LA RECHERCHE AGRONOMIQUE AU SENEGAL : ANALYSE CRITIQUE DES PRATIQUES METHODOLOGIQUES ET PISTES D AMELIORATION

This study on the application of statistical methods in agronomic research in Senegal highlights several key issues related to the methodological quality of scientific studies. The findings show that the Randomized Complete Block Design (52.54%) is the most frequently used experimental design, while ANOVA (43.31%) is the most commonly applied analytical method, followed by descriptive statistics. In addition, random sampling appears to be the predominant sampling technique used (75%) in the reviewed studies. However, the analysis also reveals significant methodological shortcomings. In many studies, the choice of experimental designs, sampling techniques, and data analysis methods is not clearly justified, which may undermine the validity, reproducibility, and reliability of the findings. Furthermore, most studies do not specify the formula used to determine sample size, which may reduce statistical precision and increase the risk of sampling errors. The study also points to the limited use of crop models in agricultural research, despite their importance as tools for simulation, yield prediction, and the assessment of the impacts of agricultural practices and climate variability. Therefore, strengthening methodological rigor, improving the transparency of statistical choices, and promoting the integration of crop models are essential to enhance the scientific quality and robustness of agronomic research in Senegal.