In this paper, the key assumptions in the M-K and effective stress ratio models are investigated for AISI 1018 steel specimens with a thickness of 0.78 mm using experimental and numerical data from Marciniak tests. The experimental procedure included Digital Imaging Correlation (DIC) to measure the major and minor in-plane strains. Strain components were obtained at points inside (i.e., the defect region) and adjacent (i.e., the safe regions) to the high strain concentrations for four different strain paths. In the numerical analysis, FEA simulations with Marc Mentat were performed with shell elements to investigate the four specimen geometries. The key assumptions of interest are the incremental major strain ratio from M-K model and the critical stress concentration factor from effective stress ratio model. Thus, the mechanics- and material-based failure phenomena in these two analytical models are examined in this paper to provide insight into the material behavior at failure. Also, data are presented that shows clearly the localization (both size and strain value) for the various strain paths.