inappropriate when it conceals authorship, fabricates evidence, or substitutes for independent academic performance.

This study develops and validates a reproducible, site-sensitive workflow for analyzing continuous electromagnetic field monitoring data published as open data by the National Spectrum Agency (ANE) of Colombia. The Bogotá subset contains 481,148 broadband electric field measurements averaged over six-minute intervals, collected by nine fixed probes between November 2, 2023, and June 30, 2024. After harmonizing timestamps and applying a completeness filter, 2,009 probe-day records were analyzed. To account for the longitudinal structure of the dataset, two ordinary least squares log-linear specifications with HC3 robust inference were compared with a random-intercept mixed-effects model, and an additional model was estimated for the daily 95th percentile. The temporal-only model explained little variation in mean daily exposure (adjusted R² = 0.007), whereas the model with probe fixed effects captured most of the variability (adjusted R² = 0.964). The mixed-effects model led to the same interpretation (marginal R² = 0.016; conditional R² = 0.964), while the p95 model retained a strongly site-dependent structure (adjusted R² = 0.951). Regression diagnostics indicated non-normality, heteroscedasticity, and positive residual autocorrelation, but no high-leverage observations or influential observation capable of explaining the high model fit were identified. The maximum six-minute field observed was 0.21011 V/m, below 1% of the public contextual reference value of 28 V/m. The differential contribution of this study is not to make a general claim that site conditions matter, but rather to validate this structure specifically for Bogotá through an auditable open-data workflow that reports complete coefficients, diagnostics, and robustness in the upper tail.