This study demonstrates substantial methodological and statistical deficiencies that compromise its credibility. The ANOVA model employed fails to confirm the homogeneity of variance and normality assumptions, potentially invalidating the significance tests and the Tukey’s HSD comparisons (p < 0.05). The reported adjusted R² values appear overstated, given the lack of thorough error propagation and inconsistency in sample pooling across experiments. Additionally, critical information about experimental controls, including fertilization baselines and substrate variability, is insufficiently detailed, raising concerns about bias and reproducibility.
The interpretation of vegetation indices such as NDVI and GLI lacks rigor, as threshold comparisons are extrapolated from unrelated conditions and species. Furthermore, discrepancies between vegetation indices and other health metrics (e.g., NPCI, PSRI) are not addressed, casting doubt on the internal consistency of the findings. The omission of the model’s lack-of-fit assessment and the absence of replication for critical trials further diminish the study’s robustness.
Key design limitations include inadequate randomization and poorly justified experimental ranges for biostimulant concentrations, which diverge significantly from established literature. The study also fails to account for the confounding effects of substrate preparation variability, undermining the validity of treatment comparisons. Together, these flaws significantly question the reliability and broader applicability of the reported outcomes.