This study introduces how the asymmetry of item characteristic curves (ICC), a different change rate below versus above the inflection point of an S-shaped curve, functions more properly than other commonly used item response models in dealing with several measurement issues. Among asymmetric models, the logistic positive exponent (LPE) model of Samejima (2000) is focused on. This study mainly explores how asymmetric item response models can be extended to accommodate various psychometric applications through linking such as differential item functioning or computerized adaptive testing. From the simulation study, a linking method using concurrent calibration in common-item and non-equivalent group design worked reasonably well in defining a consistent metric compared to non-linking separate calibration, as demonstrated by ICC recovery and adequate separation of group proficiencies in a vertical scaling context, which performs linking metrics of two groups with different proficiency levels. The limitations of this research and further study suggestions are discussed in the last section.