This report demonstrates the application of propensity score matching to construct age- and sex-balanced comparison groups between Familial and Sporadic individuals using observational data. A 1:1 nearest-neighbour matching approach without replacement was implemented to reduce baseline differences and improve comparability prior to downstream analyses. The workflow is fully reproducible in R and illustrates best practices for covariate balance assessment using standardized mean differences.

This project demonstrates a reproducible analytical workflow for evaluating pre-post training outcomes based on binary assessment data. Using the PHTKS-10 dataset (N = 60), the analysis includes score derivation, reliability assessment, distributional evaluation, Wilcoxon signed-rank testing, and effect size estimation. Results show statistically significant and practically meaningful improvements across domains, illustrating how structured R-based methods can translate simple educational data into publishable quantitative evidence.

This study applies Box-Jenkins time series methodology to monthly confirmed malaria cases collected in Togo (2022-2024). ARIMA, SARIMA, and ARIMAX models were compared to evaluate temporal dynamics and forecasting performance. Results demonstrate strong annual seasonality, with SARIMA providing the most accurate and parsimonious model for predicting malaria incidence. The analysis highlights the relevance of routine surveillance data for epidemiological forecasting and public health planning in resource-constrained settings.