Abstract
This paper identifies and corrects invalid arguments in [Edge k-Product Cordial Labeling of Graphs, Eur. J. Pure Appl. Math., 18(2), Article No. 5887 (2025)]. In particular, several published statements are shown to be false through explicit counterexamples. Revised versions of these results are established under appropriate conditions, often involving structural properties such as girth. Correct bounds or conditions are obtained for specific graph classes, including shadow graphs, splitting graphs of stars, and path unions of cycles. These corrections clarify the limitations of the earlier claims and provide a more accurate foundation for further study of edge k-product cordial labeling.