Model correlation techniques are essential tools in structural dynamics for comparing two different models across various applications. These techniques are used for damage detection and Structural Health Monitoring (SHM) when comparing two experimental models, and for mesh convergence investigations or other analyses when comparing two numerical models. In the case of comparing an experimental model with a numerical model, these techniques are crucial for model validation and model updating. The most used correlation techniques are the normalized relative frequency difference (NRFD) and the modal assurance criterion (MAC), which allow to detect differences in natural frequencies and mode shapes, respectively. According to structural dynamic modification theory, the perturbed mode shapes can be expressed as a linear combination of unperturbed mode shapes through a transformation matrix T. In this paper, this matrix T is proposed to identify discrepancies between two models in terms of mass and stiffness, denoted T-Mass and T-Stiffness, respectively. This work applies T-Mass and T-Stiffness to correlate an experimental scaled two-story building, which is perturbed with lumped masses. A numerical model was also assembled, which is used to predict the effect of the attached masses and to estimate modal masses needed in the T-Mass and T-Stiffness.